Repatch (from linux) of OSX IRAF

194 files changed, 23532 insertions, 0 deletions

diff --git a/pkg/system/README b/pkg/system/README
new file mode 100644
index 00000000..5567ce8b
--- /dev/null
+++ b/pkg/system/README
@@ -0,0 +1,5 @@
+
+This directory contains all CL-callable routines for the "system" package.
+Included are routines for file manipulation, date and time, device access,
+and so on.  Certain of these routines are machine dependent; search for 
+the string "MACHDEP" to identify such files.
diff --git a/pkg/system/bench.cl b/pkg/system/bench.cl
new file mode 100644
index 00000000..e8de1036
--- /dev/null
+++ b/pkg/system/bench.cl
@@ -0,0 +1,143 @@
+#{
+#  BENCH.CL
+#
+#  A rough benchmark script for testing IRAF on different systems.
+#
+
+procedure bench ()
+
+begin
+
+    string dum1, dum2, time1, time2
+    real t0, t1, t2, t3, t4, tf, tsec, tmake, tproc, tcomb, tmed
+    
+
+    # Load the needed tasks (if necessary) BEFORE timer starts.
+    if (!defpac("noao.artdata")) {
+    	artdata
+    }
+
+    reset imdir = "HDR$pix/"
+
+    
+    print(" ")
+    time() | scan (dum1, time1, dum2)
+    print("Bench started at ",time1)
+    time() | scan (dum1, t0, dum2)
+    print(" ")
+    
+    print ("=====> Making images...")
+    print ("         Making zero...")
+    mknoise ("zimage",
+        output="", title="zero image", ncols=2048, nlines=2048,
+        header="artdata$stdheader.dat", background=0., gain=1., rdnoise=2.,
+        poisson=no, seed=1, cosrays="", ncosrays=100, energy=30000., radius=0.5,
+        pa=0., comments=yes)
+    
+    print ("         Making flat...")
+    mknoise ("fimage",
+        output="", title="flat image", ncols=2048, nlines=2048,
+        header="artdata$stdheader.dat", background=30000., gain=1., rdnoise=1.,
+        poisson=no, seed=1, cosrays="", ncosrays=0, energy=30000., radius=0.5, 
+        pa=0., comments=yes)
+    
+    print ("         Making 3 objs...")
+    mknoise ("o1image",
+        output="", title="obj image", ncols=2048, nlines=2048,
+        header="artdata$stdheader.dat", background=500., gain=1., rdnoise=5.,
+        poisson=no, seed=1, cosrays="", ncosrays=100, energy=30000., radius=0.5,
+        pa=0., comments=yes)
+    
+    mknoise ("o2image",
+        output="", title="obj image", ncols=2048, nlines=2048,
+        header="artdata$stdheader.dat", background=500., gain=1., rdnoise=5.,
+        poisson=no, seed=1, cosrays="", ncosrays=100, energy=30000., radius=0.5,
+        pa=0., comments=yes)
+    
+    mknoise ("o3image",
+        output="", title="obj image", ncols=2048, nlines=2048,
+        header="artdata$stdheader.dat", background=500., gain=1., rdnoise=5.,
+        poisson=no, seed=1, cosrays="", ncosrays=100, energy=30000., radius=0.5,
+        pa=0., comments=yes)
+    
+    time() | scan (dum1, t1, dum2)
+    
+
+    # PROC SIMULATION SECTION
+    #
+    # Flat normalization is usually done realtime in memory in ccdproc,
+    # but writing the image might be similar to the temp image in ccdproc
+    
+    print(" ")
+    print(" ")
+    print ("=====>   Normalizing flat...")
+    imarith ("fimage",
+        "/", "30000.", "fimage", title="", divzero=1., hparams="", pixtype="",
+        calctype="", verbose=yes, noact=no)
+    
+    print(" ")
+    print(" ")
+    print ("=====>   Subtracting zero from 3 images...")
+    imarith ("o1image,o2image,o3image",
+        "-", "zimage", "o1image,o2image,o3image", title="", 
+        divzero=0., hparams="", pixtype="", calctype="", verbose=yes, noact=no)
+    
+    print(" ")
+    print(" ")
+    print ("=====>   Dividing flat into 3 images...")
+    imarith ("o1image,o2image,o3image",
+        "/", "fimage", "o1image,o2image,o3image", title="", divzero=0., 
+        hparams="", pixtype="", calctype="", verbose=yes, noact=no)
+    
+    # END of PROC SIMULATION SECTION
+    
+    time() | scan (dum1, t2, dum2)
+    
+    print(" ")
+    print(" ")
+    printf ("=====>   Combining 3 images...")
+    imcombine ("o1image,o2image,o3image",
+        "ocimage", sigma="", logfile="STDOUT", combine="average",
+        reject="crreject", project=no, outtype="real", offsets="none",
+        masktype="none", maskvalue=0., blank=0., scale="none", zero="none",
+        weight="none", statsec="", expname="", lthreshold=INDEF, 
+	hthreshold=INDEF, nlow=1, nhigh=1, nkeep=1, mclip=yes, lsigma=3., 
+	hsigma=3., rdnoise="5", gain="1", snoise="0.", sigscale=0.1, 
+	pclip=-0.5, grow=0)
+    
+    time() | scan (dum1, t3, dum2)
+    
+    print(" ")
+    print(" ")
+    print ("=====>   Median filtering image...")
+    median ("ocimage",
+        "omimage", 9, 9, zloreject=INDEF, zhireject=INDEF, boundary="nearest",
+        constant=0., verbose=yes)
+    
+    time() | scan (dum1, t4, dum2)
+    
+    print(" ")
+    print ("=====>   Deleting all images...")
+    imdel ("o1image,o2image,o3image,ocimage,omimage,zimage,fimage",
+        yes, verify=no, default_acti=yes)
+    
+    time() | scan (dum1, tf, dum2)
+    time() | scan (dum1, time2, dum2)
+    print(" ")
+    print("Bench started at ",time1)
+    print("Bench  ended at  ",time2)
+    print(" ")
+    
+    tsec  = (tf - t0) * 3600.
+    tmake = (t1 - t0) * 3600.
+    tproc = (t2 - t1) * 3600.
+    tcomb = (t3 - t2) * 3600.
+    tmed  = (t4 - t3) * 3600.
+    
+
+    printf ("Total execution time = %7.1f seconds\n",tsec)
+    printf (" Total time    Make 5 imgs    Proc 3 imgs    Combine 3 imgs    Median 1 img\n")
+    printf ("  %7.1f        %7.1f       %7.1f         %7.1f           %7.1f\n",
+	tsec, tmake, tproc, tcomb, tmed)
+    
+end
diff --git a/pkg/system/chkupdate.par b/pkg/system/chkupdate.par
new file mode 100644
index 00000000..f6800a8d
--- /dev/null
+++ b/pkg/system/chkupdate.par
@@ -0,0 +1,5 @@
+interval,i,h,0,,,Days between update checks (-1 to disable or 0 for always)
+ref_file,s,h,"iraf$.release_date",,,reference file for times
+release,s,h,)_.release,,,IRAF release version
+baseurl,s,h,"http://iraf.noao.edu/ftp",,,base URL to release timestamp directory
+verbose,b,h,yes,,,verbose output flag
diff --git a/pkg/system/chkupdate.x b/pkg/system/chkupdate.x
new file mode 100644
index 00000000..a45c6d44
--- /dev/null
+++ b/pkg/system/chkupdate.x
@@ -0,0 +1,178 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include <fio.h>
+include <fset.h>
+include <finfo.h>
+include <time.h>
+include <syserr.h>
+
+
+
+# CHKUPDATE -- Check to see whether the current IRAF system is up to date
+# relative to what's available on the NOAO server.
+
+
+procedure t_chkupdate ()
+
+
+char	version[SZ_FNAME], baseurl[SZ_FNAME], chkfile[SZ_LINE]
+char	netpath[SZ_LINE], arch[SZ_FNAME], tmpfile[SZ_FNAME]
+char	host[SZ_LINE], ref_file[SZ_LINE], release[SZ_LINE]
+char	buf[SZ_LINE]
+pointer	reply
+bool	verbose
+long	last, reldate, ndays, info[LEN_FINFO]
+int	ip, op, fd, nread, interval, tm[LEN_TMSTRUCT]
+
+int	clgeti(), envgets(), access(), open(), url_get()
+int	strlen(),  getline(), finfo()
+bool	clgetb()
+long	clktime(), ctol()
+
+define	done_		99
+
+begin
+	# Initialize strings.
+	call aclrc (baseurl, SZ_LINE)
+	call aclrc (release, SZ_LINE)
+	call aclrc (ref_file, SZ_LINE)
+	call aclrc (chkfile, SZ_LINE)
+	call aclrc (tmpfile, SZ_FNAME)
+	call aclrc (arch, SZ_FNAME)
+	call aclrc (version, SZ_FNAME)
+
+	call strcpy ("uparm$update", chkfile, SZ_LINE)
+	call mktemp ("tmp$url", tmpfile, SZ_FNAME)
+	if (envgets ("arch", arch, SZ_FNAME) == ERR) {
+	    if (verbose)
+	        call eprintf ("Error: cannot get architecture.\n")
+	    return 
+	}
+
+
+	# Get the task parameters.
+	call clgstr ("ref_file", ref_file, SZ_LINE)
+	call clgstr ("release", release, SZ_LINE)
+	call clgstr ("baseurl", baseurl, SZ_FNAME)
+	verbose = clgetb ("verbose")
+	interval = clgeti ("interval")
+
+
+	# See whether we're doing an update check.  If the interval is less
+	# than zero it means the update is disabled.  A zero value means we
+	# check at each login, otherwise we've specified the number of days
+	# since our last check.  The check time is kept in the uparm$update
+	# file which is created the first time we run, and updated after each
+	# check.
+	if (interval < 0) {
+	    return				# nothing to do
+
+	} else if (interval >= 0) {
+	    # See if current time is more than N interval days since last
+	    # check.
+	    if (access (chkfile, 0, 0) == NO) {
+	        # No update file found, create one and check for updates.
+	        fd = open (chkfile, NEW_FILE, TEXT_FILE)
+	        call fprintf (fd, "%d\n")
+		    call pargl (clktime (0))
+	        call close (fd)
+	    } else {
+		call aclrc (buf, SZ_LINE)
+	        fd = open (chkfile, READ_ONLY, TEXT_FILE)
+		if (getline (fd, buf) != EOF) {
+		    ip = 1
+		    if (ctol (buf, ip, last) > 0) {
+			ndays = (clktime(0) - last) / 86400
+			if (ndays < interval) {
+	        	    call close (fd)
+			    return		# too recent
+			}
+		    }
+		}
+	        call close (fd)
+	    }
+	}
+
+
+	# Transform the CL version string to one we can use in the URL.
+	version[1] = 'v' ; op = 2
+	for (ip=1; release[ip] != NULL && ip < SZ_LINE; ip=ip+1) {
+	    if (release[ip] != '.') {
+		version[op] = release[ip]
+		op = op + 1
+	    }
+	}
+
+	# Create the URL to the release timestamp file.
+	call sprintf (netpath, SZ_LINE, "%s/%s/releases/%s")
+	    call pargstr (baseurl)
+	    call pargstr (version)
+	    call pargstr (arch[2])
+
+
+	# Access the URL and get the reply.
+	call calloc (reply, SZ_LINE, TY_CHAR)
+
+	if (url_get (netpath, tmpfile, reply) > 0) {
+	    call aclrc (buf, SZ_LINE)
+	    fd = open (tmpfile, READ_ONLY, TEXT_FILE)
+	    if (getline (fd, buf) != EOF) {
+		ip = 1
+		if (ctol (buf, ip, reldate) > 0) {
+	    	    if (access (ref_file, 0, 0) == NO) {
+			if (verbose)
+			    call eprintf ("Error: no release file\n")
+	        	call close (fd)
+			goto done_
+	    	    }
+	    	    if (access (ref_file, 0, 0) == YES &&
+		        finfo (ref_file, info) != ERR) {
+
+			call brktime (reldate, tm)
+			call printf ("  ***  Checking update status... ")
+
+			# Add a 1-day offset to the release date to ensure
+			# the ref_file timestamp is less than the URL file,
+			# both of which are created during the packaging of
+			# a release.  This avoids a case where we would
+			# *always* get an update notification.
+			reldate = reldate + 86400
+
+		        if (FI_CTIME(info) < reldate) {
+
+			    call aclrc (release, SZ_LINE)
+			    call strcpy (buf[ip+1], release, SZ_LINE)
+			    release[strlen(release)] = NULL
+			
+			    call printf (
+			        "IRAF %s update available on %d/%d/%d\n")
+				    call pargstr (release)
+				    call pargi (TM_MONTH(tm))
+				    call pargi (TM_MDAY(tm))
+				    call pargi (TM_YEAR(tm))
+	        	    call close (fd)
+			    goto done_
+		        } else {
+			    call printf ("Your IRAF system is up to date\n")
+		        }
+		    }
+		}
+	    }
+	    call close (fd)
+
+	} else 
+	    ; 					# got a HTTP error code, quit
+
+
+	# Write the current check time to the update file to reset the
+	# interval.
+	fd = open (chkfile, READ_WRITE, TEXT_FILE)
+	call fprintf (fd, "%d\n")
+	    call pargl (clktime (0))
+	call close (fd)
+
+	# Clean up.
+done_ 	call mfree (reply, TY_CHAR)
+	if (access (tmpfile, 0, 0) == YES) 
+	    call delete (tmpfile)
+end
diff --git a/pkg/system/cmdstr.par b/pkg/system/cmdstr.par
new file mode 100644
index 00000000..ba046633
--- /dev/null
+++ b/pkg/system/cmdstr.par
@@ -0,0 +1,2 @@
+task,s,a,,,,Task name
+hidden,b,h,no,,,Print hidden parameters
diff --git a/pkg/system/cmdstr.x b/pkg/system/cmdstr.x
new file mode 100644
index 00000000..be4e965e
--- /dev/null
+++ b/pkg/system/cmdstr.x
@@ -0,0 +1,157 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+
+define	SZ_CMDSTR	4096
+
+# CMDSTR -- Read the output of LPARAM and format a command string giving
+# the values of all parameters.
+
+procedure t_cmdstr()
+
+bool	hidden, hparam
+pointer	sp, ltask, pname, ibuf, obuf, ip, op, pp, nl, last
+int	getline(), gstrcpy(), strncmp()
+bool	clgetb()
+
+begin
+	call smark (sp)
+	call salloc (ltask, SZ_FNAME, TY_CHAR)
+	call salloc (pname, SZ_FNAME, TY_CHAR)
+	call salloc (ibuf, SZ_LINE, TY_CHAR)
+	call salloc (obuf, SZ_CMDSTR, TY_CHAR)
+
+	# Get the task name and whether to print hidden parameters.
+	call clgstr ("task", Memc[ltask], SZ_FNAME)
+	hidden = clgetb ("hidden")
+
+	op = obuf + gstrcpy (Memc[ltask], Memc[obuf], SZ_CMDSTR)
+	Memc[op] = ' ';  op = op + 1
+	Memc[op] = '(';  op = op + 1
+
+	last = NULL
+	nl   = NULL
+
+	while (getline (STDIN, Memc[ibuf]) != EOF) {
+	    ip = ibuf
+
+	    # Skip white space.
+	    while (IS_WHITE (Memc[ip]))
+		ip = ip + 1
+
+	    # Check if the parameter is hidden and skip it if desired.
+	    if (Memc[ip] == '(') {
+		if (!hidden)
+		    next
+		hparam = true
+		ip = ip + 1
+	    } else
+		hparam = false
+
+	    # Check if parameter name is "mode" and skip it.
+	    if (strncmp (Memc[ip], "mode =", 6) == 0)
+		next
+
+	    # Copy or skip parameter name.
+	    pp = pname
+	    while (!IS_WHITE (Memc[ip])) {
+		if (hparam) {
+		    Memc[op] = Memc[ip]
+		    op = op + 1
+	        }
+		Memc[pp] = Memc[ip]
+		pp = pp + 1
+		ip = ip + 1
+	    }
+	    Memc[pp] = EOS
+
+	    # Copy or skip = and skip whitespace.
+	    if (hparam) {
+		Memc[op] = '='
+		op = op + 1
+	    }
+	    ip = ip + 3
+
+	    # Copy parameter value.  It is an error if there is no value.
+	    if (IS_WHITE (Memc[ip]) || (Memc[ip] == ')' && Memc[ip+1] != '_')) {
+		call sprintf (Memc[obuf], SZ_CMDSTR,
+		    "Undefined parameter value (%s.%s)")
+		    call pargstr (Memc[ltask])
+		    call pargstr (Memc[pname])
+		call error (1, Memc[obuf])
+	    }
+
+	    # If the parameter is a quoted string copy until the closing quote,
+	    # otherwise copy until whitespace or ).
+
+	    if (Memc[ip] == '"') {
+		Memc[op] = Memc[ip]
+		ip = ip + 1
+		op = op + 1
+		while (Memc[ip] != '"') {
+		    Memc[op] = Memc[ip]
+		    ip = ip + 1
+		    op = op + 1
+		}
+		Memc[op] = Memc[ip]
+		ip = ip + 1
+		op = op + 1
+
+	    } else if (Memc[ip] == ')' && Memc[ip+1] == '_') {
+		# If the value is a redirection, e.g. ")_.foo", add quotes
+		# around the value and copy as a special case.
+		Memc[op] = '"'
+		op = op + 1
+
+		# Copy the opening paren.
+		Memc[op] = Memc[ip]
+		op = op + 1
+		ip = ip + 1
+
+		# Copy the rest of the string.
+		while (!IS_WHITE(Memc[ip]) && (Memc[ip] != ')')) {
+		    Memc[op] = Memc[ip]
+		    ip = ip + 1
+		    op = op + 1
+		}
+
+		# Add the closing quote.
+		Memc[op] = '"'
+		op = op + 1
+
+	    } else {
+		while (!IS_WHITE(Memc[ip]) && (Memc[ip] != ')')) {
+		    Memc[op] = Memc[ip]
+		    ip = ip + 1
+		    op = op + 1
+		}
+	    }
+
+	    # Add a comma and a space.
+	    Memc[op] = ','
+	    op = op + 1
+	    Memc[op] = ' '
+	    op = op + 1
+
+	    # Replace the last break with a new line if the line exceeds max.
+	    if ((op - nl > 80) && (last > 0)) {
+		Memc[last] = '\n'
+		nl = last
+	    }
+	    last = op - 1
+	}
+
+	# Replace last comma and space by a parenthesis and EOS.
+	if (Memc[op-2] == ',')
+	    op = op - 2
+
+	Memc[op] = ')'
+	op = op + 1
+	Memc[op] = EOS
+
+	# Print the command string and finish up.
+	call putline (STDOUT, Memc[obuf])
+	call putci   (STDOUT, '\n')
+
+	call sfree (sp)
+end
diff --git a/pkg/system/concatenate.par b/pkg/system/concatenate.par
new file mode 100644
index 00000000..533461dd
--- /dev/null
+++ b/pkg/system/concatenate.par
@@ -0,0 +1,4 @@
+input_files,s,a,,,,list of files to be concatenated
+output_file,f,a,,,,name of output file
+out_type,s,h,"in_type",,,output file type (text|binary)
+append,b,h,no,,,append to output file?
diff --git a/pkg/system/concatenate.x b/pkg/system/concatenate.x
new file mode 100644
index 00000000..7732150b
--- /dev/null
+++ b/pkg/system/concatenate.x
@@ -0,0 +1,73 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# CONCATENATE -- Concatenate the named input files (or the standard input) to
+# produce the output file.  The input files must all be of the same type.
+# The type of the output file (text|binary) defaults to the type of the first
+# input file, or to text if reading from the standard input.  The hidden
+# parameter "out_type" may be used to force the file type of the copy operation.
+
+procedure t_concatenate()
+
+char	infile[SZ_FNAME], outfile[SZ_FNAME]
+char	out_type_string[SZ_FNAME]
+int	in, out, list, out_type
+
+int	clpopni(), clgfil()
+int	open(), access(), strmatch(), clgeti()
+bool	clgetb()
+
+begin
+	# Open the input file list, and read in the name of the first file
+	# (used to determine type of output file later).
+
+	list = clpopni ("input_files")
+	if (clgfil (list, infile, SZ_FNAME) == EOF) {
+	    call clpcls (list)
+	    return
+	}
+
+	# Determine whether the output file is text or binary, and open
+	# the file.
+
+	call clgstr ("out_type", out_type_string, SZ_FNAME)
+	out_type = NULL
+	if (strmatch (out_type_string, "^#{b}") > 0)
+	    out_type = BINARY_FILE
+	else if (strmatch (out_type_string, "^#{t}") > 0)
+	    out_type = TEXT_FILE
+
+	# Default to std output if output file not given on cmd line.
+
+	if (clgeti ("$nargs") <= 1)
+	    call strcpy ("STDOUT", outfile, SZ_FNAME)
+	else
+	    call clgstr ("output_file", outfile, SZ_FNAME)
+
+	# If the output file type has been specified (param "out_type"),
+	# use it, otherwise use type of first input file.
+
+	if (out_type == NULL) {
+	    if (strmatch (infile, "STDIN") > 0)
+		out_type = TEXT_FILE
+	    else if (access (infile, 0, TEXT_FILE) == YES)
+		out_type = TEXT_FILE
+	    else
+		out_type = BINARY_FILE
+	}
+
+	# Finally, open the output file.
+	if (clgetb ("append"))
+	    out = open (outfile, APPEND, out_type)
+	else
+	    out = open (outfile, NEW_FILE, out_type)
+
+	# Append the input files to the output file.
+	repeat {
+	    in = open (infile, READ_ONLY, out_type)
+	    call fcopyo (in, out)
+	    call close (in)
+	} until (clgfil (list, infile, SZ_FNAME) == EOF)
+
+	call close (out)
+	call clpcls (list)
+end
diff --git a/pkg/system/copy.par b/pkg/system/copy.par
new file mode 100644
index 00000000..eb33a234
--- /dev/null
+++ b/pkg/system/copy.par
@@ -0,0 +1,3 @@
+input,s,a,,,,input file or file template
+output,s,a,,,,output file or destination directory
+verbose,b,h,no,,,print names of files as they are copied
diff --git a/pkg/system/copy.x b/pkg/system/copy.x
new file mode 100644
index 00000000..2a021516
--- /dev/null
+++ b/pkg/system/copy.x
@@ -0,0 +1,62 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+
+# COPY -- Copy a file, or copy a list of files to another directory.
+# The number of input files is variable, but there can be only a single
+# output file.  If the output file is a directory, we copy the input files
+# to that directory without changing their names.  Otherwise there can
+# be only one input file, and a copy is made with the new name.
+
+procedure t_copy()
+
+int	list, root_len
+bool	verbose
+pointer	sp, infile, destination, outfile, junkstr, dirname
+bool	clgetb()
+int	clpopni(), clgfil(), clplen(), fnldir(), isdirectory()
+
+begin
+	call smark (sp)
+	call salloc (infile, SZ_FNAME, TY_CHAR)
+	call salloc (outfile, SZ_FNAME, TY_CHAR)
+	call salloc (destination, SZ_FNAME, TY_CHAR)
+	call salloc (dirname, SZ_PATHNAME, TY_CHAR)
+	call salloc (junkstr, SZ_FNAME, TY_CHAR)
+
+	list = clpopni ("input")
+	call clgstr ("output", Memc[destination], SZ_FNAME)
+	verbose = clgetb ("verbose")
+
+	# If the destination file is a directory, copy each file in the input
+	# list to the destination directory, else just copy the file and check
+	# that there is only one file in the input list.
+
+	if (isdirectory (Memc[destination], Memc[dirname], SZ_PATHNAME) > 0) {
+
+	    while (clgfil (list, Memc[infile], SZ_FNAME) != EOF) {
+		call strcpy (Memc[dirname], Memc[outfile], SZ_PATHNAME)
+		root_len = fnldir (Memc[infile], Memc[junkstr], SZ_FNAME)
+		call strcat (Memc[infile+root_len], Memc[outfile], SZ_PATHNAME)
+
+		if (verbose) {
+		    call eprintf ("%s -> %s\n")
+			call pargstr (Memc[infile])
+			call pargstr (Memc[outfile])
+		}
+
+		iferr (call fcopy (Memc[infile], Memc[outfile]))
+		    call erract (EA_WARN)
+	    }
+
+	} else if (clgfil (list, Memc[infile], SZ_FNAME) != EOF) {
+	    if (clplen (list) > 1) {
+		call clpcls (list)
+		call error (2, "cannot copy several files to a single file")
+	    }
+	    call fcopy (Memc[infile], Memc[destination])
+	}
+
+	call clpcls (list)
+	call sfree (sp)
+end
diff --git a/pkg/system/count.par b/pkg/system/count.par
new file mode 100644
index 00000000..bc78c6ef
--- /dev/null
+++ b/pkg/system/count.par
@@ -0,0 +1 @@
+files,s,a,,,,list of files to be counted
diff --git a/pkg/system/count.x b/pkg/system/count.x
new file mode 100644
index 00000000..9f9ef51c
--- /dev/null
+++ b/pkg/system/count.x
@@ -0,0 +1,131 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	<chars.h>
+include	<error.h>
+
+define	SZ_LUT		256		# "whitespace" lookup table
+define	CC_WHITE	1		# "white" character class
+define	CC_NONWHITE	0		# "word" character
+
+# COUNT -- Count the number of lines, words, and characters in the named text
+# files, or in the standard input.  Print the results on the standard output.
+# 
+# output format (single file):
+# 
+# 	nlines nwords nchars
+# 
+# output format (several files):
+# 
+# 	nlines nwords nchars filename1
+# 	nlines nwords nchars filename2
+# 	nlines nwords nchars Total
+
+procedure t_count()
+
+char	fname[SZ_FNAME]
+long	nlines, nwords, nchars
+long	totlines, totwords, totchars
+int	nfiles, list
+int	clpopni(), clgfil()
+
+begin
+	totlines = 0
+	totwords = 0
+	totchars = 0
+	nfiles = 0
+
+	list = clpopni ("files")
+
+	while (clgfil (list, fname, SZ_FNAME) != EOF) {
+	    iferr (call count_file (fname, nlines, nwords, nchars))
+		call erract (EA_WARN)
+	    else {
+		call print_stats (STDOUT, nlines, nwords, nchars, fname)
+		call flush (STDOUT)
+	    }
+
+	    totlines = totlines + nlines
+	    totwords = totwords + nwords
+	    totchars = totchars + nchars
+	    nfiles = nfiles + 1
+	}
+
+	call clpcls (list)
+
+	if (nfiles > 1)
+	    call print_stats (STDOUT, totlines, totwords, totchars, "Total")
+end
+
+
+# COUNT_FILE -- Open a file and count the number of lines, words, and
+# characters in the file.
+
+procedure count_file (fname, nlines, nwords, nchars)
+
+char	fname[SZ_FNAME], lbuf[SZ_LINE], ch, class[SZ_LUT]
+long	nlines, nwords, nchars
+int	first_time, state, fd, ip, open(), getline()
+errchk	open, getline
+data	first_time /YES/
+
+begin
+	# Initialize the lookup table, used to count words.  In this case,
+	# NEWLINE is considered to be whitespace.
+
+	if (first_time == YES) {
+	    do ip = 1, SZ_LUT
+		class[ip] = CC_NONWHITE
+
+	    class[BLANK] = CC_WHITE
+	    class[TAB] = CC_WHITE
+	    class[NEWLINE] = CC_WHITE
+
+	    first_time = NO
+	}
+
+	fd = open (fname, READ_ONLY, TEXT_FILE)
+
+	nwords = 0
+	nchars = 0
+	state = CC_WHITE
+
+	# Increment word count at the beginning of every word.  A "word"
+	# is defined as a sequence of characters delimited by whitespace.
+	# COUNT does not know anything about quoted strings.
+
+	for (nlines=0;  getline (fd, lbuf) != EOF;  ) {
+	    do ip = 1, SZ_LINE {
+		ch = lbuf[ip]
+		if (ch == EOS)
+		    break
+		else if (class[ch] != state) {
+		    nwords = nwords + state
+		    state = class[ch]
+		}
+	    }
+	    nchars = nchars + ip - 1
+	    if (lbuf[ip-1] == '\n')
+		nlines = nlines + 1
+	}   
+
+	call close (fd)
+end
+
+
+# PRINT_STATS -- Format the COUNT statistics summary line and print it on
+# the named file.
+
+procedure print_stats (fd, nlines, nwords, nchars, fname)
+
+int	fd
+long	nlines, nwords, nchars
+char	fname[ARB]
+
+begin
+	call fprintf (fd, "%7d %7d %7d %s\n")
+	    call pargl (nlines)
+	    call pargl (nwords)
+	    call pargl (nchars)
+	    call pargstr (fname)
+end
diff --git a/pkg/system/delete.par b/pkg/system/delete.par
new file mode 100644
index 00000000..a33ed606
--- /dev/null
+++ b/pkg/system/delete.par
@@ -0,0 +1,6 @@
+files,s,a,,,,list of files to be deleted
+verify,b,h,no,,,verify operation before deleting each file?
+default_action,b,h,yes,,,default delete action for verify query
+allversions,b,h,yes,,,delete all versions of each file
+subfiles,b,h,yes,,,delete any subfiles of each file
+go_ahead,b,q,yes,,," ?"
diff --git a/pkg/system/delete.x b/pkg/system/delete.x
new file mode 100644
index 00000000..0c15a2c4
--- /dev/null
+++ b/pkg/system/delete.x
@@ -0,0 +1,74 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+
+# DELETE -- Delete a list of files.  If file cannot be deleted, warn but do
+# not abort.  Verify before deleting each file if user wishes.
+
+procedure t_delete()
+
+bool	verify			# verify before deleting
+bool	allversions		# delete all versions of each file
+bool	subfiles		# delete any subfiles of a file
+char	fname[SZ_FNAME]
+int	list
+pointer	tty
+
+bool	clgetb()
+int	clpopns(), clgfil(), access(), btoi(), strncmp()
+pointer	ttyodes()
+
+begin
+	list = clpopns ("files")
+	verify = clgetb ("verify")
+	allversions = clgetb ("allversions")
+	subfiles = clgetb ("subfiles")
+	if (verify)
+	    tty = ttyodes ("terminal")
+
+	while (clgfil (list, fname, SZ_FNAME) != EOF) {
+	    if (verify) {
+		# If file does not exist, warn user (since verify mode is
+		# in effect).
+		if (access (fname, 0, 0) == NO) {
+		    call eprintf ("Warning: Nonexistent file '%s'\n")
+			call pargstr (fname)
+		    next
+		}
+
+		# Set default action of verify prompt (override learning of
+		# most recent response).
+
+		call flush (STDOUT)
+		call clputb ("go_ahead", clgetb ("default_action"))
+		call eprintf ("Delete file ")
+		call ttyso (STDERR, tty, YES)
+		call eprintf ("'%s'")
+		    call pargstr (fname)
+		call ttyso (STDERR, tty, NO)
+		call flush (STDERR)
+
+		if (!clgetb ("go_ahead"))
+		    next
+	    }
+
+	    if (strncmp ("http:", fname, 5) == 0) {
+		call eprintf ("Cannot delete URL `%s'\n")
+		    call pargstr (fname)
+		next
+	    }
+
+	    iferr (call deletefg (fname, btoi(allversions), btoi(subfiles)))
+		call erract (EA_WARN)
+	}
+
+	# Reset the go_ahead parameter, overiding learn mode, in case delete
+	# is subsequently called from the background.  Close tty descriptor.
+
+	if (verify) {
+	    call clputb ("go_ahead", true)
+	    call ttycdes (tty)
+	}
+
+	call clpcls (list)
+end
diff --git a/pkg/system/devices.cl b/pkg/system/devices.cl
new file mode 100644
index 00000000..660c91ec
--- /dev/null
+++ b/pkg/system/devices.cl
@@ -0,0 +1,3 @@
+# DEVICES -- Print information on the locally available devices.
+
+help devices
diff --git a/pkg/system/directory.par b/pkg/system/directory.par
new file mode 100644
index 00000000..444fe325
--- /dev/null
+++ b/pkg/system/directory.par
@@ -0,0 +1,6 @@
+files,s,a,,,,files or directory to be listed
+long,b,h,no,,,long format listing
+ncols,i,h,0,0,,number of columns if multicolumn format
+maxch,i,h,18,,,max filename chars to show if multicolumn format
+sort,b,h,yes,,,sort the file list
+all,b,h,no,,,list hidden files too
diff --git a/pkg/system/directory.x b/pkg/system/directory.x
new file mode 100644
index 00000000..37a1c397
--- /dev/null
+++ b/pkg/system/directory.x
@@ -0,0 +1,561 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<diropen.h>
+include	<protect.h>
+include	<finfo.h>
+include	<chars.h>
+include	<ctype.h>
+include	<fset.h>
+include	<time.h>
+
+# DIRECTORY -- List the files matching the given template.  If no template is
+# given list the contents of the current directory.  Only existing files are
+# listed, i.e., "dir file" may be used to test if "file" exists.  There are
+# two types of file listings:
+#
+#	tabular		multicolumn listing, "ncolumns" columns (default)
+#	long		long form listing; prints more info for each file
+#
+# The template may be either the name of a file or directory, or a true
+# filename template containing pattern matching metacharacters.
+
+define	DEF_MAXFILES	256		# initial max file limit
+define	DEF_SZSBUF	4096		# initial string buffer size
+define	DEF_SCREENWIDTH	80		# default screen width if envgeti fails
+define	DEF_NCOLS	4		# default number of columns in table
+define	NLINES_FLUSH	4		# nlines flushed out at a time (quick)
+define	PATCHARS	"*?["
+
+define	LEN_FLDES	6		# file list descriptor
+define	FL_NFILES	Memi[$1]	# number of files in list
+define	FL_NEXTOFF	Memi[$1+1]	# next offset in sbuf
+define	FL_OFFBP	Memi[$1+2]	# pointer to list of offsets
+define	FL_SBUFP	Memi[$1+3]	# pointer to string buffer
+define	FL_MAXFILES	Memi[$1+4]	# current length of offset array
+define	FL_SZSBUF	Memi[$1+5]	# current string buffer size
+
+define	FL_OFFSET	Memi[FL_OFFBP($1)+($2)-1]
+define	FL_FNAME	Memc[FL_SBUFP($1)+(FL_OFFSET($1,$2))-1]
+
+
+# T_DIRECTORY -- CL callable main routine for the directory program.
+
+procedure t_directory()
+
+pointer	sp, files, fname, dirname, patp, fp, ip, op, ep
+int	ncols, maxch, dirmode, fd, n, i, patlen, len_dir
+bool	long_format, is_template, is_pattern, is_dir, sort_list, match_extension
+
+bool	clgetb(), strne()
+int	clgeti(), fntopnb(), fntgfnb()
+int	diropen(), getline(), stridx(), strlen(), strncmp(), stridxs()
+int	isdirectory(), access(), envgeti(), btoi()
+string	patchars PATCHARS
+define	template_ 91
+define	done_ 92
+
+begin
+	call smark (sp)
+	call salloc (files, SZ_LINE, TY_CHAR)
+	call salloc (fname, SZ_PATHNAME, TY_CHAR)
+	call salloc (dirname, SZ_PATHNAME, TY_CHAR)
+
+	# If directory is called without any arguments, list the contents
+	# of the current directory.  Otherwise read in the file template.
+
+	if (clgeti ("$nargs") == 0)
+	    call strcpy (".", Memc[files], SZ_LINE)
+	else
+	    call clgstr ("files", Memc[files], SZ_LINE)
+
+	# Determine whether a long format listing is desired.
+	long_format = clgetb ("long")
+
+	# If not long format listing, determine the number of columns for
+	# the multicolumn format listing.
+
+	if (!long_format)
+	    ncols = clgeti ("ncols")
+
+	# Max chars of each filename to be shown.
+	maxch = clgeti ("maxch")
+
+	# Sort files?
+	sort_list = clgetb ("sort")
+
+	# Show hidden files?
+	if (clgetb ("all"))
+	    dirmode = PASS_HIDDEN_FILES
+	else
+	    dirmode = SKIP_HIDDEN_FILES
+
+	# Allocate and initialize storage for the file list descriptor and
+	# assocated array storage.
+	
+	call calloc (fp, LEN_FLDES, TY_STRUCT)
+	call malloc (FL_OFFBP(fp), DEF_MAXFILES, TY_INT)
+	call malloc (FL_SBUFP(fp), DEF_SZSBUF, TY_CHAR)
+
+	FL_MAXFILES(fp) = DEF_MAXFILES
+	FL_SZSBUF(fp)   = DEF_SZSBUF
+	FL_NEXTOFF(fp)  = 1
+
+	# The template is considered a filename template if it contains any
+	# of the metacharacters "*?[],", i.e., any string containing a comma
+	# is considered a filename template, since we don't want to mess with
+	# parsing comma delimited lists here.
+
+	is_template = false
+	is_pattern  = false
+
+	for (ip=files;  Memc[ip] != EOS;  ip=ip+1)
+	    if (IS_ALPHA (Memc[ip])) {
+		next
+	    } else if (stridx (Memc[ip], patchars) > 0) {
+		if (ip == files || Memc[ip-1] != '\\')
+		    is_pattern  = true
+	    } else if (Memc[ip] == ',' || Memc[ip] == '%')
+		if (ip == files || Memc[ip-1] != '\\')
+		    is_template = true
+
+	# Set IS_DIR flag, and get directory name if not fancy template.
+
+	if (is_template) {
+	    is_dir = false
+
+	} else if (is_pattern) {
+	    is_dir = true
+	    patp = NULL
+	    ep = NULL
+
+	    # Search for a valid directory prefix.
+	    op = dirname
+	    for (ip=files;  Memc[ip] != EOS;  ip=ip+1) {
+		Memc[op] = Memc[ip]
+		if (Memc[op] == FNLDIR_CHAR || Memc[op] == '//')
+		    if (op == dirname || Memc[op-1] != '\\') {
+			patp = ip + 1
+			ep = op + 1
+		    }
+		op = op + 1
+	    }
+
+	    # Must set patp and dirname before exiting.
+	    if (patp == NULL) {
+		patp = files
+		call strcpy ("./", Memc[dirname], SZ_PATHNAME)
+		len_dir = 0
+	    } else {
+		Memc[ep] = EOS
+		len_dir = strlen (Memc[dirname])
+	    }
+
+	} else
+	    is_dir = (isdirectory (Memc[files],Memc[dirname],SZ_PATHNAME) > 0)
+
+	# If sorting is not desired and we either don't have a template, or
+	# a simple unary match-pattern template, we can list the directory
+	# directly to the standard output as it is read.  This is the fastest
+	# possible type of of directory listing, and is particularly useful
+	# for listing very large directories.
+
+	if (is_dir && !sort_list && !long_format && !is_template) {
+	    fd = diropen (Memc[dirname], dirmode)
+	    call dir_quicklist (STDOUT, fd, Memc[patp], is_pattern, ncols)
+	    call close (fd)
+	    goto done_
+	}
+
+	# If `files' is a template, expand the template and read the file
+	# names into memory.  If `files' is not a filename template, determine
+	# if it is the name of (or pathname to) a directory.  If so, open the
+	# directory directly and read the file list into memory.  If `files' is
+	# neither a template nor a directory, it must be a filename, so merely
+	# put its name in the file list. 
+
+	if (!is_template && is_dir) {
+	    # Read file list from the named directory.
+
+	    # Set up the pattern matching code.  We recognize selecting all
+	    # files with a particular extension as a special case, since this
+	    # case is very common and can be done more efficiently if we don't
+	    # use the general pattern matching code.
+
+	    if (is_pattern) {
+		match_extension = (strncmp (Memc[patp], "*.", 2) == 0 &&
+		    stridxs (patchars, Memc[patp+2]) <= 0)
+		if (match_extension)
+		    patlen = strlen (Memc[patp])
+		else
+		    goto template_
+	    }
+
+	    fd = diropen (Memc[dirname], dirmode)
+
+	    # Accumulate the file list in memory.
+	    n = 0
+	    while (n != EOF) {
+		n = getline (fd, Memc[fname])
+		if (n <= 1)
+		    break
+		n = n - 1
+		Memc[fname+n] = EOS		# clobber the \n
+
+		# Check if file matches the given pattern.
+		if (is_pattern) {
+		    if (n < patlen)
+			next
+		    ep = fname + n - 1
+		    for (ip=patp+patlen-1;  ip > patp+1;  ip=ip-1) {
+			if (Memc[ep] != Memc[ip])
+			    break
+			ep = ep - 1
+		    }
+		    if (Memc[ip] != '.' || Memc[ep] != '.')
+			next
+		}
+
+		call dir_putstr (fp, Memc[fname], n)
+	    }
+
+	    call close (fd)
+	    if (sort_list && FL_NFILES(fp) > 1)
+		call strsrt (FL_OFFSET(fp,1), Memc[FL_SBUFP(fp)], FL_NFILES(fp))
+
+	} else {
+	    # Expand the given filename template and read the file list into
+	    # memory.  This also handles the cases of a single filename or
+	    # a simple list of filenames.
+template_
+	    fd = fntopnb (Memc[files], btoi(sort_list))
+
+	    n  = fntgfnb (fd, Memc[fname], SZ_PATHNAME)
+	    while (n != EOF) {
+		ip = fname
+		if (!is_template && is_pattern)
+		    ip = fname + len_dir;
+		if (is_pattern)
+		    call dir_putstr (fp, Memc[ip], n)
+		else if (access (Memc[fname],0,0) == YES)
+		    call dir_putstr (fp, Memc[ip], n)
+		n = fntgfnb (fd, Memc[fname], SZ_PATHNAME)
+	    }
+
+	    call fntclsb (fd)
+	}
+
+	# All done if no files were found.
+	if (FL_NFILES(fp) == 0) {
+	    call printf ("no files found\n")
+	    goto done_
+	}
+
+	# Format and output the file list in either long format (verbose)
+	# or in multicolumn format.  A one column file list is treated as
+	# a special case of the multifile format.
+
+	if (long_format || ncols == 1) {
+	    for (i=1;  i <= FL_NFILES(fp);  i=i+1) {
+		# Get filename.
+		if (is_dir && strne (Memc[dirname], "./")) {
+		    call sprintf (Memc[fname], SZ_PATHNAME, "%s%s")
+			call pargstr (Memc[dirname])
+			call pargstr (FL_FNAME(fp,i))
+		} else
+		    call strcpy (FL_FNAME(fp,i), Memc[fname], SZ_PATHNAME)
+
+		if (long_format) {
+		    # Long format output, one line per file.
+		    call dir_pfiledata (Memc[fname], STDOUT)
+		} else {
+		    call printf ("%s\n")
+			call pargstr (Memc[fname])
+		}
+	    }
+
+	} else {
+	    # Print nice multicolumn table.
+	    call strtbl (STDOUT, Memc[FL_SBUFP(fp)], FL_OFFSET(fp,1),
+		FL_NFILES(fp), 1, envgeti ("ttyncols"), maxch, ncols)
+	}
+done_
+	call mfree (FL_OFFBP(fp), TY_INT)
+	call mfree (FL_SBUFP(fp), TY_CHAR)
+	call mfree (fp, TY_STRUCT)
+	call sfree (sp)
+end
+
+
+# DIR_QUICKLIST -- List the directory directly to the standard output, without
+# first reading and sorting the entire directory.  This is the best type of
+# listing algorithm for very large directories.
+
+procedure dir_quicklist (out, fd, pat, have_pattern, a_ncols)
+
+int	out			# output file
+int	fd			# fd of directory being listed
+char	pat[ARB]		# selection pattern, if any
+bool	have_pattern		# do we have a pattern?
+int	a_ncols			# number of columns out
+
+bool	flushlines, match_extension
+pointer	sp, fname, obuf, patbuf, op, ep
+int	colwidth, patlen, ip, col, maxch, junk, lastch
+int	scrwidth, ncols, nfiles, nchars, nblanks, nlines, nflush
+int	fstati(), getline(), envgeti(), patmake(), patmatch()
+int	strncmp(), stridxs(), strlen()
+errchk	getline, ungetline, putline
+
+begin
+	call smark (sp)
+	call salloc (obuf, SZ_LINE, TY_CHAR)
+	call salloc (fname, SZ_PATHNAME, TY_CHAR)
+	call salloc (patbuf, SZ_LINE, TY_CHAR)
+
+	# Initialization.
+	flushlines = (out == STDOUT && fstati (out, F_REDIR) == NO)
+	iferr (scrwidth = envgeti ("ttyncols"))
+	    scrwidth = DEF_SCREENWIDTH
+
+	ncols = a_ncols
+	if (ncols <= 0)
+	    ncols = DEF_NCOLS
+	maxch = scrwidth / ncols
+
+	nfiles = 0
+	nchars = 0
+	nlines = 0
+	nflush = 1
+
+	# Set up the pattern matching code.  We recognize selecting all files
+	# with a particular extension as a special case, since this case is
+	# very common and can be done much more efficiently if we don't use
+	# the general pattern matching code.
+
+	if (have_pattern) {
+	    match_extension = (strncmp (pat, "*.", 2) == 0 &&
+		stridxs (PATCHARS, pat[3]) <= 0)
+	    if (match_extension)
+		patlen = strlen (pat)
+	    else {
+		# Convert file matching pattern into general pattern string.
+		Memc[fname] = '^'
+		op = fname + 1
+		lastch = 0
+		for (ip=1;  pat[ip] != EOS;  ip=ip+1) {
+		    if (pat[ip] == '*' && lastch != '?' && lastch != ']') {
+			Memc[op] = '?'
+			op = op + 1
+		    }
+		    lastch = pat[ip]
+		    Memc[op] = lastch
+		    op = op + 1
+		}
+		Memc[op] = '$'
+		op = op + 1
+		Memc[op] = EOS
+			
+		# Compile the pattern.
+		junk = patmake (Memc[fname], Memc[patbuf], SZ_LINE)
+	    }
+	}
+
+	# List the directory.
+
+	while (nchars != EOF) {
+	    # Compose the next line of the directory listing.
+	    op = obuf
+	    for (col=1;  col <= ncols;  ) {
+		nchars = getline (fd, Memc[fname])
+		if (nchars <= 1)
+		    break
+
+		# Cancel the newline.
+		nchars = nchars - 1
+		Memc[fname+nchars] = EOS
+
+		# Check if file matches the given pattern.
+		if (have_pattern)
+		    if (match_extension) {
+			if (nchars < patlen)
+			    next
+			ep = fname + nchars - 1
+			for (ip=patlen;  ip > 2;  ip=ip-1) {
+			    if (Memc[ep] != pat[ip])
+				break
+			    ep = ep - 1
+			}
+			if (pat[ip] != '.' || Memc[ep] != '.')
+			    next
+		    } else if (patmatch (Memc[fname], Memc[patbuf]) <= 0)
+			next
+
+		if (op-obuf + nchars > scrwidth && col > 1) {
+		    # Filename too long to fit in remainder of line.
+		    call ungetline (fd, Memc[fname])
+		    break
+
+		} else {
+		    # Append filename to output line.
+		    nfiles = nfiles + 1
+		    nchars = min (nchars, scrwidth - (op - obuf))
+
+		    # Copy the filename.
+		    do ip = 1, nchars {
+			Memc[op] = Memc[fname+ip-1]
+			op = op + 1
+		    }
+
+		    # Advance to the next column.
+		    colwidth = maxch
+		    while (col < ncols) {
+			nblanks = colwidth - nchars
+			if (nblanks < 1) {
+			    col = col + 1		# grab another column
+			    colwidth = colwidth + maxch
+			} else {
+			    do ip = 1, nblanks {
+				Memc[op] = ' '
+				op = op + 1
+			    }
+			    break
+			}
+		    }
+		}
+
+		col = col + 1
+	    }
+
+	    # Output one line of the directory listing.
+	    if (op > obuf) {
+		Memc[op] = '\n'
+		op = op + 1
+		Memc[op] = EOS
+
+		call putline (out, Memc[obuf])
+
+		nlines = nlines + 1
+		if (flushlines && nlines >= nflush) {
+		    call flush (out)
+		    nflush = min (NLINES_FLUSH, nflush + 1)
+		    nlines = 0
+		}
+	    }
+	}
+
+	# Was the directory empty?
+	if (nfiles == 0)
+	    call printf ("no files found\n")
+
+	call sfree (sp)
+end
+
+
+# DIR_PUTSTR -- Put a string (filename) into the file list.  If the buffer
+# fills up allocate a larger buffer.
+
+procedure dir_putstr (fp, fname, nchars)
+
+pointer	fp			# file list pointer
+char	fname[ARB]		# file name to be put in list
+int	nchars			# nchars in file name
+
+int	op
+int	fileno
+
+begin
+	# Increment file count.
+	fileno = FL_NFILES(fp) + 1
+	FL_NFILES(fp) = fileno
+
+	# Increase size of file list if it overflows.
+	if (fileno >= FL_MAXFILES(fp)) {
+	    FL_MAXFILES(fp) = FL_MAXFILES(fp) * 2
+	    call realloc (FL_OFFBP(fp), FL_MAXFILES(fp), TY_INT)
+	}
+
+	op = FL_NEXTOFF(fp)
+	FL_OFFSET(fp,fileno) = op
+
+	# Increase size of string buffer if it overflows.
+	if (op + nchars + 1 >= FL_SZSBUF(fp)) {
+	    FL_SZSBUF(fp) = FL_SZSBUF(fp) * 2
+	    call realloc (FL_SBUFP(fp), FL_SZSBUF(fp), TY_CHAR)
+	}
+
+	call strcpy (fname, FL_FNAME(fp,fileno), nchars)
+	FL_NEXTOFF(fp) = op + nchars + 1
+end
+
+
+# DIR_PFILEDATA -- Print file info in a UNIX like long directory format, on
+# the output file OUT.  Format, e.g.,: "-t-rwr-r- tody 5117 May 15 19:29 fio.x"
+
+procedure dir_pfiledata (fname, out)
+
+char	fname[ARB]			# file name
+int	out				# output file
+
+pointer	sp, date
+long	fi[LEN_FINFO]
+int	protected, ftypes[4], i
+int	finfo(), access(), bitupk(), protect()
+data	ftypes /'-', 'd', 'x', 's'/
+
+begin
+	call smark (sp)
+	call salloc (date, SZ_DATE, TY_CHAR)
+
+	# Get file directory information.
+	if (finfo (fname, fi) != OK) {
+	    call eprintf ("Cannot get info on file `%s'\n")
+		call pargstr (fname)
+	    call sfree (sp)
+	    return
+	}
+
+	# Query file protection.
+	protected = protect (fname, QUERY_PROTECTION)
+
+	# Output file mode bit flags.
+	call dir_putci (out, ftypes[FI_TYPE(fi)], '-', YES)
+	call dir_putci (out, 't', 'b', access (fname,0,TEXT_FILE))
+	call dir_putci (out, 'p', '-', protected)
+
+	# Output user,group,world protections bit flags.
+	do i = 1, 5, 2 {
+	    call dir_putci (out, 'r', '-', bitupk (int(FI_PERM(fi)),i,1))
+	    call dir_putci (out, 'w', '-', bitupk (int(FI_PERM(fi)),i+1,1))
+	}
+
+	# Output file owner, size, date, and name.
+	call cnvdate (FI_MTIME(fi), Memc[date], SZ_DATE)
+	call fprintf (out, " %-8.8s%8d %12.12s %-32.32s\n")
+	    call pargstr (FI_OWNER(fi))
+	    call pargl (FI_SIZE(fi))
+	    call pargstr (Memc[date])
+	    call pargstr (fname)
+
+	# This is slow so flush the output after each file to give the user
+	# some immediate gratification.
+
+	call flush (out)
+	call sfree (sp)
+end
+
+
+# DIR_PUTCI -- Handy procedure used to generate --rw-r-r- type strings.
+
+procedure dir_putci (fd, ch_t, ch_f, condition)
+
+int	fd			# output file
+int	ch_t			# character to output if condition is true
+int	ch_f			# character to output if condition is false
+int	condition
+
+begin
+	if (condition == NO)
+	    call putci (fd, ch_f)
+	else
+	    call putci (fd, ch_t)
+end
diff --git a/pkg/system/doc/Sys.hlp b/pkg/system/doc/Sys.hlp
new file mode 100644
index 00000000..89d66d7f
--- /dev/null
+++ b/pkg/system/doc/Sys.hlp
@@ -0,0 +1,349 @@
+.help utilities Oct83 "IRAF System Utilities"
+.sh
+Description of the Standard Utility Routines
+
+    Most of the system utilities are set up to read from either the
+standard input or a list of files.  If not reading from the standard
+input, and the input file template is not given on the command line,
+a prompt will be issued for the input file list.  If the output is a
+single file or directory, and output is not redirected, a prompt will
+be issued for the output file.
+
+.ks
+.ls 4 ALLOCATE device
+
+Allocate a device (i.e., a tape drive).  Some devices must be allocated
+before they can be accessed for i/o.  Deallocation occurs automatically
+upon logoff, or upon execution of DEALLOCATE.  The following devices may
+be allocated:
+
+.nf
+Devices:
+	mta	Magtape unit A.
+	mtb	Magtape unit B.
+	ima	Image display unit A.
+	imb	Image display unit B.
+.fi
+.le
+.ke
+
+.ks
+.ls BEEP
+Beep the terminal bell.  Used as in "> munch;beep" to signal when a
+lengthy command has finished.  Otherwise, an IRAF program will only
+ring the terminal bell when an error has occurred.
+.le
+.ke
+
+.ks
+.ls CLEAR
+
+Clear the terminal screen.
+.le
+.ke
+
+.ks
+.ls CONCATENATE files [,output_file]
+
+Concatenate the named input files to the standard output or to the named
+output file if given.  All input files should be of the same type (text or
+binary).  The default type of the output file is the same as that of the
+first input file; this may be overridden by setting the hidden parameter
+"out_type".  The standard input is assumed to be a text file unless overridden
+by setting out_type.  If no output file is given on the command line, the
+standard output is assumed.
+
+.nf
+Hidden Parameters:
+    out_type	string = "in_type"	Output file type (text|binary)
+    append	bool = no		Append to output file
+.fi
+.le
+.ke
+
+.ks
+.ls COPY file|files, output_file|directory
+
+Copy the input file to the output file, or copy each file in the input list
+to the directory given as the second argument.
+.le
+.ke
+
+.ks
+.ls COUNT files
+
+Count the number of lines, words, and characters in each file
+in the input list.  Print the totals for each file and for all files.
+.le
+.ke
+
+.ks
+.ls DEALLOCATE device
+
+Deallocate a previously allocated device (e.g., a tape drive).  Devices
+are automatically deallocated upon logoff if not explicitly deallocated
+with the deallocate command.
+.le
+.ke
+
+.ks
+.ls DELETE files
+
+Delete the named files.  A warning message is printed if a file
+cannot be deleted.  It is not considered an error to attempt to
+delete a file which does not exist.  Protected files cannot be
+deleted without first removing their protection.
+
+.nf
+Hidden Parameters:
+    verify	bool = no	Generate query to verify that each
+				  file should indeed be deleted.
+.fi
+.le
+.ke
+
+.ks
+.ls DEVSTATUS device
+
+Print status information (i.e., is the device allocated, to whom, etc.)
+for the named device.
+.le
+.ke
+
+.ks
+.ls DIRECTORY [files] [,op=string]
+
+Tabulate information on all files matching the template given as the
+first argument on the standard output.  If no template is given, the
+contents of the current directory are displayed.  The "long" format
+gives file protection status, type (text, binary, directory, or executable),
+access modes (owner, group, world), owner name, size in bytes, date of
+last modify, and the file name.
+
+.nf
+Hidden Parameters:
+    option	string = "m"	    1	one-column format
+				    a	use time of last access
+				    c	use time of creation
+				    l	long format
+				    m	multicolumn format (default)
+				    s	add size in Kb
+				    t	time sort
+				    r	reverse sort
+
+e.g.: > dir op=l	(show current directory, long form)
+      > dir "*.x",op=lt	(all ".x" files in current directory, long form)
+.fi
+.le
+.ke
+
+.ks
+.ls DISKSPACE
+
+Summarize the amount of disk space available on the local system.
+This command and its output are machine dependent.  This command
+may not be available on all systems.
+.le
+.ke
+
+.ks
+.ls HEAD files
+
+Print the first few lines of each of the named files on the standard
+output.  If more than one file is printed, a brief header is printed to
+identify each file.
+
+.nf
+hidden params:
+    nlines	int = 12	number of lines to print
+.fi
+.le
+.ke
+
+.ks
+.ls LPRINT files
+
+The named text files (or the standard input) are printed on the standard
+line printer device.  If several files are to be printed, each file is
+printed starting on a new page, and file pages are broken and printed
+with numbered headers.  If reading from the standard input, the input
+stream is copied to the printer without pagination, though pages may still
+be broken by embedding formfeed characters in the text.  The map_cc
+option may be used to turn off all control character mapping, if binary
+data is to be sent to the printer.
+
+.nf
+Hidden Parameters:
+    map_cc	bool = yes	Map control characters into a printable
+				  sequence (^L, etc.)
+.fi
+.le
+.ke
+
+.ks
+.ls MATCH pattern, files
+
+Search each text file for the given pattern.  Copy a line to the standard
+output if the pattern can be matched.  Match against the standard input
+if no files are named.
+
+.nf
+Hidden Parameters:
+    stop	bool = no	Stop matched lines and pass unmatched
+				  lines to the output instead.
+.fi
+.le
+.ke
+
+.ks
+.ls NEWS
+
+Page through the system news file, starting with the most recent entries.
+The system news file is a sort of electronic bulletin board, used to
+post announcements and notes of general interest.
+.le
+.ke
+
+.ks
+.ls PAGE files
+
+The named text files (or the standard input) are displayed a page at a
+time on the standard output.  If there is more than one input file,
+a brief header is printed at the beginning of each file.
+
+.nf
+Hidden Parameters:
+    map_cc	bool = yes	Map control characters into a printable
+				  sequence (^L, etc.)
+.fi
+.le
+.ke
+
+.ks
+.ls PROTECT files
+
+Protect the named files from deletion or clobber, accidental or otherwise.
+It is not considered an error to attempt to protect a file which is already
+protected.  Protecting a file prevents only deletion or clobber of the entire
+file: the contents of the file may still be overwritten.
+.le
+.ke
+
+.ks
+.ls RENAME file|files, output_file|directory
+
+Rename a file, or move each file in the input list to the directory
+given as the second argument.
+.le
+.ke
+
+.ks
+.ls SORT files
+
+Sort and/or merge the text files named in the list, or sort the standard
+input if no list.  Sorting is by line or column, alphabetically or
+numerically, in forward or reverse order.
+
+.nf
+Hidden Parameters:
+    numeric_sort	bool = no	Sort numerically instead of
+					  alphabetically, if datum is
+					  a number.
+    reverse_sort	bool = no	Reverse the sense of the sort.
+    column		int = 0		If 0, sort the entire line, else
+					  sort the indicated column.
+    ignore_whitespace	bool = yes	Ignore leading whitespace if
+					  sorting full lines.
+    merge		bool = no	Merge the input files, which are
+					  assumed to already have been
+					  sorted.
+.fi
+.le
+.ke
+
+.ks
+.ls SPY
+
+Get information on who is using the system, what they are up to, how
+much cpu time, etc., they have used, and so on.  This command, and the
+information it gives, are machine dependent.  This command may not be
+available on all systems.
+
+.nf
+Hidden Parameters:
+    verbose	bool = no	Give more detailed information.
+.fi
+.le
+.ke
+
+.ks
+.ls STTY [terminal]
+
+Identify the type of terminal being used, if different than the system
+default.  This is necessary to tell the system the control sequence to
+be used to clear the screen, the dimensions of the screen, etc.  If no
+arguments are given, the current terminal status is given.
+
+.nf
+Hidden Parameters:
+    baud	int = 0		If nonzero, the system takes the
+				  argument as the new baud rate.
+    ncols	int = 0		If nonzero, the argument becomes
+				  the number of columns per screen.
+    nlines	int = 0		If nonzero, the argument becomes
+				  the number of lines per screen.
+.fi
+.le
+.ke
+
+.ks
+.ls TAIL files
+
+Print the last few lines of each of the named text files on the
+standard output.
+
+.nf
+Hidden Parameters:
+    nlines	int = 12	Number of lines to print.
+.fi
+.le
+.ke
+
+.ks
+.ls TEE file
+
+Copy the standard input to the standard output, as well as to the named
+output file.
+
+.nf
+Hidden Parameters:
+    out_type	string = "text"	Type of output file (text|binary).
+    append	bool = no	Append to the output file.
+.fi
+.le
+.ke
+
+.ks
+.ls TIME
+
+Print the current time and date.
+.le
+.ke
+
+.ks
+.ls TYPE files
+
+Copy the named text files to the standard output.  If more than one file is
+to be typed, a brief header is printed at the beginning of each file,
+identifying the file.
+.le
+.ke
+
+.ks
+.ls UNPROTECT files
+
+Remove delete protection from the named files.  It is not considered
+an error to attempt to remove protection from a file which is not protected.
+.le
+.ke
+.endhelp
diff --git a/pkg/system/doc/Sys_intro.hlp b/pkg/system/doc/Sys_intro.hlp
new file mode 100644
index 00000000..03d7e797
--- /dev/null
+++ b/pkg/system/doc/Sys_intro.hlp
@@ -0,0 +1,137 @@
+.help system May83 "General Aspects of the System Package"
+.sh
+Basic IRAF System Utilities
+
+    A number of standard utilities are available as part of the IRAF
+user interface.  These utilities, most of which operate on files, are
+summarized later in this document.  A general discussion of files in
+the IRAF system follows.  The reader is assumed to have some familiarity
+with the IRAF command language (CL).
+
+
+.sh
+Virtual File Names
+
+    File names may be specified in a machine independent fashion, or with
+OS dependent pathnames.  A Virtual File Name (VFN) has the following form:
+
+
+.nf
+	ldir$root.extn
+
+where
+
+	ldir		logical directory or device name
+	root		root or base file name
+	extn		extension denoting the type of file
+.fi
+
+
+The LDIR and EXTN fields are optional.  The ROOT and EXTN fields may contain
+up to 20 characters selected from the set [a-zA-Z0-9_.+-#].  The EXTN field
+may not exceed three characters.  The EXTN field is separated from the ROOT
+by the character "." (dot).  If the ROOT field contains one or more occurrences
+of the dot character, the final dot delimited field is understood to be
+the extension.
+
+Logical directories are defined in the CL, using the SET command to associate
+an OS dependent pathname with a keyword in the environment table.  Logical
+directory prefixes are recursively expanded.
+
+A number of standard logical directories and devices are defined by the CL
+at startup time.  Additional logical directories are defined by applications
+packages upon entry, and become undefined when dictionary space is reclaimed
+on exit from the package.  The user may override environment definitions at
+will, by issuing SET commands to redefine environment variables.
+
+
+.sh
+File List Templates
+
+    Whenever it makes sense, the IRAF utilities are set up to process a list
+of files, rather than a single file.  In the descriptions of the standard
+utilities in the next section, "files" is a template or pattern specifying
+a list of files to be processed in some way.  The template is a string type
+parameter to the CL, which is expanded by a procedure (CLGFIL) in the
+compiled systems or applications task into an EOF terminated list of files.
+
+A template may consist of one or more file names, directory names,
+list file names, or patterns.  List elements are delimited by commas.
+A list file is denoted by prepending the character "@" to the pathname
+of the list file.  A pattern may be applied to the contents of either
+a list file or a directory.  The pattern is separated from the list file
+or directory name by the character "$".  A pattern all by itself is
+applied to the current directory.  Either logical or OS dependent
+pathnames may be used.
+
+A list element may refer only to a single directory.  Thus, one cannot
+specify a pattern such as "*/*.x", as is possible in UNIX.
+
+Some examples of templates follow:
+
+.nf
+	file
+	"file1, file2"
+	"*.x"
+	"[A-Z]*, *.com, ../$, ../fio/$*.x, lib$*.h, file, @list$*.vs"
+	"sia1:[iraf.lib]$*.h, uparm$cl_*.par"
+.fi
+
+The magic filenames "STDIN", "STDOUT", and "STDERR" have a special
+meaning.  Passing one of these special filenames to a task causes
+the named stream to be "reopened", transparently to the applications
+program.  Thus, for example, one can reference "STDIN" in a template,
+and the applications program will read from the standard input when
+it opens the corresponding file in the template.
+
+For example,
+
+	cl> concat "file1, @STDIN, STDIN, file2", > ofile
+
+concatenates the file "file1", the contents of the files named in the
+standard input (until the first EOF), the contents of the standard input
+itself (until the second EOF), and the contents of the file "file2",
+placing the output in the file "ofile".
+
+
+.sh
+File Protection and Clobber
+
+    The IRAF file interface (FIO) provides file protection and synchronization
+facilities, under control of user definable environment variables.
+
+File "clobber" refers to the overwriting of an existing file when a new
+file of the same name is created as an output file.  If "clobber" is
+defined as "no" in the environment, an IRAF task will abort if a new file
+would clobber an existing file.  If file clobber is enabled, FIO will try
+to overwrite the old file, and will abort if it cannot do so.
+
+	cl> set clobber = "yes"
+
+More explicit file protection is provided by the CL commands PROTECT and
+UNPROTECT.  A protected file cannot be deleted, accidentally or otherwise,
+until the protection has been removed with the UNPROTECT command or system
+call.
+
+A final form of file protection is provided to prevent a file from being
+clobbered which is already open by a task.  Thus, "COPY file,file" will
+abort when FIO discovers that "file" is already opened for reading.  This
+holds even if file clobber is enabled.
+
+
+.sh
+File Synchronization
+
+    File synchronization is useful when a process requires exclusive access
+to a file which is already open for exclusive access by another process.  
+When this situation occurs, the second task may abort with a message stating
+that it cannot access the file, or the task may wait for the file to become
+available.  File waiting is especially important for batch processes.
+
+	cl> set file_wait = "yes"
+
+The file wait option is controlled by the "file_wait" variable in the
+environment.  If file waiting is enabled, and a process finds that it
+has to wait to access a file, a warning message will be printed on the
+standard error output before the affected process goes to sleep.
+.endhelp
diff --git a/pkg/system/doc/allocate.hlp b/pkg/system/doc/allocate.hlp
new file mode 100644
index 00000000..66d3353f
--- /dev/null
+++ b/pkg/system/doc/allocate.hlp
@@ -0,0 +1,52 @@
+.help allocate Jan86 system
+.ih
+NAME
+allocate -- allocate a device
+.ih
+USAGE
+allocate device
+.ih
+PARAMETERS
+.ls device
+The device to be allocated.
+.le
+.ih
+DESCRIPTION
+\fBAllocate\fR allocates a device for exclusive access by one user, and
+readies the device for i/o by some other program.  A list of the devices
+available on the local system is maintained in the file \fBdev$tapecap\fR
+which needs to be configured by the site manager before it can be used.
+The status of given device may be obtained by calling \fIdevstatus\fR.
+.ih
+EXAMPLES
+
+1. Print a list of the allocatable devices.  The logical device names are
+given at the left in the output text; ignore the information to the right.
+\fBNote\fR: The dev$devices file should be configured by the site manager
+when new tape devices are installed.  Beginning with V2.9 it is used for
+informational purposes only.
+.ks
+.nf
+	cl> type dev$devices
+	mta ...
+	mtb ...
+	mtc ...
+	iis ...
+.fi
+.ke
+
+
+2. Allocate a tape drive after checking its status.
+
+.ks
+.nf
+	cl> devstatus mtb
+	device mtb is not currently allocated
+	cl>
+	cl> allocate mtb
+.fi
+.ke
+.ih
+SEE ALSO
+deallocate, devstatus
+.endhelp
diff --git a/pkg/system/doc/bench.hlp b/pkg/system/doc/bench.hlp
new file mode 100644
index 00000000..1e8b373e
--- /dev/null
+++ b/pkg/system/doc/bench.hlp
@@ -0,0 +1,56 @@
+.help bench Dec2010 system
+.ih
+NAME
+bench -- simple benchmark tool
+.ih
+USAGE
+bench
+.ih
+DESCRIPTION
+The BENCH task provides a simple benchmark meant to simulate CCD processing
+of three images.  The tasks used exercise both disk- and CPU-intensive
+applications and offer a crude comparison of performance between machines, 
+but it is not intended to be a comprehensive benchmark.  There are no
+parameters.
+.ih
+EXAMPLES
+1. Run the benchmark:
+
+.nf
+    ecl> bench
+            
+    Bench started at 09:45:58
+                             
+    =====> Making images...  
+             Making zero...  
+             Making flat...  
+             Making 3 objs...
+                             
+    =====>   Normalizing flat...
+    	:
+    
+    =====>   Subtracting zero from 3 images...    
+    	:
+    
+    =====>   Dividing flat into 3 images...       
+    	:
+    
+    =====>   Combining 3 images...
+    	:
+    
+    =====>   Median filtering image...
+    	:
+    
+    =====>   Deleting all images...
+    
+    Bench started at 09:45:58
+    Bench  ended at  09:46:56
+    
+    Total execution time =    58.0 seconds
+     Total time  Make 5 imgs  Proc 3 imgs  Combine 3 imgs  Median 1 img
+         58.0         25.0        10.0          4.0           19.0
+.fi
+
+.ih
+SEE ALSO
+.endhelp
diff --git a/pkg/system/doc/chkupdate.hlp b/pkg/system/doc/chkupdate.hlp
new file mode 100644
index 00000000..c4209753
--- /dev/null
+++ b/pkg/system/doc/chkupdate.hlp
@@ -0,0 +1,72 @@
+.help chkupdate Oct13 system
+.ih
+NAME
+chkupdate - Check for an available IRAF update
+.ih
+USAGE
+chkupdate
+.ih
+PARAMETERS
+.ls interval = 0
+Number of days between updates checks.  A value less than zero will disable
+the checks entirely, a value of zero will cause a check to be made with 
+each login.
+.le
+.ls ref_file = "iraf$.release_date"
+.le
+.ls release = ")_.release"
+Current IRAF release version.  This value is inherited from the CL 'release'
+parameter by default.
+.le
+.ls baseurl = "http://iraf.noao.edu/ftp"
+Base URL to the IRAF release timestamp directory.
+.le
+.ls verbose = yes
+Print verbose output?
+.le
+.ih
+DESCRIPTION
+This task compares the currently installed IRAF version with what is
+available from the NOAO servers and will indicate whether an update is
+available.  The task is executed from the login.cl each time you login 
+to the CL, however the comparison is only done if 1) the uparm$update 
+file used to indicate the time of last check does not exist, 2) the 
+\fIinterval\fR parameter is zero to indicate the check should be done
+with each login, or 3) more than \fIinterval\fR days have passed since
+the last time the servers were contacted.  An \fIinterval\fR value less
+that zero may be used to disable the version updates entirely.
+
+The update check is done by comparing the file timestamp of the file
+named in the \fIref_file\fR parameter with a distribution timestamp file
+on the NOAO servers.  The URL to this file is constructed from the 
+\fIbaseurl\fR and \fIrelease\fR parameters in addition to the \fIarch\fR
+environment value, yielding a unique URL for each version and each platform.
+If the contents of the file contain a timestamp more recent than the 
+timestamp of the \fIref_file\fR value, a message is printed to indicate
+an update is available, otherwise a message is printed indicating the
+installed system is current.
+.ih
+EXAMPLES
+
+1. Check whether an IRAF update is available, regardless of when we last
+checked.
+
+.nf
+	cl> chkupdate interval=0
+.fi
+
+2. Check for an IRAF update once a month.
+
+.nf
+	cl> chkupdate.interval = 30
+.fi
+
+.ih
+NOTES
+This task is called automatically from the login.cl file at startup.
+
+Modifying the timestamp information of the \fIref_file\fR parameter, e.g.
+by moving the IRAF tree, may invalidate the output.
+.ih
+SEE ALSO
+.endhelp
diff --git a/pkg/system/doc/concatenate.hlp b/pkg/system/doc/concatenate.hlp
new file mode 100644
index 00000000..e3cd4fdc
--- /dev/null
+++ b/pkg/system/doc/concatenate.hlp
@@ -0,0 +1,82 @@
+.help concatenate Nov84 system
+.ih
+NAME
+concatenate -- connect files together into one big file
+.ih
+USAGE
+concatenate files [output_file]
+.ih
+PARAMETERS
+.ls files
+The list of input files.  The standard input, STDIN, may be specified to
+interactively enter a few lines of text rather than read from a disk file.
+All input files should be of the same type (binary or text).
+.le
+.ls output_file
+The name of the output file.  If no file is explicitly specified the
+standard output (STDOUT) is used.
+.le
+.ls out_type = in_type
+The output file type is forced if this parameter is defined as "binary"
+or "text".  If "out_type" does not begin with a "b" (or "B"), or a
+"t" ("T"), then the output type is either "text", if the output file is
+the standard output, or is determined from the type of the first input file.
+.le
+.ls append = no
+If set to "yes", "files" are appended to "output_file".
+.le
+.ih
+DESCRIPTION
+Each file in the input file list is appended to the output file.
+If "output_file" is not the standard output, and if output redirection (">")
+was not specified on the command line, the resulting stream of data is placed
+in a file.  The input can be STDIN, which makes for an easy way to enter a
+few lines of text into a file (but \fItype\fR is usually more convenient).
+If entering data via the standard input, type the end of file character,
+e.g., <ctrl/z>, to terminate the input sequence.
+.ih
+EXAMPLES
+
+1. Write out file1, followed by file2, to the terminal screen.  Note that
+there must be no space after the comma.
+
+.nf
+	cl> concatenate file1,file2
+.fi
+
+2. Write out files file1 and file2 into the new file "outfile".
+
+.nf
+	cl> concatenate file1,file2 outfile
+.fi
+
+3. Copy what you type (up to the end-of-file character) into the file junk.
+
+.nf
+	cl> concatenate STDIN junk
+.fi
+
+4. Write out the contents of each of the files whose names are given in "list",
+one per line, and append this data to "junk".
+
+.nf
+	cl> concatenate @list junk append+
+.fi
+
+5. Concatenation is also possible using \fItype\fR, e.g., the following
+command will append the contents of "file" to the file "outfile", which will
+be created if it does not already exist.
+
+	cl> type file >> outfile
+
+The redirect-append operator ">>" may be used to append the output of any
+task to a file.
+
+.ih
+NOTES
+All input files should be of the same type, either all "text" or all
+"binary".
+.ih
+SEE ALSO
+copy, type
+.endhelp
diff --git a/pkg/system/doc/copy.hlp b/pkg/system/doc/copy.hlp
new file mode 100644
index 00000000..9c58566d
--- /dev/null
+++ b/pkg/system/doc/copy.hlp
@@ -0,0 +1,47 @@
+.help copy Nov84 system
+.ih
+NAME
+copy -- copy a file, or a set of files to a directory
+.ih
+USAGE
+copy input output
+.ih
+PARAMETERS
+.ls input
+The input file or list of files to be copied.
+.le
+.ls output
+The (new) output file when copying one file to another, or the destination
+directory when copying a set of files.
+.le
+.ls verbose = no
+If set to "yes", a line of the type " from -> to " is printed on the
+terminal for each file copied to a directory.  This parameter is not
+used when copying one file to another.
+.le
+.ih
+DESCRIPTION
+Copy makes a copy of a single file, or it copies a set of files to a different
+directory.
+.ih
+EXAMPLES
+
+1. Copy all files in the current directory with extension ".x" to the
+directory "home$src".  As each copy is made, the user is informed.
+
+	cl> copy *.x home$src ver+
+
+2. Make a copy "fred.BAK" of the file "fred".
+
+	cl> copy fred fred.BAK
+
+3. Copy the "graphcap" file from the remote node "lyra" to the current node,
+without changing the name of the file.  Note that "." is a synonym for the
+current directory.
+
+	cl> copy lyra!dev$graphcap .
+
+.ih
+SEE ALSO
+concatenate, movefiles
+.endhelp
diff --git a/pkg/system/doc/count.hlp b/pkg/system/doc/count.hlp
new file mode 100644
index 00000000..09601019
--- /dev/null
+++ b/pkg/system/doc/count.hlp
@@ -0,0 +1,44 @@
+.help count Nov84 system
+.ih
+NAME
+count -- determine number of lines, words and characters in a file
+.ih
+USAGE
+count files
+.ih
+PARAMETERS
+.ls files
+A template specifying the files to be examined.
+.le
+.ih
+DESCRIPTION
+For each file, count determines the number of lines, words, and
+characters in the file.  A word is defined as a sequence of characters
+delimited by one or more blanks or tabs, or by the end of a line.
+If \fIcount\fR is run on more than one file, each output line is identified
+by the file name, and a final output line gives the total number
+of lines, words, and characters in all files.
+.ih
+EXAMPLES
+
+1. Count the number of lines, words and characters in all files in the
+current directory with the extensions ".x" and ".h".
+
+	cl> count *.[xh]
+
+2. Count the number of .x files in the current directory.
+
+	cl> dir *.x op=1 | count
+
+3. Count the number of \fIset\fR environment definitions.
+
+	cl> set | count
+
+4. Count the number of references to the READ function in all .x files in
+the current directory.
+
+	cl> match "read#(" *.x | count
+.ih
+SEE ALSO
+directory
+.endhelp
diff --git a/pkg/system/doc/deallocate.hlp b/pkg/system/doc/deallocate.hlp
new file mode 100644
index 00000000..803c9d6a
--- /dev/null
+++ b/pkg/system/doc/deallocate.hlp
@@ -0,0 +1,34 @@
+.help deallocate Jan86 system
+.ih
+NAME
+deallocate -- deallocate a device
+.ih
+USAGE
+deallocate device
+.ih
+PARAMETERS
+.ls device
+The device to be deallocated.
+.le
+.ls rewind = yes
+Rewind the device before deallocating?
+Ignored for devices other than magtape.
+.le
+.ih
+DESCRIPTION
+Deallocate a previously allocated device.  The CL will print an error
+message if one attempts to logout while devices are still allocated,
+but if \fIlogout\fR is typed several times you will be allowed to logout
+with the devices still allocated.  The CL does not automatically
+deallocate devices upon logout.
+.ih
+EXAMPLES
+
+1. Deallocate logical magtape drive B.
+
+	cl> dealloc mtb
+
+.ih
+SEE ALSO
+allocate, devstatus, file dev$devices, dev$tapecap
+.endhelp
diff --git a/pkg/system/doc/delete.hlp b/pkg/system/doc/delete.hlp
new file mode 100644
index 00000000..78ac7df3
--- /dev/null
+++ b/pkg/system/doc/delete.hlp
@@ -0,0 +1,57 @@
+.help delete Nov84 system
+.ih
+NAME
+delete -- delete a file or files
+.ih
+USAGE
+delete files
+.ih
+PARAMETERS
+.ls files
+The list of files to be deleted.
+.le
+.ls verify = no
+Check with the user before deleting a file.  If verify is enabled the file
+name is printed and the user is queried before the default action is taken.
+.le
+.ls default_action = yes
+This is the default action to take when operating in "verify" mode.
+For example, if the default action is "yes", one need only type RETURN in
+response to the verify prompt to delete the file.
+.le
+.ls allversions = yes
+Delete all versions of a file.  This parameter has no effect on systems like
+UNIX which do not support multiple file versions.
+.le
+.ls subfiles = yes
+Delete subfiles.  Not currently used.
+.le
+.ih
+DESCRIPTION
+\fIDelete\fR destroys a file or files and returns the space they occupied to
+the host system, to be reused for other files.  Once a file has been deleted,
+it is gone forever (unless a copy exists somewhere).  Enabling \fIverify\fR
+gives one the opportunity to say yes or no before each file is deleted; this
+is particularly useful when \fIfiles\fR is a template.  Note that
+\fIprotect\fR can be used to protect files from deletion, accidental or
+otherwise.  Imagefiles are automatically protected by the system to remind
+the user to use \fIimdelete\fR to delete images (this is necessary because
+an image is stored in more than one physical file).
+.ih
+EXAMPLES
+
+1. Delete all files with extension ".x", verifying each file deletion before
+it is performed.
+
+	cl> delete *.x ver+
+
+2. List all files in the current directory, deleting only those files for
+which the user responds to the verify prompt with "yes" or "y".  Note that
+"delete *" is a very dangerous operation.
+
+	cl> delete * ver+ def=no
+
+.ih
+SEE ALSO
+protect, unprotect, imdelete
+.endhelp
diff --git a/pkg/system/doc/devstatus.hlp b/pkg/system/doc/devstatus.hlp
new file mode 100644
index 00000000..bccd7ffe
--- /dev/null
+++ b/pkg/system/doc/devstatus.hlp
@@ -0,0 +1,57 @@
+.help devstatus Jan86 system
+.ih
+NAME
+devstatus -- print status information for a device
+.ih
+USAGE
+devstatus device
+.ih
+PARAMETERS
+.ls device
+The device for which status information is requested.
+.le
+.ls verbose = no
+Print additional system dependent device information (may not be implemented
+on all systems).
+.le
+.ih
+DESCRIPTION
+\fBDevstatus\fR tells whether the named device has been allocated.
+In the case of a magtape drive allocated to the current user, additional
+information is printed noting the tape position and the type of operation
+last performed.  If the device is not currently allocated, i.e., available
+for allocation, or if the device has already been allocated by another user,
+one of the following messages is printed:
+
+.nf
+	device is not currently allocated
+	device is allocated to XXXX
+.fi
+
+A list of the allocatable devices, including the host system names for the
+devices, can be obtained by paging the file \fBdev$devices\fR providing
+that file has been properly configure by the Site Manager at installation
+time.  The dev$tapecap file is used to define the tape devices available
+on the system.
+.ih
+EXAMPLES
+
+1. Get status information for the logical tape drive "mta", which we have
+just allocated.  Note that the tape position is printed only if we are the
+owner of the drive.
+
+.nf
+	cl> dev mtb
+	# Magtape unit `mta' allocated to `smith' Fri 12:04:16 07-Jan-86
+	file = 1
+	record = 1
+	unit just allocated: no i/o has yet occurred
+.fi
+
+.ih
+BUGS
+Information can only be requested for a single device at a time.
+.ih
+SEE ALSO
+allocate, deallocate
+.endhelp
diff --git a/pkg/system/doc/directory.hlp b/pkg/system/doc/directory.hlp
new file mode 100644
index 00000000..8c4382e9
--- /dev/null
+++ b/pkg/system/doc/directory.hlp
@@ -0,0 +1,148 @@
+.help directory Mar87 system
+.ih
+NAME
+directory -- list the contents of a file directory
+.ih
+USAGE
+directory [files]
+.ih
+PARAMETERS
+.ls files
+A file template specifying the files to be listed, or the name of the directory
+whose contents are to be listed.  If omitted entirely, the contents of the
+current directory are listed.
+.le
+.ls long = no
+Long format listing.  The long format listing lists each file on a separate
+line, noting the file permissions, file type, file size, modify date, owner,
+etc. of each file.
+.le
+.ls ncols = 0
+If nonzero, the number of columns of output in multicolumn format.
+.le
+.ls maxch = 18
+The maximum number of characters to be displayed in each filename.
+Truncation may be desirable when listing a directory containing one or two
+files with very long filenames.
+.le
+.ls sort = yes
+Sort the file list alphabetically.  If sorting is disabled the directory
+program lists the files in the order in which they are read from the
+directory, which may or may not be sorted.  The directory listing is produced
+line by line as files are read from the directory, rather than accumulating
+the entire file list in memory before composing the table, hence this is the
+fastest method of listing a directory, particularly if the directory is very
+large.
+.le
+.ls all = no
+List all files, including the hidden ("." prefixed) files, and files with
+reserved filename extensions used internally by the VOS.
+.le
+.ih
+DESCRIPTION
+The \fBdirectory\fR task lists or prints information describing some subset
+of the files in a directory or directories.  If no name template is given,
+"." is assumed, i.e., all files in the current directory are listed.
+
+The long format listing gives a file type string, followed by
+the name of the owner of the file, the file size, date and time at which
+the file was last modified, and lastly the file name.
+The file type string has fields noting if the file is a directory file (d),
+an executable file (x), a text or binary file (t or b), a protected file (p),
+and summarizing the file permissions (read or write, r or b) for the owner,
+the group, and the rest of the world.  A minus sign indicates that the file
+does not have that particular attribute.
+
+All file names are printed in the IRAF virtual filename syntax, which is the
+same on all host machines.  IRAF filenames may be up to 32 characters in
+length, may contain any combination of alphanumeric characters, underscore,
+or period, and are case sensitive.  Some of the common filename extensions
+are listed below; these are mapped to and from the host filename extensions
+when a file is accessed, a directory is listed, or a filename template is
+expanded.
+
+.nf
+	.a	object library
+	.c	C source file
+	.cl	CL source file
+	.e	executable (runnable) file
+	.f	Fortran source file
+	.gX	generic source file (X=[cx])
+	.h	global header file
+	.hlp	help file
+	.o	object file
+	.par	CL parameter file
+	.s	assembler source file
+	.x	SPP source file
+.fi
+
+When listing large directories, the time required to accumulate and sort the
+entire directory in memory before producing the output listing may become
+significant (i.e., more than a few seconds).  If this happens, try setting
+the \fIsort\fR option to \fIno\fR, and the directory listing should appear
+immediately.
+.ih
+EXAMPLES
+
+1. List all the files in the current directory in tabular format.
+
+	cl> dir
+
+2. Print detailed information on all files in the current directory.
+
+.nf
+        cl> dir l+
+        -t-rwr-r- iraf         269 Oct 16  1983 README
+        dt-rwrwr- iraf        1024 Feb  7 12:48 doc
+        -t-rwr-r- iraf          60 Jan 30  1984 files.par
+        -t-rwr-r- iraf         420 Jan 30  1984 files.x
+        -b-rwrwr- system    187338 Jan 29 19:27 libpkg.a
+        xb-rwr-r- iraf      363520 Jan 29 19:29 x_system.e
+        -b-rwrwr- system      5037 Jan 19 22:15 x_system.o
+        -t-rwr-r- iraf         633 Jan 19 22:01 x_system.x
+.fi
+
+3. Print a single column listing of all the files with extension ".h"
+in the logical directory "lib$".
+
+.nf
+	cl> dir lib$*.h l+
+	lib$chars.h
+	lib$clio.h
+	lib$clpopn.h
+	    (etc)
+.fi
+
+4. While in the "system" directory, print the contents of the parallel
+directory "dataio".
+
+.nf
+	cl> cd pkg$system
+	cl> dir ../dataio
+.fi
+
+5. Test if the file "alpha" exists in the current directory.  In the example,
+the output given indicates that the file was not found.
+
+.nf
+	cl> dir alpha
+	no files found
+.fi
+
+6. Print the contents of the directory USR$2:[IRAF.LOCAL] on the remote VMS
+node "draco" (requires IRAF network access to the remote node).
+
+.nf
+	cl> dir draco!usr\$2:\[iraf.local]
+.fi
+.ih
+BUGS
+There is no provision for wildcarding directories, e.g., "dir */*.x".
+The long format listing can currently only be sorted by filename (although
+the \fIsort\fR program may be used in a pipe).  The file existence test will
+not be performed if individual files are named as list elements within
+a filename template.
+.ih
+SEE ALSO
+files, pathnames
+.endhelp
diff --git a/pkg/system/doc/diskspace.hlp b/pkg/system/doc/diskspace.hlp
new file mode 100644
index 00000000..538f308d
--- /dev/null
+++ b/pkg/system/doc/diskspace.hlp
@@ -0,0 +1,33 @@
+.help diskspace Feb85 system
+.ih
+NAME
+diskspace -- summarize disk space utilization
+.ih
+USAGE
+diskspace
+.ih
+DESCRIPTION
+\fIDiskspace\fR lists the disk devices available on the host system,
+giving the name, capacity, and free space available on each device.
+The format and contents of the output are host system dependent.
+.ih
+EXAMPLE
+
+1. Print the available diskspace (UNIX host).
+
+.nf
+	cl> disk
+	Filesystem    kbytes    used   avail capacity  Mounted on
+	/dev/hp0a       7421    5312    1366    80%    /
+	/dev/hp0g      38430    5340   29246    15%    /tmp
+	/dev/hp0h      51598   38848    7590    84%    /usr
+	/dev/ra0a       8007    5657    1549    79%    /mnt
+	/dev/ra0d      88179   46043   33318    58%    /u1
+	/dev/ra0e      54382   46694    2248    95%    /local
+	/dev/ra0f     145435   93733   37158    72%    /u2
+	/dev/ra0g      28720     350   25497     1%    /usr/spool
+	/dev/ra0h      79798   41696   30122    58%    /iraf
+	/dev/ra1h     397102  306688   50702    86%    /tmp2
+	/dev/ra2h     397102  275010   82380    77%    /tmp3
+.fi
+.endhelp
diff --git a/pkg/system/doc/fcache.hlp b/pkg/system/doc/fcache.hlp
new file mode 100644
index 00000000..b5230307
--- /dev/null
+++ b/pkg/system/doc/fcache.hlp
@@ -0,0 +1,140 @@
+.help fcache Jun2011 system
+.ih
+NAME
+fcache -- list, clean or manipulate the file cache
+.ih
+USAGE
+fcache cmd
+.ih
+PARAMETERS
+.ls cmd
+Cache command to execute.  A description of each command is given below.
+.le
+.ls pattern = "*"
+Filename substring pattern to match when initializing the cache with
+the \fIinit\fR command.
+.le
+.ls src = ""
+Source string used to generate the cache filename.  This is typically
+the full path to a local file being cached or a URL.
+.le
+.ls fname = ""
+Name of the file in the cache.
+.le
+.ls extn = ""
+Cache filename extension.
+.le
+.ls age = -1
+Age of the file (in days) to be purged with the \fIpurge\fR command.  A value
+less than zero means that the \fIcache_age\fR environment variable should 
+is used to set the age, a value of zero means to delete all files in the 
+cache  (same as the \fIinit\fR command), a value greater than zero means 
+that files older than this age will be deleted.
+.le
+.ls verbose = no
+Print status information as the task processes the command.
+.le
+.ls wait = yes
+Block on operation?  If 'yes' then the task will block until the requested
+file becomes available in the cache.
+.le
+.ls cache = "cache$"
+Cache directory to be used.
+.le
+
+.ih
+DESCRIPTION
+The \fIFCACHE\fR command is used to list or manage the system file cache
+named by the \fIcache\fR parameter.  If the \fIcache\fR directory does not
+exist, it will be created when required.  The \fIcache_age\fR environment
+variable determines the default maximum age of files in the cache, older
+files are automatically removed by the login.cl as part of the startup
+process.
+
+The IRAF file cache is used primarily to cache local copies of URLs in the
+system to prevent repeated downloads when accessing URLs from tasks.  This
+allows a URL to be passed to multiple tasks without explicitly requiring
+the user to create a named (temporary) file themselves.
+
+The \fIcmd\fR parameter determines the action to take, other parameters are
+used as needed depending on the command according to the following table:
+
+.nf
+    Command	Input Pars	Output Pars	Action
+    -------	----------	-----------	------
+    init	pattern				Initialize the cache
+    purge	age				Purge old files
+    destroy					Destroy the cache
+    list					List cache contents
+    lookup	src		fname,extn	Lookup a file in the cache
+    access	src				Is file in cache?
+    add		src extn	fname		Add file to the cache
+    delete	src		fname		Delete file from cache
+    wait	src				Wait for access to file
+.fi
+
+The \fIlookup\fR command works in two ways:  If a \fIsrc\fR string is
+provided then the \fIfname\fR parameter will contain the matching cached
+file (and \fIextn\fR will contain the optional extension), if the \fIfanme\fR
+parameter is specified then on output \fIsrc\fR will contain the original
+filename/URL.
+
+.ih
+EXAMPLES
+
+1. Remove all "url" files from the cache.
+.nf
+    cl> fcache init pattern="url"
+.fi
+
+2. List the contents of the file cache.
+.nf
+    cl> fcache list
+.fi
+
+3. Destroy a cache directory (i.e. remove it entirely).
+.nf
+    cl> fcache destroy cache="/tmp/cache"
+.fi
+
+4. Purge all cache files older than 7 days:
+.nf
+    cl> fcache purge age=7
+.fi
+
+5. Determine if a URL is already in the cache:
+.nf
+    cl> fcache add src="/tmp/dpix.fits"
+    cl> fcache list
+         f1128531670  1  /tmp/dpix.fits
+          f789045894  1  http://iraf.noao.edu/vao/dpix.fits
+    cl> fcache access src="/tmp/dpix.fits"
+    yes
+    cl> fcache access src="http://iraf.noao.edu/vao/dpix.fits"
+    yes
+.fi
+
+6. Delete a cached URL:
+.nf
+    cl> fcache delete src="http://iraf.noao.edu/vao/dpix.fits"
+.fi
+
+7. Add a local file to the cache, then look it up:
+.nf
+    cl> fcache add src="/tmp/test.fits"
+    cl> fcache lookup src="/tmp/test.fits"
+    cl> =fcache.fname
+    f1295587026
+    cl> fcache lookup fname="f1295587026"
+    cl> =fcache.src
+    /tmp/test.fits
+.fi
+
+
+.ih
+BUGS
+
+.ih
+SEE ALSO
+head
+.endhelp
diff --git a/pkg/system/doc/files.hlp b/pkg/system/doc/files.hlp
new file mode 100644
index 00000000..f2a0996f
--- /dev/null
+++ b/pkg/system/doc/files.hlp
@@ -0,0 +1,72 @@
+.help files Jun86 system
+.ih
+NAME
+files -- expand a file name template into a list of files
+.ih
+USAGE
+files template
+.ih
+PARAMETERS
+.ls template
+A file name template specifying the set of files to be listed.
+.le
+.ls sort = "yes"
+Sort the file list.
+.le
+.ih
+DESCRIPTION
+\fIFiles\fR lists all files matching the given template.  The existence of
+the listed files is checked only if pattern matching is used, hence \fIfiles\fR
+may also be used to parse a comma delimited list of strings which are not
+necessarily filenames.  \fIFiles\fR performs the same function as "dir l+"
+but is simpler and more convenient to use when generating file lists.
+
+The \fIfiles\fR task and all other tasks which operate upon groups of files
+use the \fBfile template\fR facility to specify the set of files to be
+operated upon.  This should not be confused with the \fBimage template\fR
+facility, used by tasks which operate upon sets of images and which is
+documented in the manual page for the \fIsections\fR task.
+
+Pattern matching in a file template is provided by the usual pattern matching
+meta-characters "*?[]", documented in the CL User's Guide.  Pattern matching 
+is used to select files from one or more directories.  In addition, the
+filename template notation provides two operators for generating new filenames
+from the matched filenames.  These are the \fBconcatenation\fR operator "//",
+and the \fBstring substitution\fR operator "%chars%newchars%".
+The concatenation operator concatenates either a prefix to a filename,
+or a suffix to the root of a filename.  The string substitution operator
+uses the "chars" to match filenames, and then replaces the "chars" by the
+"newchars" to generate the final output filename.  Either string may be null
+length to insert into or delete characters from a filename.
+.ih
+EXAMPLES
+
+1. Generate a single column list of files in the current directory,
+sorted in ASCII collating sequence.
+
+	cl> files
+
+2. Generate an unsorted single column list of files in logical directory
+"lib$".  Each entry in the output list is of the form "lib$...".
+
+	cl> files lib$ sort-
+
+3. Generate a file list to be used to make a set of new files.  The new file
+names will be the old file names with "_1" concatenated to the root, e.g.,
+"root.x" would map to "root_1.x" and so on.
+
+	cl> files root.*//_1
+
+4. Generate a file list similar to that in [3], adding a directory prefix
+to each filename.
+
+	cl> files dir$//root.*
+
+5. Use string substitution to change the filename extension of a set of files
+to ".y".
+
+	cl> files root.%*%y%
+.ih
+SEE ALSO
+directory, pathnames, images.sections
+.endhelp
diff --git a/pkg/system/doc/gripes.hlp b/pkg/system/doc/gripes.hlp
new file mode 100644
index 00000000..3938e1ea
--- /dev/null
+++ b/pkg/system/doc/gripes.hlp
@@ -0,0 +1,67 @@
+.help gripes Mar86 system
+.ih
+NAME
+gripes -- send gripes or suggestions to the system
+.ih
+USAGE
+gripes subject
+.ih
+PARAMETERS
+.ls subject
+The subject of the gripe; any string, usually the name of package or task
+to which the gripe refers.
+.le
+.ls verbose = yes
+Print instructions on how to enter text whenever \fIgripes\fR is run.
+.le
+.ih
+DESCRIPTION
+The \fBgripes\fR task is used to post complaints, suggestions, or any other
+formal or informal comments regarding the IRAF system.  Each gripe is
+appended to the system gripesfile "hlib$gripesfile", a public file which
+can be read by anyone by simply typing "page hlib$gripesfile" within IRAF.
+Use \fBtail\fR instead of \fBpage\fR to read only the most recent gripes.
+A copy of each gripe is also sent immediately to one or members of the IRAF
+group via electronic mail, to insure that the gripe gets read promptly (this
+feature is not available on all host systems).
+
+Gripe text is read from the standard input.  A line containing only a period
+terminates the gripe, as does the end of file character (e.g., <ctrl/z>).
+If the line containing only "~e" is entered a text editor will be called up
+to edit the text of the gripe.
+
+Users are encouraged to use the gripe facility at will.  Be assured that
+someone will at least read the gripe, although there is no guarantee that
+any action will be taken.  In many cases there will be no response from the
+system, but nonetheless the gripe will be seen and it may well influence
+the direction in which the system evolves.  Do not avoid posting gripes for 
+fear that you do not understand something about the system; if enough users
+find some aspect of the system or a program confusing then that is more
+than sufficient reason for a gripe.
+.ih
+EXAMPLES
+
+1.  The user has discovered some nasty features of the \fIimdelete\fR task
+and enters the following gripe.  Note the use of the "." to terminate
+the text.
+
+.nf
+	cl> gripe
+	Subject: image deletion
+	Enter text of gripe.  Type <eof> or '.' to quit:
+
+	IMDEL * will delete non image files as well as images!
+	It should be possible to delete images with the normal
+	DELETE command.
+	.
+	cl>
+.fi
+.ih
+BUGS
+There is currently no provision for communicating gripes from a remote site
+back to the site that wrote the software, unless some person manually mails
+a gripe (or the accumulated gripesfile).
+.ih
+SEE ALSO
+news
+.endhelp
diff --git a/pkg/system/doc/head.hlp b/pkg/system/doc/head.hlp
new file mode 100644
index 00000000..0b2f6393
--- /dev/null
+++ b/pkg/system/doc/head.hlp
@@ -0,0 +1,39 @@
+.help head Nov84 system
+.ih
+NAME
+head -- print the first few lines of the specified files
+.ih
+USAGE
+head files
+.ih
+PARAMETERS
+.ls files
+The list of files to be dealt with, quite possibly given as
+a template, such a "image*".
+.le
+.ls nlines = 12
+The number of lines to be printed.
+.le
+.ih
+DESCRIPTION
+\fIHead\fR prints, on the standard output, the first \fInlines\fR of each
+file that matches the given file list.  If the file list has more than one
+name in it, a short header precedes each listing.
+.ih
+EXAMPLES
+
+1. Print the first 12 lines of each help file in the current directory.
+
+	cl> head *.hlp
+
+2. Print the first line of each help file.
+
+	cl> head *.hlp nl=1
+
+3. Print the most recently defined \fIset\fR environment definitions.
+
+	cl> set | head
+.ih
+SEE ALSO
+tail, page
+.endhelp
diff --git a/pkg/system/doc/help.hlp b/pkg/system/doc/help.hlp
new file mode 100644
index 00000000..74a53917
--- /dev/null
+++ b/pkg/system/doc/help.hlp
@@ -0,0 +1,599 @@
+.help help Oct01 system
+.ih
+NAME
+help -- print online documentation for the named modules or packages
+.ih
+USAGE
+help [template]
+.ih
+PARAMETERS
+.ls template
+A string listing the modules or packages for which help is desired.
+Each list element may be a simple name or a pattern matching template.
+Abbreviations are permitted.  If \fItemplate\fR is omitted a long format
+menu will be printed for the current package, listing each task (or
+subpackage) and describing briefly what it is.
+.le
+.ls file_template = no
+If this switch is set the template is interpreted as a filename matching
+template, and all help blocks found in the named files are output.  The help
+database is not searched, hence manual pages can be printed or documents
+may be formatted without entering the files into a help database.
+In other words, "help file.hlp fi+" makes it possible to use \fIhelp\fR as
+a conventional text formatter.
+.le
+.ls all = no
+Print help for all help modules matching \fItemplate\fR, rather than only the
+first one found.
+.le
+.ls parameter = "all"
+If the value of this parameter is not "all", only the help text
+for the given parameter will be printed.
+.le
+.ls section = "all"
+If the value of this parameter is not "all", only the help text for the
+given section (e.g. "usage", "description", "examples") will be printed.
+.le
+.ls option = help
+The option parameter specifies the type of help desired, chosen from
+the following:
+.ls help
+Print the full help block for the named module.
+.le
+.ls source
+Print the source code for the module (which often contains additional
+detailed comments).
+.le
+.ls sysdoc
+Print the technical system documentation for the named module.
+.le
+.ls directory
+Print a directory of all help blocks available for the named package.
+.le
+.ls alldoc
+Print all help blocks in the file containing the help block for
+the named procedure (i.e., both the user and system documentation).
+.le
+.ls files
+Print the names of all help files associated with the named modules or
+packages.
+.le
+.ls summary
+Print only the titles and sizes of help blocks in referenced help files.
+The contents of the blocks are skipped.  Titles are printed for \fIall\fR
+help blocks found in the file containing the help block for the named module.
+.le
+.le
+.ls page = yes
+Pause after every page of output text.  Turning this off for large documents
+speeds up output considerably.
+.le
+.ls nlpp = 59
+The number of lines per page if output is redirected, e.g., to \fIlprint\fR.
+.le
+.ls lmargin = 1
+Left margin on output.
+.le
+.ls rmargin = 72
+Right margin on output.
+.le
+.ls search = no
+If enabled the 
+.hr #l_template template
+is interpreted as a search string and the task
+is started with the search panel open with the results of the search.  The
+.hr #l_file_template file_template
+parameter is ignored with search turned on.
+.le
+.ls home = ""
+The home page for the task.  If not set and no 
+.hr #l_template template
+is specified
+the task will start with the online help in the main window, otherwise it
+may be set to a filename to be displayed when the task starts.  This file
+may contain a text help block which will be formatted before display,  or
+it may be a valid HTML file.  See below for a description of the format of
+a homepage file which provides links to tasks.
+.le
+.ls printer = "printer"
+Default hardcopy printer name. If the \fIvalue\fR of the parameter is the
+reserved string "printer", the actual device is the value of the CL
+environment variable \fIprinter\fR.
+.le
+.ls showtype = no
+Add task-type suffix in package menus?
+.le
+.ls quickref = "uparm$quick.ref"
+Name of the quick-reference file used for searching.  This file is created
+the first time the task is run in GUI mode or whenever it doesn't exist, 
+or when any help database file has been updated.
+.le
+.ls uifname = "lib$scr/help.gui"
+The user interface file.   This file is what defines the look and behavior
+of all the graphical user interface elements.   Experts may create variants
+of this file.
+.le
+.ls helpdb = "helpdb"
+The filename of the help database to be searched.  If the \fIvalue\fR of the
+parameter is the reserved string "helpdb", the actual filename is the value
+of the CL environment variable \fIhelpdb\fR.
+.le
+.ls device = "terminal"
+Output device if the standard output is not redirected.  Allowable values
+include:
+.ls terminal
+If the \fIvalue\fR of
+the parameter is the reserved string "terminal",  the actual device name is
+the value of the CL environment variable \fIterminal\fR.  
+.le
+.ls text
+Output the formatted help page as plain text.
+.le
+.ls gui
+Invoke the GUI for browsing the help system.  This option will only work if
+the \fIstdgraph\fR environment variable is set the \fIxgterm\fR, and the
+user is running IRAF from an \fIXGterm\fR window.
+.le
+.ls html
+Output the formatted help page as HTML text.
+.le
+.ls ps (or postscript)
+Output the formatted help page as postscript.
+.le
+.le
+.ih
+BASIC USAGE
+Despite the complex appearing hidden parameters, \fBhelp\fR is easy to use
+for simple tasks.  \fBHelp\fR is most commonly used to get help on the current
+package, and to get help on a program named in a CL menu.  To get help on
+the current package one need only type \fBhelp\fR without any arguments.
+For example, if the current package is \fBplot\fR, the command and its output
+might appear as follows:
+
+.nf
+	pl> help
+		contour - Make a contour plot of an image
+		  graph - Graph one or more image sections or lists
+		   pcol - Plot a column of an image
+		  pcols - Plot the average of a range of image columns
+		   prow - Plot a line (row) of an image
+		  prows - Plot the average of a range of image lines
+		surface - Make a surface plot of an image
+	pl>
+.fi
+
+To get help on a module one supplies the module name as an argument,
+
+	pl> help graph
+
+and the manual page for the \fBplot.graph\fR program will be printed on the
+terminal.  To get a hardcopy of the manual page on the printer, the output
+may be redirected to the line printer, as follows:
+
+	pl> help graph | lprint
+.ih
+DESCRIPTION
+The function of the \fBhelp\fR program is to perform a depth first search
+of the help database \fIhelpdb\fR, printing help for all packages and modules
+matching the template.  By default the standard IRAF help database is searched,
+but any other help database may be searched if desired.  A help database is
+precompiled with the \fBmkhelpdb\fR program to speed up runtime searches for
+help modules.  The standard IRAF help database contains the documentation and
+source for all CL programs and system and math library procedures installed
+in IRAF.
+
+A help template is a string type parameter to the CL.  The form of a template
+is a list of patterns delimited by commas, i.e.,
+
+	"pattern1, pattern2, ..., patternN"
+
+The form of a pattern is
+
+	package_pattern.module_pattern
+
+If the "." is omitted \fImodule_pattern\fR is assumed.  The standard pattern
+matching meta-characters, i.e., "*?[]", are permitted in patterns.
+Simple patterns are assumed to be abbreviations.
+
+.ih
+GUI OPERATION
+
+The GUI component of the task is a front-end to the IRAF 
+.hr system.help \fBhelp\fR
+task which provides on-the-fly conversion of help documents to HTML for
+presentation in the GUI or formatted PostScript for hardcopy.  
+The GUI is started by setting the 
+.hr #l_device \fIdevice\fR
+parameter to the special value \fIgui\fR, it is only available when using
+an XGterm window to start IRAF and assuming the \fIstdgraph\fR environment
+variable is set to \fRxgterm\fR.
+
+Help pages may be loaded on the command line, through use of a
+file browser, or by navigating the help databases using a familiar CL
+package menu scheme.   It also features a search capability similar to the 
+.hr system.references \fBreferences\fR
+task and a complete history mechanism. 
+
+When invoked with no command line arguments the task starts as a browser
+and the user is presented with a GUI that has the toplevel CL package menu
+in the upper navigation window.  The main display window below will contain
+any help page specified in the 
+.hr #l_template template
+parameter or loaded on
+the command line by specifying the 
+.hr #l_template template
+and 
+.hr #l_file_template file_template
+parameters. If the 
+.hr #l_search search
+parameter is enabled the 
+.hr #l_template template
+is taken to be a search phrase and the database is searched for tasks
+matching the keyword and the GUI will appear with the search panel mapped
+so the user can select the task help to
+view.  When no 
+.hr #l_template template
+is given the main display window will start with the page specified by the 
+.hr #l_home home
+parameter, this can be a user-defined HTML file giving links to specific tasks
+(see below for details) or if 
+.hr #l_home home
+is empty the display will contain the online help for the task.
+
+The first time the task is run, or whenever the help database is updated,
+a quick reference file (specified by the task 
+.hr #l_quickref quickref 
+parameter) and package menu file will be created in the user's \fIuparm\fR
+directory to speed up help searching and subsequent startups of the task.
+
+.ih
+NAVIGATING THE HELP SYSTEM
+When run as a GUI browser \fIHELP\fR works very much like any WWW browser.
+The top panel is a list widget that will always contain a CL package listing,
+at startup this will be the toplevel \fI"Home"\fR package menu one would see
+when first logging into the CL containing the core system packages, NOAO
+package, and any site-specific external package, or in the case of starting
+with a specific task it will be the parent package for the task.  Additionally,
+system documents for the 
+.hr os \fBos\fR
+HSI routines and the 
+.hr sys.imfort \fBimfort\fR
+and
+.hr math \fBmath\fR
+interfaces will be available in the \fIHome\fR package although
+these are programmatic interfaces and not tasks which can be executed.
+
+New packages or task help pages are loaded by selecting an item from the 
+package menu list using the left mouse button.  If the requested item is a 
+package, the menu listing will change as though the package were loaded in
+the CL, and the help display panel will contain a listing of the package
+tasks with a one-line description for each task such as would be seen with 
+a \fI"help <package>"\fR command using the standard task.  New items may then
+be selected using either the menu list or links in the display panel.  If the
+item is a task, the help page for the task will appear in the display panel.
+In either case new pages may be selected from the menu listing.  
+
+Specific help documents may also be requested by entering the task/package
+name in the \fBTopic\fR text widget above the menu list.  As when selecting
+from the package menu list, items selected this way will cause the menu
+list to change to the package menu for the parent package if the item is a
+task (displaying the help page in the display panel) or the package menu
+if the item is a package (displaying the one-liner package listing in the
+display panel).
+
+Using the \fBBack\fR button will revert to the previous page in the history
+list which will either be the previously loaded package or help page.
+Similarly, selecting the \fBForward\fR button will move the next page further
+down in the history list, either button will become insensitive when the 
+end of the list on either end is reached.  Selecting the \fBUp\fR button will
+cause the browser to immediately jump up the previous package, skipping 
+over any help pages that were loaded in between.  The \fBHome\fR button will
+cause the default homepage (either the user-defined page if specified by the
+task \fIhome\fR parameter or the online help) to be displayed.  Browsing
+in this way can also be done using the navigation menu created by hitting
+the right mouse button while in the main display panel.
+
+Users can also jump to specific pages in the history list using the
+\fBHistory\fR button on the main menubar.   The right column of the menu
+will indicate whether the item is a task, package, internal link or a text
+file.  The history list is truncated at about 40 entries in the menu but
+the user may work back incrementally by selecting the last item of the 
+menu, after which the History button will display the previous 40 entries.
+The history list may be cleared except for the current page by selecting
+the \fIClear History\fR menu item.
+
+.ih
+BROWSING A HELP DOCUMENT
+Once a help page is loaded the middle menubar above the display panel
+will change to activate widgets based on the position within the history
+list and options available for a particular page.  The left-most group
+of buttons are the standard navigation buttons described above.
+The middle group of buttons contains the \fBSections\fR and
+\fBParameters\fR buttons which are used to browse within a help document.
+The \fISections\fR button is a menu listing all of the sections found
+within a help page, allowing the user to jump to a specific section
+rather than scrolling through the entire document. The \fISections\fR
+menu is also available using the middle mouse button from the
+main display area.  The \fIParameters\fR button is similarly a menu
+listing of all task parameter help sections found within the document.
+Both or either of these buttons will become insensitive when no section
+or parameter information is found in the document.
+
+The right-most group of buttons represent the various help options available
+for each page.  The default is to get the task help, however help pages
+may have an associated \fBsource\fR file or \fBsysdoc\fR (e.g. if the task is
+a CL script there may be a pointer to the script source itself, or a package
+may have a general overview document listed as the system document).  Once
+a help page is loaded these buttons will change become sensitive if that option
+is available, simply select the button to view the option.  Selecting the
+\fBFiles\fR button will bring up a panel listing all the files associated
+with a particular help topic.  When a help topic is selected and an option is
+defined but the file does not exist, the options button will display a yellow
+diamond icon even if the button is insensitive, a green icon indicates the
+currently selected option.  This feature may be disabled by selecting the
+"Show missing files" item from the main menubar \fBOptions\fR menu.
+
+.ih
+SEARCHING
+Searching the help database is done by selecting the \fBSearch\fR button
+from the main menubar to bring up the search panel.  Users may then enter 
+one or more keywords into the \fBTopic\fR field at the bottom of the panel
+and initiate the search with either a carriage return or hitting the
+\fISearch\fR button just beside it.  The panel will then show a list of all
+tasks and packages which match the search phrase along with a one-line
+description of the task.  Help pages may be displayed by selecting either the
+task or package link with the left mouse button, in both case the package
+menu list on the main help window will be updated to list the package
+contents allowing other tasks from that package to be selected in the normal
+way.
+
+By default the exact phrase entered in the topic window will be used for the
+search.  This can be relaxed by toggling the  "Require exact match" button
+at the top of the panel.  For example,  to search for all tasks matching
+\fIeither\fR the keyword "flat" or "field" turn off the exact match
+toggle and the search will return not only tasks matching "flat field" but 
+also any task description containing only one of the words such as the
+VELVECT task which plots velocity \fIfield\fRs.
+
+Within a help document itself one can search for a string by selecting
+the \fBFind\fR button from the main menubar to bring up a panel used to
+enter the search string.  When the text is entered the main display 
+window will reposition itself and highlight the text found within the
+document.  Searches can be repeated and will wrap around the document
+automatically, searches can be done either forward or backward through
+the text and may be case insensitive.
+
+.ih
+USER_DEFINED HOME PAGES
+By default the \fIhelp\fR GUI will start with the online help page displayed
+in the main help window.  The user can change this by setting the task
+\fBhome\fR parameter to be a path to any valid file.  This file may be plain
+text, a help document in LROFF format which will be converted to HTML for
+display, or a native HTML document.
+
+HTML files may contain URLs of the form
+.nf
+	\fB<a href=\fI[package.]task\fB>\fIurl_text\fB</a>
+.fi
+
+where \fIurl_text\fR is the text to appear in the window and the URL itself
+consists of an optional package and task name delimited by a period.  For
+example, to create a link to the 
+.hr onedspec.splot \fBsplot\fR
+task in a document one would use the URL
+.nf
+	\fB<a href=onedspec.splot>splot</a>\fR
+.fi
+
+In this way users can create a homepage which serves as a \fI"bookmark"\fR
+file or index of shortcuts to the most commonly accessed help pages.
+
+.ih
+LOADING FILES
+Text files may be loaded on the command line when starting the task by
+specifying the filename and setting the
+.hr #l_file_template file_template
+task parameter.  The named file
+will be searched for a \fI.help\fR LROFF directing indicating it contains
+a help block that will be converted to HTML for display.  If no help
+block is found the file will be displayed as-is, meaning existing
+HTML documents can be loaded and will be formatted correctly.
+
+Once the task is running users may load a file by selecting the \fBOpen
+File...\fR menu item from the main menubar \fBFile\fR menu or the
+right-mouse-button menu from within the main display area.  This will
+open a file browser allowing users to change directories by using the
+navigation buttons at the top of the panel, or selecting items from the
+leftmost directory listing.  Selecting a file on the rightmost list will
+cause it to be loaded and automatically formatted if it contains a help
+block.  The file list may be filtered to select only those files matching
+a particular template by changing the \fBFilter\fR box at the top of
+the panel.  Filenames or directories may be entered directly using the
+\fBSelection\fR box at the bottom of the panel.
+
+.ih
+SAVING FILES
+Once a file has been loaded in the browser it may be saved to disk as 
+either \fIsource\fR (i.e. the original LROFF file if that was converted
+for the display, or whatever file is currently displayed regardless of
+format), \fItext\fR to save formatted plain text such as that produced
+by the standard \fBhelp\fR task, \fIHTML\fR to save the converted HTML
+used in the display, or \fIPostScript\fR to save formatted PostScript of
+the document such as that sent to the printer using the \fBPrint\fR 
+button.  Not all options will be available depending on the format of the
+input text, unavailable options will be insensitive in the GUI.
+
+The \fBSave\fR panel is opened by selecting the \fBSave As...\fR menu
+item from the  main menubar \fBFile\fR menu or the right-mouse-button
+menu from within the main display area.   The file browser operates the
+same as when loading images, the only difference is that file selection 
+simply defines the filename to be used and does not cause the save to
+occur automatically.  Users can overwrite existing files by selecting the
+\fIOptions\fR toggle at the bottom of the panel.
+
+.ih
+HARDCOPY OUTPUT AND SAVING DISK FILES.
+Help pages may be output to any configured IRAF printer by selecting the
+main menubar \fBPrint\fR button to bring up the print panel.  Task help pages
+will be converted to formatted PostScript and may be sent to either a
+printer or saved to disk depending on the selection made in the printer 
+panel.  If the printer name is set to the special value \fI"printer"\fR then
+the device named by the CL \fIprinter\fR environment variable will be used.
+When saving to disk files the default action is to save to a filename whose
+name is the task name plus a ".ps" extension.  Either of these are changeable
+within the GUI as is the default page size to be used when generating the
+PostScript.
+
+The main menubar \fBFile\fR button can also be used to bring up the file
+browser in order to save the current document to disk.  Help pages may be
+saved as either the origin LROFF source for the file, formatted text as you
+would get from the standard help task, HTML as is displayed in the GUI, or
+formatted PostScript.  The choice of formats is dictated by the type of file
+being displayed (e.g. you cannot save PostScript of a program source).
+
+.ih
+LROFF DIRECTIVE EXTENSIONS FOR HTML
+To better support HTML links within documents and to other help pages two
+new directives have been added to the LROFF text formatter.  These are
+\fB.hr\fR to specify a link (an HTML \fIHREF\fR directive) and \fB.hn\fR
+to specify a name (an HTML \fINAME\fR directive).  The syntax for these are
+as follows:
+.nf
+
+	\fB.hn\fI <name>\fR
+	\fB.hr\fI <link> <text> \fR
+.fi
+
+where \fI<name>\fR is the destination name of an internal link, \fI<link>\fR
+is the URL of the link to be created, and \fI<text>\fR is the text to be
+displayed in the HTML.  The URL syntax is either a '#' character followed
+by a destination name, a simple \fItask\fR name or \fIpackage\fR name,
+or a \fIpackage.task\fR pair giving a more precise task.  For internal links
+the current document is repositioned so the name is at the top of the display,
+for task help links new help pages will be loaded in the browser.  
+
+These directives are ignored when converting the LROFF to either formatted
+plain text or PostScript.
+
+.hn examples_target
+.ih
+GUI EXAMPLES
+1) Start \fIhelp\fR as a GUI browser:
+.nf
+
+	cl> help dev=gui
+.fi
+
+2) Begin by searching for the phrase 'gauss', tasks and packages may be
+selected from the search panel which will appear when the task starts:
+.nf
+
+	cl> help gauss dev=gui search+
+.fi
+
+3) Load an LROFF help page in the browser at startup
+.nf
+
+	cl> help mytask.hlp dev=gui file+
+.fi
+
+.ih
+EXAMPLES
+
+1. Print the help text for the program \fIdelete\fR in the package
+\fIsystem\fR (output will be directed to the terminal):
+
+.nf
+	cl> help system.delete
+or
+	cl> help delete
+or
+	cl> help del
+.fi
+
+2. Print the help text on the line printer:
+.nf
+
+	cl> help delete | lprint
+.fi
+
+3. Print help for the current package:
+.nf
+
+	cl> help
+.fi
+
+4. Print the usage section of all modules in the package \fBimages\fR:
+.nf
+
+	cl> help images.* section=usage
+.fi
+
+5. Print a directory of all help blocks in the packages \fBclpackage\fR
+and \fBclio\fR (and any others whose names begin with the string "cl"):
+.nf
+
+	cl> help cl* op=dir
+.fi
+
+6. Print a directory of each package in the database (useful for getting an
+overview of the contents of a help database):
+.nf
+
+	cl> help * op=dir
+.fi
+
+7. Print the source for all of the string utilities in the system library
+package \fBfmtio\fR:
+.nf
+
+	cl> help fmtio.str* op=source
+.fi
+
+8. Find all tasks that delete something:
+.nf
+
+	cl> help * | match delete
+.fi
+
+9. Print the manual pages for the \fIhelp\fR and \fIlprint\fR tasks on the
+default printer device:
+.nf
+
+	cl> help help,lprint | lpr
+.fi
+
+10. Capture the manual page for task \fIhedit\fR in a text file, in a form
+suitable for printing on any device.
+.nf
+
+	cl> help hedit dev=text > hedit.txt
+.fi
+
+11. Print the manual page for task \fIhedit\fR as a Postscript file. 
+.nf
+
+	cl> help hedit dev=ps | lprint
+.fi
+
+.ih
+BUGS
+On some systems, typing the next command keystroke before the end-of-page
+prompt is printed may result in the character being echoed (messing up the
+output) and then ignored when raw mode is enabled for the prompt.
+
+.ih
+SEE ALSO
+.hr system.references references
+,
+.hr system.phelp phelp
+,
+.hr system.mkhelpdb mkhelpdb
+,
+.hr system.hdbexamine hdbexamine
+,
+.hr system.mkmanpage mkmanpage
+,
+.hr system.lroff lroff
+, the online task help documents.
+.endhelp
diff --git a/pkg/system/doc/lprint.hlp b/pkg/system/doc/lprint.hlp
new file mode 100644
index 00000000..33997b0e
--- /dev/null
+++ b/pkg/system/doc/lprint.hlp
@@ -0,0 +1,65 @@
+.help lprint Nov84 system
+.ih
+NAME
+lprint -- print a file or files
+.ih
+USAGE
+lprint files
+.ih
+PARAMETERS
+.ls files
+A filename template specifying the files to be printed.
+.le
+.ls device = "printer"
+The output device.  If the value of \fIdevice\fR is the reserved string
+"printer", the name of the actual printer device is taken from the value
+of the environment variable "printer".
+.le
+.ls map_cc = yes
+If set to "yes", any unprintable characters embedded in the text are printed
+in the form "^X", where ^A is <ctrl/A> (ASCII 1), and so on.
+.le
+.ls paginate = "auto"
+If \fIpaginate\fR is set to "auto" and the standard input is not redirected,
+pages are broken and a header is printed at the top of each page.
+If \fIpaginate\fR is set to "auto" and the standard input \fIis\fR redirected,
+the input text is not paginated, allowing proper operation when \fIlprint\fR
+is used in a pipe, e.g., taking input from \fIhelp\fR.
+If "paginate" is set to "yes", pages are broken even if the input text
+is being read from STDIN.
+.le
+.ls label = "STDIN"
+If displaying a header with input from the standard input, use the
+"label" string where the filename would appear in a normal header.
+.le
+.ih
+DESCRIPTION
+The named files, or the standard input, are printed on the standard
+line printer device.  Each file is printed starting at the top of a new
+page, with a header giving the page number and the date of last modification
+for the file.  Pagination and headers are normally suppressed when reading
+input from the standard input, but may be enabled if desired.
+.ih
+EXAMPLES
+
+1. Print all files with an extension of either ".x" or ".h", followed by
+all files with the extension ".com".  Note that filename sorting occurs only
+within a comma delimited field of the filename template, hence the "*.[xh]"
+files are printed in sort order, followed by the ".com" files.
+
+	cl> lprint *.[xh],*.com
+
+2. Print the output of the \fIimstat\fR task on the versatec printer,
+paginating the output with the given label on each page.  Note that the
+command may be broken after the "pipe" character without need for
+explicit backslash continuation.
+
+.nf
+	cl> imstat nite1.* |
+	>>> lprint pag+ label="Image Statistics" device=versatec
+.fi
+
+.ih
+SEE ALSO
+type
+.endhelp
diff --git a/pkg/system/doc/match.hlp b/pkg/system/doc/match.hlp
new file mode 100644
index 00000000..3338af3d
--- /dev/null
+++ b/pkg/system/doc/match.hlp
@@ -0,0 +1,77 @@
+.help match Nov84 system
+.ih
+NAME
+match -- match a pattern against the lines in a file or files
+.ih
+USAGE
+match pattern files
+.ih
+PARAMETERS
+.ls pattern
+The pattern to be matched.  A pattern may contain any of the
+pattern matching \fImeta-characters\fR described below.
+.le
+.ls files
+A template specifying the file or files to be searched.  Omitted if the
+standard input is redirected.
+.le
+.ls meta-characters = yes
+Set to "no" to disable the pattern matching meta-characters, e.g., when
+you want to explicitly match one of the meta-characters as a regular character.
+.le
+.ls stop = no
+If \fIstop\fR is enabled, lines with match the pattern are "stopped" (not
+passed to the output), otherwise only those lines with match the pattern
+are output.
+.le
+.ls print_file_names = yes
+If more than one file is being searched, preface each printed line
+with the "file_name: ".
+.le
+.ih
+DESCRIPTION
+The listed files are searched for the given pattern, copying each line that
+matches to the standard output.  If "stop" is set the action is reversed,
+i.e., all lines are passed on to the output except those which match the
+pattern.  If no files are named text is read from the standard input.
+The pattern matching meta-characters are described in the table below.
+
+.nf
+	^		matches the beginning of a line
+	$		matches the end of a line
+	?		matches any single character
+	*		matches zero or more of whatever is at the left
+	[12345]		matches any single character in the given set
+	[1-5]		matches any single character in a range
+	[^a-z]		matches any character NOT in the range/set
+	#		matches whitespace
+	{chars}		turns off case sensitivity inside the braces
+	\		used to include a meta-character in the pattern
+.fi
+
+If more than one file is being searched, each output line is prefixed
+with its file name.
+.ih
+EXAMPLES
+
+1. From all the lines displayed by "set", print only those that have
+the string "tty" somewhere in them.
+
+	cl> set | match tty
+
+2. Find all tasks that delete something.
+
+	cl> help * | match delete
+
+3. Delete all the "red" objects from the list file "catalog".
+
+	cl> match red catalog stop+ > newcatalog
+
+4. Type out the file "spool", omitting all lines that end in a colon,
+and paginating the output.
+
+	cl> match ":$" spool stop+ | page
+.ih
+SEE ALSO
+lcase, ucase, translit, sort, unique
+.endhelp
diff --git a/pkg/system/doc/mkdir.hlp b/pkg/system/doc/mkdir.hlp
new file mode 100644
index 00000000..de7a03c4
--- /dev/null
+++ b/pkg/system/doc/mkdir.hlp
@@ -0,0 +1,34 @@
+.help mkdir May85 system
+.ih
+NAME
+mkdir -- make a new directory
+.ih
+USAGE
+mkdir newdir
+.ih
+PARAMETERS
+.ls newdir
+New directory or subdirectory to be made.
+.le
+.ih
+DESCRIPTION
+\fIMkdir\fR creates a new directory with the given name.
+\fINewdir\fR may be an IRAF virtual directory name (not a logical name)
+or a host directory name.
+.ih
+EXAMPLES
+
+1. Make a subdirectory named "sub1".
+
+	cl> mkdir sub1
+
+2. Make a subdirectory "sub2" below "sub1".  The subdirectory "sub1" must
+already exist.
+
+	cl> mkdir sub1/sub2
+
+3. Make a directory "blue" at the same level in the directory hierarchy as
+the current directory (".." is a synonym for the previous directory).
+
+	cl> mkdir ../blue
+.endhelp
diff --git a/pkg/system/doc/mkscript.hlp b/pkg/system/doc/mkscript.hlp
new file mode 100644
index 00000000..5aa33204
--- /dev/null
+++ b/pkg/system/doc/mkscript.hlp
@@ -0,0 +1,161 @@
+.help mkscript Nov85 system
+.ih
+NAME
+mkscript -- make a script for a command sequence to be run in batch
+.ih
+USAGE	
+mkscript script task submit
+.ih
+PARAMETERS
+.ls script
+Script file name.  Commands will be successively added to this file.
+.le
+.ls task   
+Task name of command to be added to the script.  If given on the command
+line then only commands for this task may be added to the script.
+If not given on the command line then the task will query for a task
+name for each new command.  Currently the task name must not be abbreviated.
+.le
+.ls submit
+Submit the completed script as a background job as the last act of the task?
+If not given on the command line the task will query before submitting the
+script.
+.le
+.ls append = no
+Append new commands to an existing script file?
+If no the file will be deleted first.   If \fIverify\fR = yes
+the user will be asked to confirm the deletion.
+.le
+.ls hidden = yes
+Include hidden parameters in each command?
+.le
+.ls verify = yes
+Verify each command, any file deletions, and the final script?
+.le
+.ls logfile = "script.log"
+Script log file name.  When the script is submitted as a background job
+by this task any command and error output is directed to this file.
+.le
+.ih
+DESCRIPTION
+A command script is created consisting of a number of commands to be
+executed sequentially.  The task assumes the responsibility of formatting
+the command and placing it in the script file.  The user sets the
+parameter values using the parameter editor \fBeparam\fR.  As an optional
+final stage the task will optionally submit the script as a background job.
+
+The sequence of steps are outline as follows:
+.ls (1)
+If the script already exists and \fIappend\fR = no the script file
+is deleted.  When \fIverify\fR = yes the deletion is verified with the
+user.
+.le
+.ls (2)
+If the task is not specified on the command line then the user
+is queried for a task name.
+.ls (2a)
+The task must be loaded.  If it has not been loaded a message is printed
+and the task query is repeated.
+.le
+.le
+.ls (3)
+\fBEparam\fR is now invoked to allow the user to set the task
+parameters.
+.le
+.ls (4)
+If \fIverify\fR = yes the command is printed and the user is asked if the
+command is ok.  If ok the command is added to the script.
+.le
+.ls (5)
+The user is asked if another command is to be added to the script.  While
+the response is yes steps 2 to 5 are repeated.
+.le
+.ls (6)
+If \fIverify\fR = yes the script is paged and the user is asked if the
+script is ok.  If not ok the script is deleted, with user confirmation,
+and steps 2 to 6 are repeated.
+.le
+.ls (7)
+If the submit parameter is not specified on the command line the user
+is asked if the script should be submitted as a background job.
+.le
+
+The parameter \fIhidden\fR is important for the following reason.  If
+the hidden parameters are not explicitly included in the script commands
+then the values of the hidden parameters will be those in the parameter
+file at the time of execution.  Thus, in changes in the hidden parameters
+with \fBeparam\fR or explicit changes may produce unexpected results.
+However, if the hidden parameters are never changed then the commands
+are more readable when the hidden parameters are not included.
+.ih
+EXAMPLES
+One of the most common usages in data reductions is to create repeated
+commands with different input data or parameters.
+
+.nf
+cl> mkscript script.cl transform
+
+[\fIeparam\fR is called to set the parameter values for \fItransform\fR]
+
+transform ("n1r.008", "n1r.008a", "disp012,distort,disp013",
+database="identify.db", interptype="spline3", x1=1., x2=256., dx=1.,
+nx=256., xlog=no, y1=4300., y2=6300., dy=INDEF, ny=800., ylog=no,
+flux=yes, logfiles="STDOUT,logfile")
+
+Is the command ok? (yes):
+Add another command? (yes):
+
+[\fIeparam\fR is called again for task \fItransform\fR]
+
+transform ("n1r.010", "n1r.010a", "disp013,distort",
+database="identify.db", interptype="spline3", x1=1., x2=256., dx=1.,
+nx=256., xlog=no, y1=4300., y2=6300., dy=INDEF, ny=800., ylog=no,
+flux=yes, logfiles="STDOUT,logfile")
+
+Is the command ok? (yes):
+Add another command? (yes): no
+
+[The script is paged]
+
+Is the script ok? (yes):
+Submit the script as a background job? (yes):
+Script script.cl submitted at:
+Fri 10:32:57 01-Nov-85
+[1]
+.fi
+
+To combine several tasks:
+
+.nf
+cl> mkscript script.cl ver- sub- hid-
+Task name of command to be added to script: response
+
+[\fIeparam\fR is called for \fIresponse\fR and parameter values are set]
+
+Add another command? (yes):
+Task name of command to be added to script: imarith
+...
+Add another command? (yes): no
+.fi
+
+To run the command script as a foreground job:
+
+cl> cl < script.cl
+
+To run the command script as a background job:
+
+cl> cl < script.cl >& logfile &
+
+Note that the output including possible error output is redirected to a logfile.
+.ih
+BUGS
+The current implementation is preliminary.  It is done with a script which
+makes it seem somewhat slow.  The most important bug is that the command
+formatter is based on the output of \fBlparam\fR.  If a task parameter
+name exceeds 12 characters it is truncated by \fBlparam\fR and is then
+also truncated by the command formatter.  The script will then fail when
+executed!  Also the task name may not be abbreviated.
+.ih
+SEE ALSO
+eparam
+.endhelp
diff --git a/pkg/system/doc/movefiles.hlp b/pkg/system/doc/movefiles.hlp
new file mode 100644
index 00000000..73374474
--- /dev/null
+++ b/pkg/system/doc/movefiles.hlp
@@ -0,0 +1,38 @@
+.help movefiles Nov84 system
+.ih
+NAME
+movefiles -- move files to a directory
+.ih
+USAGE
+movefiles files directory
+.ih
+PARAMETERS
+.ls files
+A template specifying the file or files to be moved.
+.le
+.ls directory
+The directory to which the files are to be moved.
+.le
+.ls verbose = no
+If set to "yes", tell user as each file is moved.
+.le
+.ih
+DESCRIPTION
+This routine moves the specified files to the named directory.
+If a subdirectory and a logical directory both exist with the same
+name as the destination directory, the subdirectory is used.
+.ih
+EXAMPLES
+
+1. Move all files whose names start with `im' and end with `ab' to
+the directory `dir'.  Since "verbose" defaults to "no", do the work silently.
+
+	cl> movefiles im*ab dir
+
+2. Move all files in the current directory into the directory one level up.
+
+	cl> move * ..
+.ih
+SEE ALSO
+copy, rename
+.endhelp
diff --git a/pkg/system/doc/netstatus.hlp b/pkg/system/doc/netstatus.hlp
new file mode 100644
index 00000000..059f6819
--- /dev/null
+++ b/pkg/system/doc/netstatus.hlp
@@ -0,0 +1,44 @@
+.help netstatus Feb86 system
+.ih
+NAME
+netstatus -- print the status of the local network
+.ih
+USAGE
+netstatus
+.ih
+PARAMETERS
+None.
+.ih
+DESCRIPTION
+\fINetstatus\fR prints the status of the local network as perceived by the
+system process x_system.e (the network status may differ for each subprocess).
+The status output identifies the local node, lists all nodes in the local
+network, and lists all the aliases (recognized names) for each node.
+A message will be printed if networking is disabled on the local machine.
+The local network is defined by the table files "dev$hosts", "dev$uhosts",
+and "dev$hostlogin".
+.ih
+EXAMPLES
+
+.nf
+cl> netstatus
+Local node `draco' (5), default node `draco', 12 nodes in local network
+
+    NODE SERVER NREFS STATUS  ALIASES
+       1      0     0  00000  aquila vax1 a 1 class1 plot print
+       2      0     0  00000  lyra vax2 b 2 class2
+       3      0     0  00000  vela vax3 3 v class3
+       4      0     0  00000  carina vax5 c 5 class5
+       5      0     0  00000  draco vax6 6 d class6 0
+       6      0     0  00000  tucana sun1 t
+       7      0     0  00000  hydra sun2 h
+       8      0     0  00000  mensa pc1 m
+       9      0     0  00000  pictor pc2
+      10      0     0  00000  octans sun3 o
+      11      0     0  00000  pavo mvax1 p
+      12      0     0  00000  volans lsi1
+.fi
+
+.ih
+SEE ALSO
+.endhelp
diff --git a/pkg/system/doc/news.hlp b/pkg/system/doc/news.hlp
new file mode 100644
index 00000000..1b12f24c
--- /dev/null
+++ b/pkg/system/doc/news.hlp
@@ -0,0 +1,58 @@
+.help news Mar90 system
+.ih
+NAME
+news -- print the revisions summary for the current IRAF version
+.ih
+USAGE
+news
+.ih
+DESCRIPTION
+The \fInews\fR task uses the standard IRAF file pager to review a formatted
+summary of the system revisions for the version of IRAF being run.
+The revisions summaries for older versions of the system are also provided:
+use the \fIN\fR and \fIP\fR pager keys to display the next or previous
+system revisions summary.  The revisions summary is given in the file
+"doc$newsfile".
+
+For reasons of brevity, only the revisions summary is printed.  For detailed
+information on the revisions made to a particular science package, type
+
+    cl> help <pkg>.revisions op=sys
+
+where "pkg" is the name of the CL package for which revisions information
+is desired.  For detailed information on the revisions to the system
+software and programming interfaces, examine the system notes file,
+given in the file "notes.*" in the directory "iraf$local".  The system
+notes files for older versions of the system will be found in the "doc"
+directory.
+.ih
+BUGS
+The revisions summary is often lengthy and may be easier to read if a
+printed copy is made.
+
+Redirecting the output of \fInews\fR, e.g., to \fIlprint\fR, doesn't work
+at present.
+.ih
+EXAMPLES
+1. Page the revisions summary for the current IRAF release.
+
+    cl> news
+
+2. Print the revisions summary.
+
+    cl> lprint doc$newsfile
+
+3. Page the system notes file.  Anyone who develops software for IRAF
+should review this file with each new release, to see what has changed.
+Documentation for new system facilities is often given in the system
+notesfile.
+
+    cl> page iraf$local/notes.*
+
+4. Review the revisions summary for the IMAGES package.
+
+    cl> phelp images.revisions op=sys
+.ih
+SEE ALSO
+help, phelp, page
+.endhelp
diff --git a/pkg/system/doc/page.hlp b/pkg/system/doc/page.hlp
new file mode 100644
index 00000000..92cfd105
--- /dev/null
+++ b/pkg/system/doc/page.hlp
@@ -0,0 +1,146 @@
+.help page Nov86 system
+.ih
+NAME
+page -- display a file or files one page at a time
+.ih
+USAGE
+page files
+.ih
+PARAMETERS
+.ls files
+The list of files.  If omitted, text is read from the standard input.
+.le
+.ls map_cc = yes
+Map non-printing control characters into printable form, e.g., BEL
+(ctrl/G, ASCII 7) is printed as "^G".
+.le
+.ls clear_screen = no
+If set, the screen is cleared before each page is written.
+.le
+.ls first_page = 1
+The first page to be printed.  Pages are defined by form feed characters
+embedded in the text.
+.le
+.ls prompt = ""
+Optional prompt string for the end-of-page prompt.  If no prompt string is
+given the name of the file being paged is used.
+.le
+.ls device = "terminal"
+The output device.  The special device \fItext\fR may be specified to
+represent standout mode with upper case rather than reverse video characters.
+.le
+.ih
+DESCRIPTION
+\fIPage\fR displays a file on the terminal, one page of text at a time,
+pausing between successive pages of output until a key is typed on the
+terminal.  Pages are normally broken when the screen fills, but may also
+be delimited by form feed characters embedded in the input text.
+A prompt is printed after each page of text naming the current file,
+showing the percentage of the file which has been displayed, and the numeric
+order of the file within the file list if a file template was given.
+
+When the end of page prompt is displayed any of the following command
+keystrokes may be entered.  Command keystrokes are input in raw mode,
+i.e., no carriage return is required to pass the command to the program.
+
+.ks
+.nf
+	.		go to the beginning of the current file    [BOF]
+	:		colon escape (see below)
+	?		display a one-line command summary
+	G		go to the end of the current file          [EOF]
+	N,<ctrl/n>	go to the next file
+	P,<ctrl/p>	go back to the previous file
+	b		back up one page
+	d		scroll down half a page of text
+	e		edit the current file
+	f,space		advance to the next page
+	j,return	scroll down one line
+	k		back up one line
+	n		search for the next occurrence of a pattern
+	q		quit
+	u		back up half a screen
+
+	<ctrl/c>	quit (interrupt)
+	<ctrl/z>	quit (EOF)
+	<ctrl/d>	quit (EOF)
+.fi
+.ke
+
+If an unrecognized keystroke is entered the terminal will beep.  The following
+colon commands are recognized in addition to the single keystroke commands
+described above.
+
+.ks
+.nf
+	:!<clcmd>	send a command to the CL (:!! for host command)
+	:/<pattern>	advance to line matching the given pattern
+	:file <fname>	display file "fname" (may be abbreviated)
+	:help		print summary of colon commands
+	:line [+/-]N	goto line N (relative move if +/- given)
+	:spool <fname>	spool output to the named file
+.fi
+.ke
+
+The \fI:clcmd\fR facility is used to send commands to the CL from within
+the context of the pager.  For example, ":!cl" will temporarily suspend the
+pager, allowing CL commands to be entered until the command "bye" is entered,
+causing execution of the pager to resume.  Note that since the \fIpage\fR
+task resides in the system process \fIx_system.e\fR, it will be necessary
+for the CL to connect a second system process if the command issued calls
+another task in the system package, since the first system process will
+still be running, i.e., executing the \fIpage\fR task.  This is harmless,
+but the second process may be removed from the process cache with
+\fIflprcache\fR if desired, after exiting the original \fIpage\fR task.
+
+The sequence ":/" followed by a pattern will cause the current input stream
+to be searched for the next occurrence of the pattern given.  A pattern once
+entered is retained indefinitely and may be used in subsequent searches by
+typing the single keystroke `n', without need to reenter the pattern.
+Searching stops at the end of the current file, requiring a `.' to wrap back
+around to the beginning of the file, or a `N' to advance to the next file.
+
+The \fI:file\fR command is used to change the current position within the
+file list specified by \fIfiles\fR, and may not be used to page a file not
+specified in the initial template.  Note that the filename may be abbreviated,
+and that searching stops with the first file lexically greater than or equal
+to the given string (hence ":file x" might return file "y").
+
+The \fI:line N\fR command may be used to randomly position to the indicated line
+within the current file.  If the line number argument N is preceded by a plus
+or minus the argument is taken to be an offset from the current position.
+
+The \fI:spool\fR command is used to spool output to a file.  Each time a
+file line is printed on the screen, it is appended to the named file as well.
+One can interactively position to the desired line of the file and then turn
+on spooling to extract a portion of the file or stream being displayed.
+A subsequent \fI:spool\fR command with no filename will turn spooling off.
+Issuing a \fI:spool\fR to begin spooling on a new file when already spooling
+to some other file will cause the old spool file to be closed.
+.ih
+EXAMPLES
+
+1. Page through all of the files in the directory "lib" which have
+the extension ".h".
+
+	cl> page lib$*.h
+
+2. Use \fIhelp\fR to format the text in the file "doc$spp.hlp", displaying
+the formatted document beginning on page 5 (the entire document has to be
+formatted first so it takes a minute or so to get any output).
+
+	cl> help doc$spp.hlp fi+ | page first=5
+
+3. Run \fIrfits\fR to print a long format listing of the headers of a series
+of FITS images from a magnetic tape, directing the output through \fIpage\fR
+so that it does not flash by when you aren't looking.
+
+	cl> rfits mta make- long+ | page
+.ih
+BUGS
+Since \fIpage\fR does not currently buffer any input text, backwards motions
+and absolute line positioning are not permitted when paging the standard input.
+.ih
+SEE ALSO
+type, match, head, tail
+.endhelp
diff --git a/pkg/system/doc/pathnames.hlp b/pkg/system/doc/pathnames.hlp
new file mode 100644
index 00000000..3719fe28
--- /dev/null
+++ b/pkg/system/doc/pathnames.hlp
@@ -0,0 +1,42 @@
+.help pathnames Nov84 system
+.ih
+NAME
+pathnames -- print the host pathnames for a set of files
+.ih
+USAGE
+pathnames [template]
+.ih
+PARAMETERS
+.ls template
+A filename template specifying the set of files for which pathnames
+are desired.  If omitted, the pathname of the current working or default
+directory is printed.  The list of file names can come from the standard input.
+.le
+.ls sort = yes
+Sort the output in ASCII collating sequence.
+.le
+.ih
+DESCRIPTION
+\fIPathnames\fR converts a list of IRAF virtual file names into their host
+system equivalents.  When called with no arguments, the function of
+\fIpathnames\fR is to print the current default directory.
+.ih
+EXAMPLES
+
+1. Print the pathname of the current default directory (sample output for
+a VMS host system).
+
+.nf
+	cl> path
+	draco!DRB1:[IRAF.SYS.FIO]
+.fi
+
+2. Translate the file "vfiles", containing a list of virtual filenames, into
+the equivalent list of host system filenames, e.g., for use as input to a
+foreign task.
+
+	cl> path @vfiles > hfiles
+.ih
+SEE ALSO
+directory, files
+.endhelp
diff --git a/pkg/system/doc/phelp.hlp b/pkg/system/doc/phelp.hlp
new file mode 100644
index 00000000..2230eb29
--- /dev/null
+++ b/pkg/system/doc/phelp.hlp
@@ -0,0 +1,61 @@
+.help phelp Mar90 system
+.ih
+NAME
+phelp -- page the output of the HELP task
+.ih
+USAGE
+phelp template
+.ih
+PARAMETERS
+The \fIphelp\fR parameters are the same as for \fIhelp\fR except that
+the \fIpage\fR and \fInlpp\fR parameters are omitted.
+.ih
+DESCRIPTION
+The \fIphelp\fR task is a front end to \fIhelp\fR which spools the output
+of \fIhelp\fR in a scratch file, then calls the file pager \fIpage\fR to
+view the output text.  The advantage is that while \fIhelp\fR pages its
+output, one can only move forward through the output text.  By using
+\fIphelp\fR it is possible to randomly scan the spooled help text, e.g.,
+skipping forward to get a quick overview and then backing up to read the
+details more carefully.  This capability is especially useful when viewing
+large multipage help pages, or when viewing a number of related help pages
+all at once.
+.ih
+EXAMPLES
+1. Page the help page for the \fImkpkg\fR task.
+
+    cl> phelp mkpkg
+
+2. View the help pages for all the tasks in the IMAGES package.
+
+    cl> phelp images.*
+
+When viewing multiple help pages as in this last example, note that the
+'N' and 'P' keystrokes in the pager may be used to move to the next or
+previous help page.  "." will return to the first help page (the start
+of the spooled help text) and 'G' will skip to the end of file.  Type '?'
+while in the pager to get a summary of the most often used keystrokes.
+
+3. Format and page the Lroff (IRAF HELP) format document "MWCS.hlp" in
+the system directory "mwcs".
+
+.nf
+    cl> cd mwcs
+    cl> phelp MWCS.hlp fi+
+.fi
+
+In this case the text being viewed is not part of the on-line help system,
+but is a technical document describing one of the IRAF programming interfaces.
+Any .hlp file may be viewed in this way.
+.ih
+TIME REQUIREMENTS
+\fIphelp\fR is not quite as fast as \fIhelp\fR since it must fully format
+the help text into a temporary file before the file can be viewed.  For
+small help pages, or to view only the first few screens of a help page,
+the \fIhelp\fR task will be faster.
+.ih
+BUGS
+.ih
+SEE ALSO
+page, help, references
+.endhelp
diff --git a/pkg/system/doc/protect.hlp b/pkg/system/doc/protect.hlp
new file mode 100644
index 00000000..6a0049e4
--- /dev/null
+++ b/pkg/system/doc/protect.hlp
@@ -0,0 +1,36 @@
+.help protect Nov84 system
+.ih
+NAME
+protect -- protect files from deletion
+.ih
+USAGE
+protect files
+.ih
+PARAMETERS
+.ls files
+A template specifying the file or files to be protected.
+.le
+.ih
+DESCRIPTION
+\fIProtect\fR asserts protection from deletion for the specified files.
+A protected file can be deleted only by first "unprotecting" it.
+File protection is preserved when a file is copied or renamed,
+even when copied or renamed to a remote network node,
+but may be lost when a file is backed up on tape and later restored
+(depending upon what utility one uses).  Note that imagefiles are
+automatically protected to prevent accidental deletion of the header
+file, leaving a "zombie" pixel file somewhere on disk.
+.ih
+EXAMPLES
+
+1. Protect the file "paper.ms" from deletion, accidental or otherwise.
+
+	cl> protect paper.ms
+
+2. Protect all the ".ms" files from deletion.
+
+	cl> protect *.ms
+.ih
+SEE ALSO
+unprotect, delete
+.endhelp
diff --git a/pkg/system/doc/references.hlp b/pkg/system/doc/references.hlp
new file mode 100644
index 00000000..a67b4182
--- /dev/null
+++ b/pkg/system/doc/references.hlp
@@ -0,0 +1,78 @@
+.help references Jun89 system
+.ih
+NAME
+references -- find all help database references to a given topic
+.ih
+USAGE
+references topic
+.ih
+PARAMETERS
+.ls topic
+The topic for which help is desired, i.e., a keyword, phrase, or pattern
+which the help database or quick-reference file is to be searched for.
+.le
+.ls quickref = "uparm$quick.ref"
+The name of the optional quick-reference file.
+.le
+.ls updquick = no
+Create or update the quick-reference file, e.g., because new packages
+have been added to the global help database.  Updating the quick-reference
+file automatically enables \fIusequick\fR, discussed below.
+.le
+.ls usequick = no
+Use the quick-reference file.  By default, a runtime search of all the package
+menus in the full help database is performed, which ensures that all packages
+are searched, but which is comparatively slow.
+.le
+.ih
+DESCRIPTION
+The \fIreferences\fR task is used to search the help database for all tasks
+or other help modules pertaining to the given topic, where \fItopic\fR may be
+a keyword, phrase, or any general pattern as would be input to the \fImatch\fR
+task.  By default the full help database will be searched.  If desired,
+the "one-liner" information used for topic searching may be extracted and
+used to prepare a quick-reference file to speed further searches.
+This is not done by default because the help database is too dynamic, e.g., 
+new external packages may be installed at any time, by any user, or new
+tasks may be added to existing packages at any time.
+
+References to tasks (or other objects) are printed in the form
+
+.nf
+       keyword1 - one line description
+       keyword2 - one line description
+.fi
+
+and so on.  To determine the \fIpackage pathname\fR of each named task,
+get \fIhelp\fR on the named \fIkeyword\fR and the pathname will be seen at
+the top of the help screen, followed by additional information about the
+referenced object.  If there are multiple objects with the same name,
+a "help <keyword> \fIall+\fR" may be required to locate a particular one.
+.ih
+EXAMPLES
+1. Find all help on CCDs.
+
+	cl> ref ccd
+
+2. Create or update your private quick-reference file.
+
+	cl> ref upd+
+
+3. Examine the quick-reference file to get a summary of all the tasks
+or other help modules in the help database.
+
+	cl> page (ref.quickref)
+.ih
+TIME REQUIREMENTS
+If a quick-reference file is used searching takes seconds, otherwise it
+might take a minute or so for the typical large help database containing
+all help modules for the core system and several external, layered packages.
+.ih
+BUGS
+Only the one-liner (NAME) field describing each help module is used for
+the searches.  With a little work searching could be made much more
+comprehensive as well as faster.
+.ih
+SEE ALSO
+help, match
+.endhelp
diff --git a/pkg/system/doc/rename.hlp b/pkg/system/doc/rename.hlp
new file mode 100644
index 00000000..226045da
--- /dev/null
+++ b/pkg/system/doc/rename.hlp
@@ -0,0 +1,69 @@
+.help rename Aug97 system
+.ih
+NAME
+rename -- rename a file or set of files
+.ih
+USAGE
+rename file newname
+.ih
+PARAMETERS
+.ls file
+A template specifying the file or files to be renamed.
+.le
+.ls newname
+If a single file is being renamed, the new filename, else the new name of
+the field being renamed in a set of filenames.  If \fInewname\fR is a
+directory the input files will be moved to this directory with the same
+name.
+.le
+.ls field = all
+If set to "all" the file name remains unchanged and the \fInewname\fR is
+assumed to be a destination directory in the case of multiple input files,
+or the new filename (which may contain a new directory path) in the case of
+a single input file.  If set to \fIldir\fR the \fInewname\fR value is taken
+to be a destination directory and the file is moved to this directory.
+Setting to \fIroot\fR will rename only the root portion of the filename,
+a value of \fIextn\fR will change or append the extension. \fInewname\fR
+cannot contain a directory path when changing the root or extn field.
+.le
+.ih
+DESCRIPTION
+\fIRename\fR renames either a single file to "newname", or a set of files,
+changing either the ldir, root or the extension part of each name.  
+If \fInewname\fR is a directory or  \fIfield\fR is "ldir" the input files
+are moved to this directory and the filenames remain the same.  When
+modifying the root or extension part of the filename \fInewname\fR is the
+new root or extension name for the input files, an extension will be added
+to the file name if it doesn't already exist and the extension field is being
+modified.  For multiple input files it is assumed
+that \fInewname\fR is a directory if the value of \fIfield\fR is "all", 
+otherwise an error is generated to prevent overwriting files.
+.ih
+EXAMPLES
+
+1. Rename file "fred" to "jay".
+
+    cl> rename fred jay
+
+2. Change the root name of a set of files from "out" to "pkout".
+
+    cl> rename out.x,out.o,out.par pkout field=root
+
+3. Change the extension of all ".f77" files from ".f77" to ".f".
+
+    cl> rename *.f77 f field=extn
+
+4. Move all files with a ".dat" extension to a new directory.
+
+    cl> rename *.dat data$
+    cl> rename *.dat /data/user
+
+5. Add a ".fits" extension to all files in a directory.
+
+    cl> rename im00* fits field=extn
+.ih
+BUGS
+.ih
+SEE ALSO
+movefiles, copy
+.endhelp
diff --git a/pkg/system/doc/rewind.hlp b/pkg/system/doc/rewind.hlp
new file mode 100644
index 00000000..0519f81c
--- /dev/null
+++ b/pkg/system/doc/rewind.hlp
@@ -0,0 +1,36 @@
+.help rewind Apr92 system
+.ih
+NAME
+rewind -- rewind a previously allocated device
+.ih
+USAGE
+rewind device
+.ih
+PARAMETERS
+.ls device
+The device to be rewound.
+.le
+.ls initcache = yes
+Initialize the magtape device position cache for the device.  This causes
+the magtape i/o system to "forget" what it thinks it knows about things
+like the number of files on the tape, the amount of tape used, and so on.
+.le
+.ih
+DESCRIPTION
+\fIRewind\fR rewinds the specified device, which is most likely
+a magnetic tape, and which has been previously allocated to the user.
+
+By default \fIrewind\fR will initialize the device position cache.  When
+changing tapes, one should always either rewind or deallocate and reallocate
+the device, to force the magtape system to recompute the number of files
+on the tape and to ensure that the tape is left in a defined position.
+.ih
+EXAMPLES
+
+1. Rewind logical tape drive a.
+
+	cl> rewind mta
+.ih
+SEE ALSO
+allocate, deallocate, devstatus
+.endhelp
diff --git a/pkg/system/doc/sort.hlp b/pkg/system/doc/sort.hlp
new file mode 100644
index 00000000..7d89657b
--- /dev/null
+++ b/pkg/system/doc/sort.hlp
@@ -0,0 +1,62 @@
+.help sort Mar87 system
+.ih
+NAME
+sort -- sort a file or the standard input
+.ih
+USAGE
+sort input_file
+.ih
+PARAMETERS
+.ls input_file
+The text file to be sorted.  If the standard input is redirected the standard
+input is sorted.
+.le
+.ls column = 0
+If 0, sort entire text lines, else sort based on data/text starting
+in the specified column.  Columns are delimited by whitespace.  Thus,
+.nf
+	12   abc   34   56
+.fi
+has four columns, "abc" being in the second.
+.le
+.ls ignore_whitespace = no
+Ignore leading whitespace.  Useful only when column = 0 and the sort is
+non-numeric.
+.le
+.ls numeric_sort = no
+If set, make numerical (rather than ASCII text) comparisons.
+.le
+.ls reverse_sort = no
+If set, sort in reverse text/numeric order.
+.le
+.ih
+DESCRIPTION
+\fISort\fR sorts the contents of the given text file, or the
+standard input, either on a textual (based on the ASCII collating
+sequence), or on a numeric basis.  If a numeric sort is requested,
+and either field in any comparison is non-numeric, a string (ASCII)
+comparison will be made instead.
+.ih
+EXAMPLES
+
+1. Sort the output of the set command into alphabetic (ASCII collating)
+order.
+
+	cl> set | sort
+
+2. Sort the contents of "file", in reverse ASCII order, ignoring the
+contents of columns 1 through 4.
+
+	cl> sort file rev+ col=5
+
+3. Print a long form directory listing with the files sorted by size,
+largest files first.
+
+	cl> dir | sort num+ rev+ col=3
+.ih
+BUGS
+Only one file can be sorted per call, and only one column or all columns can
+be used for the sort.  The current program is inefficient for large numeric
+sorting tasks because the same numeric field may be decoded into its
+corresponding binary value many times.
+.endhelp
diff --git a/pkg/system/doc/spy.hlp b/pkg/system/doc/spy.hlp
new file mode 100644
index 00000000..33ffa52c
--- /dev/null
+++ b/pkg/system/doc/spy.hlp
@@ -0,0 +1,26 @@
+.help spy Feb85 system
+.ih
+NAME
+spy -- tell who is logged in and what they are doing
+.ih
+USAGE
+spy [v]
+.ih
+DESCRIPTION
+\fISpy\fR prints a summary of who is on the system and what they are doing.
+The optional argument \fIv\fR (short for \fIverbose\fR) causes more detailed
+information to be given.  The operation of this task is machine dependent,
+as is the quantity and format of the information returned.
+.ih
+EXAMPLES
+.nf
+cl> spy
+  4:36pm  up 24 days,  5:42,  2 users,  load average: 0.29, 0.15, 0.18
+User     tty       login@  idle   JCPU   PCPU  what
+roger    ttyh8     4:21pm    15      8      5  -csh 
+alice    ttyh9     4:26pm           44     27  w 
+.fi
+.ih
+SEE ALSO
+diskspace
+.endhelp
diff --git a/pkg/system/doc/tail.hlp b/pkg/system/doc/tail.hlp
new file mode 100644
index 00000000..bd91e617
--- /dev/null
+++ b/pkg/system/doc/tail.hlp
@@ -0,0 +1,50 @@
+.help tail Nov84 system
+.ih
+NAME
+tail -- print the last few lines of the specified files
+.ih
+USAGE
+tail files
+.ih
+PARAMETERS
+.ls files
+A template specifying the files to be listed.
+.le
+.ls nlines = 12
+The number of lines to be printed.  If negative, the number
+of lines to be skipped, counting from the beginning of the file.
+.le
+.ih
+DESCRIPTION
+For each file in the input file list, \fItail\fR copies the last \fInlines\fR
+of the file to the standard output.  If there is more than one file in the
+input file list, as one line header is printed before each file.
+If "nlines" is negative, then abs(nlines) lines are skipped, and the rest
+of the file is printed, i.e., the tail of the file is still printed, but
+the offset is measured from the beginning of the file rather than the end.
+.ih
+EXAMPLES
+
+1. Prints the last 12 lines of each help file in the current directory.
+
+	cl> tail *.hlp
+
+2. Print the last line of each help file.
+
+	cl> tail *.hlp nl=1
+
+3. Prints the third through fifth lines of "file".  The same thing
+might be done (at least conceptually) by "head file,nlines=5"
+piped into "tail ,nlines=3".  However, \fItail\fR does not work on STDIN.
+
+.nf
+	cl> tail file nl=-2 | head nl=3
+.fi
+.ih
+BUGS
+\fITail\fR does not presently work on standard input, and therefore cannot
+be used in pipes.
+.ih
+SEE ALSO
+head
+.endhelp
diff --git a/pkg/system/doc/tee.hlp b/pkg/system/doc/tee.hlp
new file mode 100644
index 00000000..466b3751
--- /dev/null
+++ b/pkg/system/doc/tee.hlp
@@ -0,0 +1,36 @@
+.help tee Nov84 system
+.ih
+NAME
+tee -- tee the standard output to a file
+.ih
+USAGE
+tee file
+.ih
+PARAMETERS
+.ls file
+The name of the output file.
+.le
+.ls out_type = "text"
+The type of output file to be created, either "text" or "binary".
+.le
+.ls append = no
+If set, append to an existing file, otherwise create a new file.
+.le
+.ih
+DESCRIPTION
+\fITee\fR copies its input to both the standard output and the named file.
+Its primary use is in pipes where one wants to capture some intermediate output.
+.ih
+EXAMPLES
+
+1. The results of the \fIset\fR command are captured in the file "temp",
+and are also passed on to the "match" command.  The result is
+a "temp" file of perhaps 100 lines, with the output to the screen
+only around 5 lines.
+
+	cl> set | tee temp | match tty
+.ih
+BUGS
+Since the processes in an IRAF pipe execute serially rather than concurrently,
+the teed output will not appear until all tasks to the left have completed.
+.endhelp
diff --git a/pkg/system/doc/touch.hlp b/pkg/system/doc/touch.hlp
new file mode 100644
index 00000000..aedb35a0
--- /dev/null
+++ b/pkg/system/doc/touch.hlp
@@ -0,0 +1,71 @@
+.help touch Jan04 system
+.ih
+NAME
+touch -- change file access and modification times
+.ih
+USAGE
+touch files
+.ih
+PARAMETERS
+.ls files
+List of files to be created or touched.
+.le
+.ls create = yes
+If enabled, the file will be created as a zero-length text file if it doesn't
+already exist.
+.le
+.ls atime = yes
+Change the access time of the file.  Will not change the modification time
+unless \fImtime\fR is also set.
+.le
+.ls mtime = yes
+Change the modification time of the file.  Will not change the access time
+unless \fIatime\fR is also set.
+.le
+.ls time = ""
+Time and date to set for the file.  The format of this string may be any
+of DD/MM/YY or CCYY-MM-DD (in which case time is assumed to be midnight of
+that day), or CCYY-MM-DDTHH:MM:SS[.SSS...] to specify both date and time.
+If not specified, the current system time is used unless the \fIref_file\fR
+parameter is set.  If specified, \fIref_file\fR will be ignored.
+.le
+.ls ref_file = ""
+Use the corresponding times of the specified file for modifying the
+times of the \fIinput_files\fR.  If not specified, the current time is
+used unless the \fItime\fR parameter is set.
+.le
+.ls verbose = no
+Print verbose output of the files and times being reset.
+.le
+.ih
+DESCRIPTION
+The \fItouch\fR task sets the access and modification times of each file
+in the \fIfiles\fR list.  The file will be created if it does not already
+exist when the \fIcreate\fR parameter is set.  The time used can be
+specified by \fItime\fR parameter or by the corresponding fields of the
+file specified by \fIref_file\fR, otherwise the current system time will
+be used.  The task will update both the modification and access times of
+the file unless disabled by the \fIatime\fR or \fImtime\fR parameter.
+
+.ih
+EXAMPLES
+
+1.  Update the times of all SPP source files in the current directory:
+
+	cl> touch *.x
+
+2.  Create an empty file on a remode node:
+
+	cl> touch ursa!/data/trigger_file
+
+3.  Reset the file modification time to 2:33:45 pm on June 5, 2003:
+
+	cl> touch nite1.fits time="2003-06-05T14:23:45"
+
+4.  Reset the file modification time to match dev$hosts:
+
+	cl> touch nite1.fits ref_file=dev$hosts
+
+.ih
+SEE ALSO
+.endhelp
diff --git a/pkg/system/doc/type.hlp b/pkg/system/doc/type.hlp
new file mode 100644
index 00000000..f8af91cd
--- /dev/null
+++ b/pkg/system/doc/type.hlp
@@ -0,0 +1,43 @@
+.help type Nov84 system
+.ih
+NAME
+type -- type a file or files on the standard output
+.ih
+USAGE
+type input_files
+.ih
+PARAMETERS
+.ls input_files
+A template specifying the file or files to be typed.
+.le
+.ls map_cc = yes
+If set, output any non-printing control characters in the input text in
+a printable form, e.g., ctrl/c (ASCII 3) would be output as ^C.
+.le
+.ls device = "terminal"
+The output device, defaulting to the user's terminal.  If the special device
+"text" is named, any standout mode control characters embedded in the text
+will cause the enclosed text to be output in upper case.
+.le
+.ih
+DESCRIPTION
+\fIType\fR copies the named files (or the files selected by
+the file template) to the standard output.
+If there is more than one file in the input list, a header naming the file
+to be printed will precede each output file.
+.ih
+EXAMPLES
+
+1. Type all files in the current directory with the extension ".x" on the
+standard output.  Do not pause between files or pages (unlike \fIpage\fR).
+
+	cl> type *.x
+
+2. Capture the manual page for task \fIhedit\fR in a text file, in a form
+suitable for printing on any device.
+
+	cl> help hedit | type dev=text > hedit.doc
+.ih
+SEE ALSO
+page, head, tail, concatenate, lprint
+.endhelp
diff --git a/pkg/system/doc/unprotect.hlp b/pkg/system/doc/unprotect.hlp
new file mode 100644
index 00000000..99120c07
--- /dev/null
+++ b/pkg/system/doc/unprotect.hlp
@@ -0,0 +1,27 @@
+.help unprotect Nov84 system
+.ih
+NAME
+unprotect -- remove file protection
+.ih
+USAGE
+unprotect files
+.ih
+PARAMETERS
+.ls files
+A template specifying the file or files from which delete protection is
+to be removed.
+.le
+.ih
+DESCRIPTION
+\fIUnprotect\fR removes delete protection from the named files.
+A "protected" file cannot be deleted or clobbered.
+.ih
+EXAMPLES
+
+1. Remove file protection from the listed files.
+
+	cl> unprotect alpha,beta,gamma.x,*.db
+.ih
+SEE ALSO
+protect, delete
+.endhelp
diff --git a/pkg/system/doc/urlget.hlp b/pkg/system/doc/urlget.hlp
new file mode 100644
index 00000000..da6ede83
--- /dev/null
+++ b/pkg/system/doc/urlget.hlp
@@ -0,0 +1,84 @@
+.help urlget Sep2011 system
+.ih
+NAME
+urlget -- Get a (http) URL to the named file
+.ih
+USAGE
+urlget url fname
+.ih
+PARAMETERS
+.ls url
+The URL to download.
+.le
+.ls fname
+The name of the file to create containing the URL contents.  If not
+specified, the trailing part of the URL is used as the filename.
+.le
+.ls use_cache = yes
+Use the system file cache?  If 'yes' and the file already exists in the
+cache, the cached file will be copied to the output filename.
+If 'no' then the URL will be downloaded again.
+.le
+.ls extn = ""
+Optional filename extension to put on the cached filename.  This can be
+used to force files to be saved as a particular type in the cache.
+.le
+.ls verbose = no
+Print verbose output?
+.le
+.ls cache = "cache$"
+Logical cache directory name.
+.le
+
+.ih
+DESCRIPTION
+The \fIURLGET\fR task is used to download a URL (HTTP protocol only) to a 
+local file named by the \fIfname\fR parameter.  If no \fIfname\fR is given, 
+a filename is constructed from the last part of the URL (i.e.
+characters trailing any of the '?', '/', or '&' delimiters).  Because 
+the URL may not point to a static file, use of the \fIfname\fR parameter
+is recommended.
+
+If the \fIuse_cache\fR parameter is set, the URL will only be downloaded if
+it does not already exist in the file cache pointed to by the \fIcache\fR
+parameter, otherwise the cached file will be copied to the output filename.
+The \fIextn\fR parameter may be to used to force a specific file extension
+to be appended to the filename in the cache, this allows a URL to be passed
+to a task and treated as if it were a file of a specific type.
+
+.ih
+EXAMPLES
+
+1. Download a FITS image from a URL (these are equivalent):
+
+.nf
+    cl> urlget http://iraf.noao.edu/foo.fits 
+    cl> urlget http://iraf.noao.edu/foo.fits foo.fits
+.fi
+
+2. Force a URL to be downloaded again:
+.nf
+    cl> urlget http://iraf.noao.edu/foo.fits use_cache=no
+.fi
+
+3. Download a URL with special characters:
+.nf
+    cl> urlget http://iraf.noao.edu/scripts/tget?f=foo.fits
+or
+    cl> s1 = "http://iraf.noao.edu/scripts/tget?f=foo.fits"
+    cl> urlget(s1)
+or
+    cl> s1 = "http://iraf.noao.edu/scripts/tget?f=foo.fits&d=/iraf/web"
+    cl> urlget(s1,"foo.fits",verbose+)
+.fi
+
+Escaping special characters isn't required from the commandline since the
+URL is assumed to be whitespace or comma delimited.
+
+.ih
+BUGS
+
+.ih
+SEE ALSO
+
+.endhelp
diff --git a/pkg/system/fcache.par b/pkg/system/fcache.par
new file mode 100644
index 00000000..0c69ba6f
--- /dev/null
+++ b/pkg/system/fcache.par
@@ -0,0 +1,9 @@
+cmd,s,a,,,,cache command
+pattern,s,h,"*",,,filename pattern
+src,s,h,"",,,file or url to cache
+fname,s,h,"",,,cache filename
+extn,s,h,"",,,cache filename extension
+age,i,h,-1,,,file age (in days) to purge
+verbose,b,h,no,,,print verbose output?
+wait,b,h,yes,,,block on operation?
+cache,s,h,"cache$",,,cache directory
diff --git a/pkg/system/files.par b/pkg/system/files.par
new file mode 100644
index 00000000..20e6a230
--- /dev/null
+++ b/pkg/system/files.par
@@ -0,0 +1,2 @@
+template,s,a,,,,file template
+sort,b,h,yes,,,sort file list
diff --git a/pkg/system/files.x b/pkg/system/files.x
new file mode 100644
index 00000000..9bd94f91
--- /dev/null
+++ b/pkg/system/files.x
@@ -0,0 +1,32 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# FILES -- Expand a file name template into a list of file names on the
+# standard output.
+
+procedure t_files()
+
+int	list
+pointer	sp, template, fname
+
+bool	clgetb()
+int	fntopnb(), fntgfnb(), clgeti(), btoi()
+
+begin
+	call smark (sp)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+	call salloc (template, SZ_LINE, TY_CHAR)
+
+	if (clgeti ("$nargs") > 0)
+	    call clgstr ("template", Memc[template], SZ_LINE)
+	else
+	    call strcpy ("*", Memc[template], SZ_LINE)
+
+	list = fntopnb (Memc[template], btoi(clgetb("sort")))
+	while (fntgfnb (list, Memc[fname], SZ_FNAME) != EOF) {
+	    call printf ("%s\n")
+		call pargstr (Memc[fname])
+	}
+
+	call fntclsb (list)
+	call sfree (sp)
+end
diff --git a/pkg/system/hdbexamine.par b/pkg/system/hdbexamine.par
new file mode 100644
index 00000000..9ef63b4a
--- /dev/null
+++ b/pkg/system/hdbexamine.par
@@ -0,0 +1,2 @@
+helpdb,s,h,"helpdb",,,"name of help database to be examined"
+verbose,b,h,no,,,"long form output"
diff --git a/pkg/system/head.par b/pkg/system/head.par
new file mode 100644
index 00000000..2379807e
--- /dev/null
+++ b/pkg/system/head.par
@@ -0,0 +1,2 @@
+input_files,s,a,,,,list of files to be printed
+nlines,i,h,12,0,,number of lines to be printed from each file
diff --git a/pkg/system/head.x b/pkg/system/head.x
new file mode 100644
index 00000000..71b6db2b
--- /dev/null
+++ b/pkg/system/head.x
@@ -0,0 +1,56 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+
+# HEAD -- Print the head (first few lines) of each of the named text files on
+# the standard output.  If more than one file is to be printed, a brief header
+# is printed for each file.
+# 
+# Params:
+# 	input_files		file matching template
+# 	nlines [h,8]		number of lines to be printed
+
+procedure t_head()
+
+char	fname[SZ_FNAME]
+bool	multiple_files
+int	nlines, list
+int	clpopni(), clplen(), clgfil(), clgeti()
+
+begin
+	list = clpopni ("input_files")
+	multiple_files = (clplen (list) > 1)
+	nlines = clgeti ("nlines")
+
+	while (clgfil (list, fname, SZ_FNAME) != EOF)
+	    iferr (call print_head (STDOUT, fname, nlines, multiple_files))
+		call erract (EA_WARN)
+
+	call clpcls (list)
+end
+
+
+# PRINT_HEAD -- Print the first few lines of a file on the output stream given
+# as the first argument, optionally plus a header.
+
+procedure print_head (out, fname, nlines, print_file_name)
+
+int	out, nlines
+bool	print_file_name
+char	fname[SZ_FNAME], lbuf[SZ_LINE]
+int	in, line, open(), getline()
+errchk	open, getline, putline
+
+begin
+	in = open (fname, READ_ONLY, TEXT_FILE)
+
+	if (print_file_name) {
+	    call fprintf (out, "\n\n===> %s <===\n")
+		call pargstr (fname)
+	}
+
+	for (line=1;  line <= nlines && getline(in,lbuf) != EOF;  line=line+1)
+	    call putline (out, lbuf)
+
+	call close (in)
+end
diff --git a/pkg/system/help.par b/pkg/system/help.par
new file mode 100644
index 00000000..ef6f2060
--- /dev/null
+++ b/pkg/system/help.par
@@ -0,0 +1,31 @@
+#  Parameter file for the HELP task
+
+template,s,a,"",,,"Module name or package.module style template"
+
+file_template,b,h,no,,,"Template is a filename matching template"
+option,s,h,"help",,,"Type of help (help,source,sysdoc,files,summary,etc)"
+device,s,h,"terminal","terminal|gui|text|html|postscript|ps",,"output device"
+
+# Output parameters.
+all,b,h,no,,,"print help for all modules matching template"
+parameter,s,h,"all",,,"parameter for which help is desired"
+section,s,h,"all",,,"section for which help is desired"
+
+# Terminal display parameters.
+page,b,h,yes,,,"paginate output"
+nlpp,i,h,59,1,,"number of lines per page if output redirected"
+lmargin,i,h,1,1,,"left margin"
+rmargin,i,h,72,2,,"right margin"
+
+# GUI parameters
+search,b,h,no,,,"Search database for help topic?"
+home,s,h,"",,,"Homepage to be used"
+printer,s,h,"printer",,,"Printer name"
+showtype,b,h,no,,,"Indicate packages in list?"
+quickref,s,h,"uparm$quick.ref",,,"Quick-reference file"
+uifname,s,h,"lib$scr/help.gui",,,"User interface filename"
+coords,*gcur,h,"",,,"Cursor"
+
+# Help Parameters.
+helpdb,s,h,"helpdb",,,"Help database to be used"
+curpack,s,h,"AskCL",,,"current package"
diff --git a/pkg/system/help/README b/pkg/system/help/README
new file mode 100644
index 00000000..a448779b
--- /dev/null
+++ b/pkg/system/help/README
@@ -0,0 +1,12 @@
+The HELP package -- online documentation utilties.  This directory contains
+the source and documtation for the HELP utility.  The subdirectory lroff
+contains the source for the Lroff text formatter, called by HELP to format
+help blocks.
+
+Jul84	Added the help database facility, i.e., pre-compilation of the help
+	directories to produce a binary database.  Numerous other improvements.
+	See the revs file for additional details.
+
+Note:	The TASK declaration for HELP is in pkg$system/system.cl.
+	The TASK declarations for MKHELPDB and HDBEXAMINE are in the softools
+	    package, file pkg$softools/softools.cl.
diff --git a/pkg/system/help/design.hlp b/pkg/system/help/design.hlp
new file mode 100644
index 00000000..9d236da6
--- /dev/null
+++ b/pkg/system/help/design.hlp
@@ -0,0 +1,500 @@
+.helpsys help Mar84 "Detailed Design of the Help Utility"
+.nh
+The Help Database
+
+    Help has to know where to find help files and sources for modules.
+This information is stored in two types of files.  The package list file,
+maintained in lib$ by the system manager, contains an entry for each
+installed package.  The module list file, maintained in the package
+directory by the programmer, contains the same information for each
+module in the package.  The format of a help list file is as follows:
+
+.nf
+	$defdir = pathname|ldir
+	$ldir1	= pathname|ldir
+	$ldir2	= pathname|ldir
+		    ...
+
+	name1	hlp=file, src=file, sys=file, pkg=file
+	name2	hlp=file, src=file, sys=file, pkg=file
+		    ...
+.fi
+
+Logical directories may be defined in the header area of the helpdir file
+as shown above.  These may be omitted if the package directory is already
+defined in the CL environment.  The "defdir" prefix will be added to any
+file name that does not contain an explicit logical or OS directory
+specification.  The default directory may be an OS pathname or a logical
+directory guaranteed to be defined in the CL environment at the time that
+\fBhelp\fR is run.  There should be no OS dependent file names in the
+module lists.
+
+Each module entry consists of a number of keyword equals value type
+fields.  The fields are optional and may be given in any order.  If the
+last character on the line is a comma, the list is assumed to be continued
+on the next line.  The \fBhlp\fR field is the name of the file containing the
+help block.  The \fBsrc\fR file contains the source for the named module.
+The \fBsys\fR file contains the system documentation.  The \fBpkg\fR field,
+present only in the package helpdir file, is the name of the module helpdir
+file for the package.  Although the help database is organized by
+package, the package list is linear, not a tree.
+
+If multiple entries are given for a module, the entry nearest the tail
+of the list is used.
+
+.nh
+Program Structure
+
+    If no args, print help block for the current package, otherwise expand
+the template into a list of package.module elements.  Process the help
+text for each module.  In general, the \fBlroff\fR text formatter is called
+to process all help text.  \fBLroff\fR is called with the names of two external
+procedures used to get lines of text from the input and put formatted text
+to the output.
+
+\fBLroff\fR is essentially a numerical procedure which effects a
+format conversion on a list of text lines.  Additional specialized processing
+is carried out in the input and output procedures.  Thus, when processing
+only the help text for a single section, the input procedure skips input
+lines until the desired section is detected, and returns EOF to \fBlroff\fR
+when the end of the input section is detected.  Similarly, the output
+procedure handles all the details of paginating the output on the user
+terminal.
+
+.nf
+t_help
+	clio to get params
+	salloc ctrl structure
+	ttyodes
+	get_option					[match option]
+	process_template
+		hd_open					[open helpdir]
+			open,getline,close
+			hd_putldiry
+				hd_getc
+				hd_putstr
+					realloc
+			hd_putmodule
+				hd_getc
+				hd_putstr
+					realloc
+				getline
+			hd_sort_modules
+		expand_template				[encode pat  ]
+			malloc
+			patmake
+		read_template				[get modname ]
+			hd_getname
+				hd_getldir
+			patmatch
+		print_help				[do a module ]
+			hd_findmod
+			hd_getname
+				hd_getldir
+			pr_filenames
+				hd_findmod
+				hd_getname
+					hd_getldir
+				output
+					nomore
+						clgetb
+					ttyputline
+					putline
+			pr_sourcefile
+				hd_findmod
+				hd_getname
+					hd_getldir
+				input
+					getline
+				output
+					...
+			pr_summary
+				hd_findmod
+				hd_getname
+					hd_getldir
+				getline
+				output
+					...
+			pr_helpblock			[default help]
+				find_help_block
+					getline
+					strmatch
+				pr_header
+					ttyclear
+				LROFF			[format text ]
+					input
+						getline
+					output
+						...
+		close_template
+			mfree
+		hd_close
+			mfree
+.fi
+
+.nh
+Data Structures
+
+    For simplicity, helpdir files are read into an internal buffer in
+one operation, and searching for modules, file name extraction, etc. is
+thereafter done on the buffered helpdir.  The helpdir is represented
+in memory by the following structures.
+
+.nf
+struct helpdir {
+	char	*hd_sbuf		# string buffer
+	int	hd_nextch		# index of next char in sbuf
+	int	hd_szsbuf		# size of string buffer
+	int	hd_defdir		# index of defdir str
+	int	hd_nldirs		# number of ldirs
+	int	hd_nmodules		# number of modules
+	int	hd_ldir[MAX_LDIR]	# indices of ldirs
+	struct	module hd_module[MAX_MODULES] # module names, files
+}
+
+
+struct module {
+	int	m_name			# index of module name
+	int	m_hlp			# index of help file name
+	int	m_sys			# sys help file
+	int	m_src			# source file
+	int	m_pkg			# package helpdir file
+}
+.fi
+
+All character strings are stored in the string buffer, which is initially
+allocated with malloc, and which is reallocated with realloc if it fills
+up while the helpdir is being read.  The maximum number of logical
+directory definitions and module entries are fixed when \fBhelp\fR is
+compiled.  All strings are referred to by an integer index into the
+string buffer (if the string buffer is moved by realloc, the integer
+offset does not change, whereas a char pointer would).
+
+.nh
+Semicode
+
+HELP -- The main procedure.  Expand the module list, determine the
+type of help desired, fetch the names of the doc files and call the
+appropriate routine to process the help text.
+
+.nf
+procedure help
+
+begin
+	# Get control parameters.
+	fetch and decode option string
+	if (option "param")
+	    fetch parameter name
+	fetch margin params, pagination flag, open tty descriptor,
+	    set up "ctrl" structure
+
+	# Get module template: "mod", "pack.mod", "pack.", etc.  If no
+	# args or template is null, default to "curpack.".
+
+	if (we were called with no positional arguments)
+	    template = null
+	else
+	    fetch template from cl
+
+	# Format and output the help text.  The template list consists
+	# of a series of templates delimited by commas.
+
+	for (each subtemplate in the template list)
+	    call process_template (template, ctrl)
+end
+.fi
+
+
+PROCESS_TEMPLATE -- Called with a template defining the packages and
+modules for which help is desired, and the control parameters defining
+the type of help desired.  Expand the template into a list of packages
+and modules and process the help for each module.
+
+.nf
+procedure process_template (template, ctrl)
+
+begin
+	# Open the system help directory; gives help files and name of
+	# package help directories for the individual packages.
+
+	hp_sys = hd_open (system helpdir)
+
+	# If null template, print hlp for current package.  If template
+	# not null, but no package named, make package list the current
+	# CL package search path.  Otherwise the template contains all
+	# necessary information.
+
+	if (entire template is null) {
+	    set package template to the name of the current package
+	    set module template to null
+	} else {
+	    extract package part of template, delimited by "."
+	    if (null package template)
+		set package template to CL package search path
+	}
+
+	# By the time we get here we always have a non-null template.
+	paklist = expand_template (hp_sys, pak_template)
+
+	# If package help is desired, print help on the package as if
+	# it were a module.  Otherwise, expand module template for
+	# the package and process help on each module.
+
+	while (read_template (paklist, pakname) != EOF)
+	    if (null module template)
+		call print_help (hp_sys, pakname, ctrl)
+	    else {
+		hp_pak = hd_open (package helpdir)
+		modlist = expand_template (hp_pak, mod_template)
+		while (read_template (modlist, modname) != EOF)
+		    call print_help (hp_pak, modname, ctrl)
+		close_template (modlist)
+		hd_close (hp_pak)
+	    }
+
+	close_template (paklist)
+	hl_close (hp_sys)
+end
+.fi
+
+
+EXPAND_TEMPLATE -- Expand a template into a list of module names,
+given a help directory defining the pattern matching domain.
+The permissible complexity of a help template is somewhat less than
+that of a file template.  A template consists of a list of patterns;
+there is nothing corresponding to the logical directories and list
+files used in file templates.  Examples of templates include
+
+.nf
+	(null)			package-help for the current package
+	cl.			package-help clpackage
+	cl.*			all tasks in clpackage
+	*			all tasks in the current package
+	alpha			module alpha
+.fi
+
+The simple alphanumeric string is a special type of pattern.  The string
+"cl", for example, is equivalent to "cl*" or "cl?*".  Thus template
+expansion is a simple matter of scanning through the module list of
+a help directory and extracting all names which match the pattern.
+
+
+.nf
+pointer procedure expand_template (hp, template)
+
+hp:	helpdir descriptor
+
+begin
+	allocate template descriptor, containing space for the encoded
+	    pattern, hp, and the module index.
+
+	save hp
+	encode template into descriptor; turn "*" meta-characters
+	    into "?*" if not following character class
+	set module index to zero
+
+	return (pointer to template descriptor)
+end
+.fi
+
+
+READ_TEMPLATE -- Get next module name from help directory matching the
+encoded pattern.  Return EOF when directory is exhausted.
+
+.nf
+int procedure read_template (template_descriptor, module_name)
+
+begin
+	for (index+=1;  hp_getname (hp, module_name) != 0;  index+=1) {
+	    if (pattern matches module name)
+		return
+	}
+	return (EOF)
+end
+.fi
+
+
+PRINT_HELP -- Print help documentation for the named module or parameter.
+We are called with the name of a single module; all fiddling with packages
+and templates has been performed by the time we are called.  The help
+directory is open and contains the names of the files containing the help
+source for the module.  Our main task is to determine what kind of help
+is desired and call the appropriate routine.
+
+Recall that the principal options are
+
+.nf
+       \fBoption\fR                     \fBmeaning\fR                 \fBfile\fR
+
+	help		print help block for module	   hlp
+	param		print help for single param	   hlp
+	section		print a single section		   hlp
+	files	        print file names		   ...
+	source		print source file		   src
+	sysdoc		print system documentation	   sys
+	alldoc		print all documentation		 hlp,sys
+	summary		print help block titles		 hlp,sys
+.fi
+
+
+.nf
+procedure print_help (hp, modname, ctrl)
+
+begin
+	# Handle options which do not access a help file.
+	if (option == files) {
+	    pr_filenames (hp, modname, ctrl)
+	    return
+	} else if (option == source) {
+	    get source file name from hp_getname
+	    pr_sourcefile (sourcefile, ctrl)
+	    return
+	} else if (option == summary) {
+	    # Scan hlp file and print summary of help blocks.
+	    get helpfile name from hp_getname
+	    pr_summary (helpfile, ctrl)
+	    return
+	}
+
+	# Process help block.  Processing is controlled by the
+	# "ctrl" structure.
+
+	if (option == help or alldoc) {
+	    get helpfile name from hp_getname
+	    pr_helpblock (helpfile, module, ctrl)
+	if (option == sysdoc or alldoc) {
+	    get sysdocfile name from hp_getname
+	    pr_helpblock (sysdocfile, module, ctrl)
+	}
+end
+.fi
+
+
+PR_HELPBLOCK -- Format and print a help block.  Open the help file and search
+for the named module.  Interpret block header and print help title.
+Process the remainder of the help block with lroff.
+
+.nf
+procedure pr_helpblock (helpfile, module, ctrl)
+
+begin
+	iferr (in = open help file)
+	    print warning message
+	if (find_help_block (in, module, lbuf) == not found)
+	    cannot find help block for module 'modname'
+
+	clear EOF flag in ctrl
+	initialize line counter
+
+	if (not printing just single param or section)
+	    pr_header (lbuf, ctrl)
+	if (lroff (input, ctrl, output, ctrl, ...) == ERR)
+	    print warning message
+	close help file
+end
+.fi
+
+
+INPUT -- Lroff line input procedure.  Called by Lroff to get lines of
+input text.  Function varies slightly depending on the Help option.
+If printing only single param or single section, our job is to eat all
+input which is not part of the indicated section of the help block.
+A parameter block begins with ".ls paramname" and ends with a matching
+".le" or ".ih", if the text is formatted.  Similarly, the single section
+begins with a ".ih" followed by the section name on the next line.
+
+.nf
+int procedure input (ctrl, outbuf)
+
+ctrl:	Control structure.  Contains fd of input file, and the EOF
+	flag which may be set by the \fBoutput\fR procedure if we are
+	called interactively.
+
+begin
+	# Eof flag is set after param block has been fully input,
+	# but normally before the actual end of file.
+	if (eof flag was set by user interaction)
+	    return (EOF)
+	else if (processing full help block)
+	    return (getline (infile, outbuf))
+
+	# We get here only if special processing is required to
+	# filter out all but a section of the help text.
+
+	if (first call for new file) {
+	    # Determine whether or not the help block is formatted.
+	    get first line of help block
+	    if (line is an Lroff directive)
+		formatted = true
+	    else
+		formatted = false
+	    return, passing line to Lroff
+	}
+
+	if (second call for new file) {
+	    if (single param mode) {
+		if (formatted) {
+		    eat input lines until a ".ls" directive is
+			found which contains the param name as
+			a substring.
+		} else {
+		    eat input lines until a line beginning with
+			the parameter name is found.
+		}
+	    } else if (print a single section) {
+		if (formatted) {
+		    eat input lines until ".ih\nSECNAME" is found
+		} else {
+		    eat input lines until a line beginning with the
+			parameter name or SECNAME is found.
+		}
+		put a section indent directive into line buffer
+	    }
+	    return, passing line to Lroff
+	}
+
+	# By the time we get here we are in the parameter or single
+	# section.
+	get input line
+	if (line is a .ih directive)
+	    set EOF flag
+	else if (in single param mode and have a matching .le)
+	    set EOF flag
+
+	return, passing line to Lroff
+end
+.fi
+
+
+OUTPUT -- Output a line of text to the "out" file.  If the standard output
+is redirected, put lines out as is with \fBputline\fR.  Otherwise put lines out
+to the interactive terminal with \fBttyputline\fR, which processes standout
+mode etc.  Count lines and pause between pages, if enabled by control
+flag.  Pause is implemented as a request for a CL query mode boolean
+parameter, just as in the \fBpage\fR utility.
+
+.nf
+procedure output (dev, linebuf)
+dev:	The "ctrlpar" structure, containing output file descriptor,
+	pagination flag, TTY descriptor pointer, left and right
+	margins, flag if output is redirected, and so on.
+
+begin
+	if (output is redirected) {
+	    call putline to output line without further processing
+	    return
+	}
+
+	call ttyputline to output line
+	bump line counter
+
+	# The NOMORE procedure produces a query on the user terminal,
+	# giving the user a chance to continue when they have finished
+	# reading the current screen, or of terminating the help block
+	# currently being processed.
+
+	if (output page is full)
+	    if (nomore)
+		set EOF flag for input procedure
+end
+.fi
diff --git a/pkg/system/help/filetemp.x b/pkg/system/help/filetemp.x
new file mode 100644
index 00000000..1270abe2
--- /dev/null
+++ b/pkg/system/help/filetemp.x
@@ -0,0 +1,28 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	"help.h"
+
+# DO_FILE_TEMPLATE -- Called with a template matching a list of files containing
+# help blocks.  All help blocks therein are processed and output.
+
+procedure do_file_template (template, ctrl)
+
+char	template[ARB]			# filename matching template
+pointer	ctrl				# Help control structure
+
+char	fname[SZ_FNAME]
+int	option
+pointer	list
+pointer	fntopn()
+int	fntgfn()
+
+begin
+	list = fntopn (template)
+	option = H_OPTION(ctrl)
+
+	while (fntgfn (list, fname, SZ_FNAME) != EOF)
+	    call pr_helpblock (fname, "", "", NULL, ctrl)
+
+	call fntcls (list)
+end
diff --git a/pkg/system/help/getoption.x b/pkg/system/help/getoption.x
new file mode 100644
index 00000000..a7504531
--- /dev/null
+++ b/pkg/system/help/getoption.x
@@ -0,0 +1,52 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	"help.h"
+
+# GET_OPTION -- Map the possibly abbreviated option string into an integer
+# option code.  It is an error if the abbreviation is not unique.
+
+define	EQUAL		0
+
+int procedure get_option (opstr)
+
+char	opstr[ARB]		# option string, always lower case
+int	keycode[MAX_OPTIONS]
+int	oplen, key, ip, match
+int	strlen(), strncmp()
+
+string	keywords "help|sources|sysdoc|alldoc|files|summary|directory|references"
+data	keycode /O_HELP,O_SOURCE,O_SYSDOC,O_ALLDOC,O_FILES,O_SUMMARY,O_DIR,
+		 O_REFERENCES/
+
+begin
+	oplen = strlen (opstr)
+	ip = 1
+	match = 0
+
+	# Search the keyword table.  If the option string matches a keyword,
+	# it is either an exact match (oplen = keyword length) or a legal
+	# abbreviation.  If an abbreviation matches two keywords it is
+	# ambiguous and an error.
+
+	for (key=1;  keywords[ip] != EOS;  key=key+1) {
+	    if (strncmp (keywords[ip], opstr, oplen) == EQUAL) {
+		if (keywords[ip+oplen] == '|')
+		    return (keycode[key])		# exact match
+		else {
+		    if (match != 0)
+			call error (1, "ambiguous help option abbreviation")
+		    else
+			match = key
+		}
+	    }
+	    repeat {
+		ip = ip + 1
+	    } until (keywords[ip-1] == '|' || keywords[ip] == EOS)
+	}
+
+	if (match == 0)
+	    call error (2, "unknown help option")
+	else
+	    return (keycode[match])
+end
diff --git a/pkg/system/help/hbgetblk.x b/pkg/system/help/hbgetblk.x
new file mode 100644
index 00000000..a39db034
--- /dev/null
+++ b/pkg/system/help/hbgetblk.x
@@ -0,0 +1,195 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	"help.h"
+
+.help hb_getnextblk
+.nf ___________________________________________________________________________
+HB_GETNEXTBLK -- Scan a file for the next help block, i.e., block of lines
+delimited by ".help" and ".endhelp".  The syntax of a help block header is
+as follows:
+
+	.help[typestr] key1,key2,...,keyN section title
+
+Whitespace is NOT permitted in the keyword list unless it is quoted.  Quotes
+are optional for each of the three strings.  If the line ends with the
+backslash character or a comma we assume that the header is continued on the
+next line.  The typestr, denoting the type of help block, is optional.
+If absent it defaults to "hlp".  System help blocks are of type "sys".
+
+We keep track of the file line number when searching for a help block.  The
+line number is updated in the HB structure.  For this number to be accurate,
+the caller must see that it is initialized before we are called.
+.endhelp ______________________________________________________________________
+
+define	SZ_SBUF		(MAX_KEYS * SZ_KEY)
+define	IS_KEYWCHAR	(IS_ALNUM($1) || $1 == '_' || $1 == '$')
+define	exit_		90
+
+
+int procedure hb_getnextblk (hb, ctrl)
+
+pointer	hb
+pointer	ctrl
+
+bool	at_eof
+char	key[SZ_KEY]
+int	fd, n, ip, op, junk
+pointer	sp, lbuf, sbuf, p
+int	hb_getstr(), strmatch(), getline(), strlen()
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+	call salloc (sbuf, SZ_SBUF, TY_CHAR)
+
+	# Search forward for the next help block.
+	fd = H_IN(ctrl)
+	at_eof = true
+
+	while (getline (fd, Memc[lbuf]) != EOF) {
+	    HB_LINENO(hb) = HB_LINENO(hb) + 1
+	    if (strmatch (Memc[lbuf], "^.{help}") > 0) {
+		at_eof = false
+		break
+	    }
+	}
+
+	if (at_eof) {
+	    call sfree (sp)
+	    return (EOF)
+	}
+
+	# Initialize everything in case of an early exit.
+	HB_TYPE(hb)	= TY_UNKNOWN
+	HB_NKEYS(hb)	= 0
+	HB_TYPESTR(hb)	= EOS
+	HB_SECTION(hb)	= EOS
+	HB_TITLE(hb)	= EOS
+
+	# Decode help block into the HB structure.  First step is to
+	# get the type suffix string, if any.
+
+	ip = strlen (".help") + 1
+	p = lbuf + ip - 1
+	if (IS_ALPHA (Memc[p])) {
+	    for (n=0;  IS_ALNUM (Memc[p+n]);  n=n+1)
+		;
+	    call strcpy (Memc[p], HB_TYPESTR(hb), min(n,SZ_TYPESTR))
+	    if (     strmatch (HB_TYPESTR(hb), "{sys}") > 0)
+		HB_TYPE(hb) = TY_SYS
+	    else if (strmatch (HB_TYPESTR(hb), "{hlp}") > 0)
+		HB_TYPE(hb) = TY_HLP
+	    else
+		HB_TYPE(hb) = TY_UNKNOWN
+	} else {
+	    HB_TYPE(hb) = TY_HLP
+	    n = 0
+	}
+	ip = ip + n
+
+	# Now get the keyword string, and break the keywords out into
+	# the keyword list.
+
+	if (hb_getstr (fd, Memc[lbuf], ip, Memc[sbuf], SZ_SBUF,
+	HB_LINENO(hb)) == 0)
+	    goto exit_
+
+	p = sbuf
+	while (IS_WHITE (Memc[p]) || Memc[p] == ',')
+	    p = p + 1
+	for (n=1;  n <= MAX_KEYS && Memc[p] != EOS;  n=n+1) {
+	    for (op=1;  IS_KEYWCHAR (Memc[p]);  op=op+1) {
+		key[op] = Memc[p]
+		p = p + 1
+	    }
+	    key[op] = EOS
+	    call strcpy (key, HB_KEY(hb,n), SZ_KEY)
+	    while (IS_WHITE (Memc[p]) || Memc[p] == ',')
+		p = p + 1
+	}
+	HB_NKEYS(hb) = n - 1
+		
+	# Fetch section label string and title string.
+
+	if (hb_getstr (fd, Memc[lbuf], ip, HB_SECTION(hb), SZ_SECTION,
+	HB_LINENO(hb)) == 0)
+	    goto exit_
+	junk = hb_getstr (fd, Memc[lbuf], ip, HB_TITLE(hb), SZ_TITLE,
+	    HB_LINENO(hb))
+
+exit_	call sfree (sp)
+	return (HB_LINENO(hb))
+end
+
+
+# HB_GETSTR -- Fetch a string (optionally quoted) from the line buffer.
+# Handle everything having to do with continuation.
+
+int procedure hb_getstr (fd, lbuf, ip, outstr, maxch, lineno)
+
+int	fd
+char	lbuf[ARB]
+int	ip
+char	outstr[ARB]
+int	maxch
+int	lineno
+
+char	ch
+int	op, dstart
+int	stridx(), getline()
+string	delim " \t'\""
+
+begin
+	while (IS_WHITE (lbuf[ip]))
+	    ip = ip + 1
+	op = 1
+
+	# If quoted string, only a quote can end the string.
+	dstart = 1
+	if (lbuf[ip] == '\'' || lbuf[ip] == '"') {
+	    ip = ip + 1
+	    dstart = 3
+	}
+
+	# Fetch the string.
+	for (ch=lbuf[ip];  stridx (ch, delim[dstart]) == 0;  ch=lbuf[ip]) {
+	    if (op >= maxch)
+		break
+
+	    switch (ch) {
+	    case '\\', ',':
+		if (lbuf[ip+1] == '\n' || lbuf[ip+1] == EOS) {
+		    # Continue on next line.
+		    if (getline (fd, lbuf) == EOF)
+			lbuf[1] = EOS
+		    else
+			lineno = lineno + 1
+		    ip = 1
+		    if (ch == '\\')
+			next
+		} else
+		    ip = ip + 1
+		outstr[op] = ch
+		op = op + 1
+
+	    case '\n', EOS:
+		if (delim[dstart] == '\'') {
+		    call eprintf ("Missing right quote in helpfile, line %d\n")
+			call pargi (lineno)
+		}
+		break
+
+	    default:
+		outstr[op] = ch
+		op = op + 1
+		ip = ip + 1
+	    }
+	}
+
+	if (ch != EOS)
+	    ip = ip + 1
+	outstr[op] = EOS
+
+	return (op - 1)
+end
diff --git a/pkg/system/help/hdbexamine.hlp b/pkg/system/help/hdbexamine.hlp
new file mode 100644
index 00000000..a7498c23
--- /dev/null
+++ b/pkg/system/help/hdbexamine.hlp
@@ -0,0 +1,55 @@
+.help hdbexamine Feb86 softools
+.ih
+NAME
+hdbexamine -- examine a help database
+.ih
+USAGE
+hdbexamine
+.ih
+PARAMETERS
+.ls helpdb = "helpdb"
+The filename of the help database to be examined.  The reserved name "helpdb"
+causes the actual filename to be taken from the environment variable of
+the same name.
+.le
+.ls verbose = no
+If this switch is enabled, \fIhdbexamine\fR will print a detailed description
+of the help database listing the modules in each package, the date the entry
+for the package was last modified, and other information.  A more concise
+summary listing only the packages and the number of help modules in each
+package is printed by default.
+.le
+.ih
+DESCRIPTION
+The \fIhdbexamine\fR task is used to examine the contents of a help
+database.  By default the standard IRAF help database is examined.
+Examining the help database with \fIhdbexamine\fR verifies that it can
+be read by \fIhelp\fR, and may be useful as a diagnostic in the event
+that an invalid help directory file (".hd") somewhere in the help
+directory tree, causes the database to be compiled incorrectly.
+.ih
+EXAMPLES
+1. Print a concise summary of the contents and structure of the standard
+help database.
+
+.nf
+cl> hdbexamine
+Help database dev$help.db created Feb 14 21:34 by tody
+Database contains 794 modules in 43 packages, file size 105460 bytes
+
+_clpackage   Nov  4  1984 tody     clpackage$_clpackage.hd
+clpackage    May 29 17:44 rooke    clpackage$clpackage.hd
+os           Feb 13 11:06 tody     host$os/doc/os.hd
+root         Nov  4  1984 tody     lib$root.hd
+_math        Apr  1 13:38 tody     math$_math.hd
+curfit       Jan  6 16:23 tody     math$curfit/doc/curfit.hd
+gsurfit      Jan  2 14:46 davis    math$gsurfit/doc/gsurfit.hd
+iminterp     Aug  6 16:31 davis    math$iminterp/doc/iminterp.hd
+bias         Dec 17  8:53 valdes   pkg$imred/bias/bias.hd
+coude        Dec 31 14:38 valdes   pkg$imred/coude/coude.hd
+vtel         Jan 22  8:36 lytle    pkg$imred/vtel/vtel.hd
+plot         Jan 28 14:04 hammond  pkg$plot/plot.hd
+.fi
+.ih
+SEE ALSO
+mkhelpdb, help
diff --git a/pkg/system/help/help.h b/pkg/system/help/help.h
new file mode 100644
index 00000000..e1ec7070
--- /dev/null
+++ b/pkg/system/help/help.h
@@ -0,0 +1,115 @@
+# Help Definitions.
+
+# Control Structure.  Contains all control parameters.  Pointer to structure
+# is passed to Lroff and on by Lroff to the input and output procedures.
+# With the exceptions of H_EOF and H_NLINES, the control parameters are
+# read only outside the main routine.
+
+define	LEN_CTRLSTRUCT	220
+define	SZ_CURPACK	39		# current package
+define	SZ_SECNAME	39		# section name, single section mode
+define	SZ_PARNAME	39		# parameter name
+define	SZ_TEMPLATE	79		# the original modules template
+define	SZ_HELPDB	1024		# max chars in helpdb file list
+
+define	H_IN		Memi[$1]	# input file descriptor
+define	H_OUT		Memi[$1+1]	# output file descriptor
+define	H_OPTION	Memi[$1+2]	# option code (see below)
+define	H_TTY		Memi[$1+3]	# TTY device descriptor
+define	H_LMARGIN	Memi[$1+4]	# permanent left margin for Lroff
+define	H_RMARGIN	Memi[$1+5]	# permanent right margin for Lroff
+define	H_RAWIN		Memi[$1+6]	# if YES, do not look at input
+define	H_RAWOUT	Memi[$1+7]	# if YES, do not process output
+define	H_FILTER_INPUT	Memi[$1+8]	# if YES, filter out part of input
+define	H_PAGINATE	Memi[$1+9]	# paginate output?
+define	H_MANPAGE	Memi[$1+10]	# manpage style output?
+define	H_NLPP		Memi[$1+11]	# number of lines per manual page
+define	H_NLINES	Memi[$1+12]	# number of output lines on page
+define	H_STATE		Memi[$1+13]	# input state, hinput()
+define	H_EOF		Memi[$1+14]	# input should return EOF to Lroff
+define	H_QUIT		Memi[$1+15]	# stop program
+define	H_LENTL		Memi[$1+16]	# length of the TL template list
+define	H_ALLMODULES	Memi[$1+17]	# process all modules matching template
+define	H_FORMAT	Memi[$1+18]	# output format type
+define	H_SOFLAG	Memi[$1+19]	# standout flags
+			# (extra space)
+define	H_CURPACK	Memc[P2C($1+20)]	# current package
+define	H_SECNAME	Memc[P2C($1+60)]	# section name
+define	H_PARNAME	Memc[P2C($1+100)]	# parameter name
+define	H_TEMPLATE	Memc[P2C($1+140)]	# module name template
+
+# The nomore flag is set whenever the user responds negatively to the more?
+# query.  A nomore at the beginning of a help block or two nomores in a row
+# stop the program.
+
+define	NOMORE		(-1)
+
+# Option codes.  Max_options is used by the get_option keyword recognizer,
+# which handles abbreviations.
+
+define	O_HELP		1		# print full help block
+define	O_SOURCE	2		# print source code
+define	O_SYSDOC	3		# print technical system documentation
+define	O_ALLDOC	4		# print all documentation (!source)
+define	O_FILES		5		# print file names
+define	O_SUMMARY	6		# summarize contents of help file
+define	O_DIR		7		# print directory of help blocks
+define	O_REFERENCES	8		# keyword search output format
+define	MAX_OPTIONS	8
+
+define	O_PARAM		7		# output text for single parameter
+define	O_SECTION	8		# output text for single section
+define	O_MENU		9		# print package menu
+
+# Type codes for filenames.  Passed to hd_getname to fetch the module name
+# or a filename from a help directory.
+
+define	TY_MODNAME	0
+define	TY_HLP		1
+define	TY_SYS		2
+define	TY_SRC		3
+define	TY_PKG		4
+define	TY_MEN		5
+define	TY_UNKNOWN	6
+
+# Format codes.  This is used to determine whether we output formatted
+# text, HTML or Postscript.
+
+define	HF_DEVICES	"|terminal|text|gui|html|postscript|ps|"
+define	HF_TERMINAL	1
+define	HF_TEXT		2
+define	HF_GUI		3
+define	HF_HTML		4
+define	HF_POSTSCRIPT	5
+define	HF_PS		6
+
+# Help block header structure.  A ".help" directive is decoded into this
+# structure.  The line number counter should be zeroed when the structure
+# is allocated.
+
+define	LEN_HBSTRUCT	1124
+define	MAX_KEYS	50
+define	SZ_TYPESTR	9		# help block type string
+define	SZ_KEY		19		# max size of a key
+define	SZ_SECTION	39		# section label, i.e., (Mar84)
+define	SZ_TITLE	63		# block title
+
+define	HB_TYPE		Memi[$1]	# type of block (default TY_HLP)
+define	HB_LINENO	Memi[$1+1]	# line number within file
+define	HB_NKEYS	Memi[$1+2]	# number of keys
+define	HB_TYPESTR	Memc[P2C($1+10)]		# blk type string
+define	HB_KEY		Memc[P2C($1+20+($2-1)*20)]	# keys
+define	HB_SECTION	Memc[P2C($1+1020)]		# section label
+define	HB_TITLE	Memc[P2C($1+1060)]		# block title
+
+# Pagination Control Codes.  The pagination directives BP, TP, KS, and KE
+# are ignored when output is directed to the terminal, but are important
+# when output is piped to the printer.  When the line input routine HINPUT sees
+# one of these directives in the input it places a control code in the
+# data stream read by Lroff.  Lroff passes control codes on to the output,
+# i.e., to HOUTPUT, where pagination takes place.
+
+define	BREAK_PAGE	1		# .bp
+define	TEST_PAGE	2		# .tp n
+define	START_KEEP	3		# .ks
+define	END_KEEP	4		# .ke
diff --git a/pkg/system/help/helpdb.x b/pkg/system/help/helpdb.x
new file mode 100644
index 00000000..04af40db
--- /dev/null
+++ b/pkg/system/help/helpdb.x
@@ -0,0 +1,1203 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<mach.h>
+include	<syserr.h>
+include	<error.h>
+include	<finfo.h>
+include	<fset.h>
+include	<time.h>
+include	<mii.h>
+include	"help.h"
+include	"helpdir.h"
+
+.help helpdb
+.nf ___________________________________________________________________________
+HELPDB -- Code to manage the help database.  The help database is a set of
+compiled package help directories.  Each help directory file is compiled with
+HD_OPEN and the resultant structure is saved in the database, indexed by the
+name of the directory file.  At run time, instead of opening and interpreting
+multiple help directory files to expand templates, we merely read in the
+compiled database and search the index for the entry for the named package,
+returning a pointer to the compiled help directory structure for the package.
+The help database must be recompiled with HDB_COMPILE whenever the system help
+directory or a package help directory is modified.
+
+		hdb_compile (root_helpdir_filename, database_filename)
+
+	db =	   hdb_open (database)
+		  hdb_close (db)
+	hp =	   hdb_load (db, help_directory_file)
+		   hdb_free (db, hp)
+
+There is no provision for updating the database; the whole thing must be
+recompiled with HDB_COMPILE.  At HELP time the root help directory is opened
+with HDB_OPEN, naming either a raw root help directory file or a compiled
+help database file.  Normally a precompiled help database is used, but if
+desired the help directory files will be accessed directly.  The directory
+for an individual package is accessed with HDB_LOAD, which is functionally
+equivalent to HD_OPEN (helpdir.x).  HDB_FREE is functionally equivalent to
+HD_CLOSE.
+
+Currently the entire database is read into memory to speed up database
+searches.  In the future this implementation should be changed to use the
+DBIO faciltities if the database becomes large.
+.endhelp ______________________________________________________________________
+
+define	DATA_OFFSET	513		# offset to data area, chars
+define	MAX_ENTRIES	100		# initial max db entries
+define	INC_ENTRIES	50		# increment if overflow
+define	MAX_DEPTH	20		# max nesting of packages
+define	MAX_MENUSIZE	500		# max modules in a table
+define	MAX_NAMELEN	20		# max chars in a module name in table
+define	FIRST_COL	6		# indentation of tables
+
+# Help database header structure.  Stored at the beginning of a help
+# database file.
+
+define	LEN_HDBHEADER	14
+define	HDB_MAGICVAL	110104B
+
+define	HDB_MAGIC	Memi[$1]	# helpdb file type code
+define	HDB_RAW		Memi[$1+1]	# access compiled or raw database
+define	HDB_RHD		Memi[$1+2]	# if raw, HP of root help directory
+define	HDB_INDEX	Memi[$1+3]	# index of root help directory
+define	HDB_CRDATE	Meml[$1+4]	# creation date
+define	HDB_NENTRIES	Memi[$1+5]	# number of help directories in db
+define	HDB_MAXENTRIES	Memi[$1+6]	# maximum no. of help directories in db
+define	HDB_NMODULES	Memi[$1+7]	# count of the total number of modules
+define	HDB_INDEXOFFSET	Meml[$1+8]	# file offset of index, chars
+define	HDB_INDEXPTR	Memi[$1+9]	# pointer to loaded index, ty_struct
+define	HDB_INDEXLEN	Memi[$1+10]	# length of index structure, su
+define	HDB_DATAOFFSET	Meml[$1+11]	# file offset of data area, chars
+define	HDB_DATAPTR	Memi[$1+12]	# pointer to loaded data area, ty_struct
+define	HDB_DATALEN	Memi[$1+13]	# length of data area, struct units
+
+# Index structure.  Identifies the contents of the database and tells where
+# they are stored.  There is one index entry for each help directory, i.e.,
+# for each package.
+
+define	LEN_HDBINDEX	34
+define	SZ_DBIKEY	63
+define	LEN_DBIDATA	2
+
+define	DBI_KEY		Memc[P2C($1)]	# entry name
+define	DBI_OFFSET	Memi[$1+32]	# offset of entry into data area, su
+define	DBI_MTIME	Meml[$1+33]	# modification date of entry
+
+
+# HDB_COMPILE -- Compile a set of help directories to produce a new help
+# database.  The root help directory is read first and entered into
+# the database, followed by each subdirectory named in the root help
+# directory.
+
+procedure hdb_compile (root_helpdir_file, helpdb_file, verbose)
+
+char	root_helpdir_file[ARB]	# name of root help directory file
+char	helpdb_file[ARB]	# name of new database file
+bool	verbose			# print informative messages
+
+pointer	db, index, p1, p2, ix
+bool	no_entries_interchanged
+int	fd, temp[LEN_HDBINDEX], i
+
+int	open()
+bool	strgt(), streq()
+long	clktime(), note(), fstatl()
+errchk	open, note, seek, hdb_compile_rhd, mii_writec, mii_writei
+
+begin
+	# Open the output database file.
+	fd = open (helpdb_file, NEW_FILE, BINARY_FILE)
+
+	# Allocate and initialize the database header and index structures.
+	# The data area begins at a fixed offset and will be written out
+	# as we go.  The index and db header are written out after all
+	# help directories have been processed.
+
+	call calloc (db, LEN_HDBHEADER, TY_STRUCT)
+	call calloc (index, MAX_ENTRIES * LEN_HDBINDEX, TY_STRUCT)
+
+	HDB_MAGIC(db)      = HDB_MAGICVAL
+	HDB_MAXENTRIES(db) = MAX_ENTRIES
+	HDB_CRDATE(db)     = clktime (long(0))
+
+	# Write zeros into the header area of the file, so that the next
+	# sequential write will place the first data record at the offset
+	# DATA_OFFSET.  We assume that the empty index is larger than the
+	# header area.
+
+	call write (fd, Memi[index], DATA_OFFSET - 1)
+	HDB_DATAOFFSET(db) = note (fd)
+
+	# Compile root directory followed by all subdirectories.
+	# The root directory is expanded into a set of package help
+	# directories which are compiled and written into the data
+	# area of the database.  The database descriptor and index
+	# structures are returned.
+
+	call hdb_compile_rhd (fd, root_helpdir_file, db, index, verbose)
+
+	# Sort the index and append it to the database file.  A simple
+	# interchange sort is sufficient here; we have assumed that the
+	# database is small in this implementation of helpdb.
+
+	if (HDB_NENTRIES(db) > 1)
+	    repeat {
+		no_entries_interchanged = true
+		do i = 1, HDB_NENTRIES(db) - 1 {
+		    p1 = index + (i - 1) * LEN_HDBINDEX
+		    p2 = p1 + LEN_HDBINDEX
+		    if (strgt (DBI_KEY(p1), DBI_KEY(p2))) {
+			call amovi (Memi[p1], temp,     LEN_HDBINDEX)
+			call amovi (Memi[p2], Memi[p1], LEN_HDBINDEX)
+			call amovi (temp,     Memi[p2], LEN_HDBINDEX)
+			no_entries_interchanged = false
+		    }
+		}
+	    } until (no_entries_interchanged)
+
+	HDB_INDEXOFFSET(db) = note (fd)
+	HDB_INDEXLEN(db) = HDB_NENTRIES(db) * LEN_HDBINDEX
+	HDB_MAXENTRIES(db) = HDB_NENTRIES(db)
+
+	# Get the offset into the index of the "_index" help directory
+	# (package name list).
+
+	do i = 1, HDB_NENTRIES(db) {
+	    ix = index + (i - 1) * LEN_HDBINDEX
+	    if (streq (DBI_KEY(ix), "_index")) {
+		HDB_INDEX(db) = i
+		break
+	    }
+	}
+	if (HDB_INDEX(db) == 0)
+	    call eprintf ("Help warning: cannot find _index\n")
+
+	# Write the index structure to the database file.
+	do i = 1, HDB_NENTRIES(db) {
+	    ix = index + (i - 1) * LEN_HDBINDEX
+	    call mii_writec (fd, DBI_KEY(ix), SZ_DBIKEY + 1)
+	    call mii_writei (fd, DBI_OFFSET(ix), LEN_DBIDATA)
+	}
+
+	# Update the database file header.
+	call seek (fd, BOFL)
+	call mii_writei (fd, Memi[db], LEN_HDBHEADER)
+
+	call printf ("\nTotal of %d help modules in %d packages, ")
+	    call pargi (HDB_NMODULES(db))
+	    call pargi (HDB_NENTRIES(db) - 1)
+	call printf ("file size %d bytes\n")
+	    call pargl (fstatl (fd, F_FILESIZE) * SZB_CHAR)
+	    
+	# All done; close files and deallocate buffers.
+	call close (fd)
+	call mfree (index, TY_STRUCT)
+	call mfree (db, TY_STRUCT)
+end
+
+
+# HDB_COMPILE_RHD -- Compile the root help directory and all subdirectories
+# into the database.  The package structure is hierarchical but the database
+# is linear.  Subdirectories (subpackages) are recursively expanded starting
+# with all packages in the root help directory.  Each subdirectory is entered
+# into the database as it is found.  We end up with an inverted index keyed
+# by the filename of the help directory file for each package.  Since the
+# key is a filename each key is guaranteed to be unique.
+
+procedure hdb_compile_rhd (fd, root_helpdir_file, db, index, verbose)
+
+int	fd			# database file, written sequentially
+char	root_helpdir_file[ARB]	# name of root help directory file
+pointer	db			# database descriptor
+pointer	index			# database index
+bool	verbose			# print notes on structure of database
+
+bool	found_a_subpackage
+pointer	hp_stk[MAX_DEPTH]	# help directory pointer stack
+int	pk_stk[MAX_DEPTH]	# subpackage index stack
+char	fname[SZ_FNAME]		# helpdir filename
+
+int	sp, pk, len_index
+pointer	ix, hp, modname, data, sv_sbuf
+long	mtime, fi[LEN_FINFO], savepos
+
+long	note(), clktime()
+int	hd_getname(), finfo()
+pointer	hd_open(), hdb_make_rhd()
+errchk	finfo, seek, note, mii_writei
+errchk	hd_getname, hdb_make_rhd, hdb_getdata
+
+begin
+	# Initialize the stacks and open the next help directory file to be
+	# processed, i.e., the root directory file.
+
+	sp = 0
+	pk = 1
+	call strcpy (root_helpdir_file, fname, SZ_FNAME)
+
+	iferr (hp = hd_open (root_helpdir_file))
+	    call error (2, "cannot open root help directory file")
+	if (finfo (root_helpdir_file, fi) == ERR)
+	    call error (3, "cannot get finfo on root help directory file")
+	mtime = FI_MTIME(fi)
+
+	# We enter the compile loop ready to scan the next help directory file,
+	# which has been pushed onto the top of the stack.  The help directory
+	# file is first entered into the database, then we scan the directory
+	# until a subdirectory is found.  If a subdirectory is found it is
+	# opened and pushed onto the stack and the process repeats (hd_open
+	# does not leave the physical file open so running out of file
+	# descriptors is not a problem).  When the end of a directory is
+	# reached the stack is popped, closing the current directory, and 
+	# we continue to scan the previous directory until another subdirectory
+	# is found.
+
+	repeat {
+	    # Append the current compiled package help directory to the
+	    # database.  Update index; make index larger if it overflows.
+
+	    HDB_NMODULES(db) = HDB_NMODULES(db) + HD_NMODULES(hp)
+	    HDB_NENTRIES(db) = HDB_NENTRIES(db) + 1
+
+	    if (HDB_NENTRIES(db) > HDB_MAXENTRIES(db)) {
+		HDB_MAXENTRIES(db) = HDB_MAXENTRIES(db) + INC_ENTRIES
+		len_index = HDB_MAXENTRIES(db) * LEN_HDBINDEX
+		iferr (call realloc (index, len_index, TY_STRUCT))
+		    call fatal (1, "cannot reallocate index buffer")
+	    }
+	    ix = index + ((HDB_NENTRIES(db) - 1) * LEN_HDBINDEX)
+
+	    # Prepare the index entry for the help directory file.
+	    call strcpy (fname, DBI_KEY(ix), SZ_DBIKEY)
+	    DBI_MTIME(ix) = mtime
+	    DBI_OFFSET(ix) = HDB_DATALEN(db)
+
+	    # Write the compiled help directory structure to the database
+	    # file.  The directory structure consists of the HD structure
+	    # and a string buffer.  We write out the HD structure followed
+	    # by the string buffer.  Each must be aligned to TY_STRUCT.
+	    # The offset of the string buffer from the start of the helpdir
+	    # struct is saved in the HD_NEXTCH field of the HD structure,
+	    # to be used when loaded and referenced to compute a pointer to
+	    # the buffer.
+
+	    HD_NEXTCH(hp) = HD_LENHD(hp)
+	    sv_sbuf = HD_SBUF(hp)
+	    HD_SBUF(hp) = 0
+	    call mii_writei (fd, Memi[hp], HD_LENHD(hp))
+	    HD_SBUF(hp) = sv_sbuf
+	    call mii_writec (fd, Memc[HD_SBUF(hp)], HD_SZSBUF(hp))
+
+	    # Keep track of the amount of struct storage that will be
+	    # required later to hold the UNPACKED helpdir data.
+
+	    #HDB_DATALEN(db) = HDB_DATALEN(db) +
+	    #	HD_LENHD(hp) + ((HD_SZSBUF(hp) + SZ_STRUCT-1) / SZ_STRUCT)
+	    HDB_DATALEN(db) = HDB_DATALEN(db) +
+	    	HD_LENHD(hp) + ((HD_SZSBUF(hp) + SZ_STRUCT32-1) / SZ_STRUCT32)
+
+	    call printf ("%3d %15s (%s): %d help modules\n")
+		call pargi (sp + 1)
+		if (HD_PAKNAME(hp) != NULL)
+		    call pargstr (Memc[HD_SBUF(hp)+HD_PAKNAME(hp)])
+		else
+		    call pargstr ("")
+		call pargstr (fname)
+		call pargi (HD_NMODULES(hp))
+	    call flush (STDOUT)
+
+	    # Now scan the directory for subdirectories.  If one is found,
+	    # open it and push it on the stack, otherwise pop the stack and
+	    # resume scanning the previous directory.
+
+	    repeat {
+		# Search for a module which is a subpackage.
+		found_a_subpackage = false
+		for (;  pk <= HD_NMODULES(hp);  pk=pk+1) {
+		    if (verbose) {
+			modname = M_NAME(HD_MODULE(hp,pk))
+			call printf ("\t\t[%d.%02d] %s\n")
+			    call pargi (sp + 1)
+			    call pargi (pk)
+			    call pargstr (Memc[HD_SBUF(hp) + modname])
+		    }
+		    if (hd_getname (hp, pk, TY_PKG, fname, SZ_FNAME) > 0) {
+			found_a_subpackage = true
+			break
+		    }
+		}
+
+		if (found_a_subpackage) {
+		    if (finfo (fname, fi) == ERR) {
+			call eprintf ("\t\t%4w(cannot access `%s')\n")
+			    call pargstr (fname)
+			# ...and continue searching the current helpdir
+			pk = pk + 1
+
+		    } else {
+			# Got one; push it on the stack.  Bump PK so that we
+			# resume scanning the package with the module following
+			# the subpackage.
+			sp = sp + 1
+			if (sp > MAX_DEPTH)
+			    call fatal (3, "packages nested too deeply")
+			hp_stk[sp] = hp
+			pk_stk[sp] = pk + 1
+			iferr (hp = hd_open (fname)) {
+			    hp = hp_stk[sp]
+			    sp = sp - 1
+			    call eprintf ("cannot open helpdir `%s'\n")
+				call pargstr (fname)
+			    next
+			}
+			pk = 1
+	    		mtime = FI_MTIME(fi)
+			break			# go process new helpdir
+		    }
+
+		} else {
+		    # Helpdir has been exhausted.  Close it and pop the
+		    # stack, continue scanning on the previous helpdir.
+
+		    call hd_close (hp)
+		    if (sp > 0) {
+			hp = hp_stk[sp]
+			pk = pk_stk[sp]
+			sp = sp - 1
+			if (verbose)
+			    call printf ("\t\t\t[end of package]\n")
+
+		    } else if (sp == 0) {
+			# Root helpdir file has been fully expanded.  Scan
+			# all compiled helpdirs and produce a master helpdir
+			# containing an entry for each package in the database.
+			# This is similar to the root helpdir, but contains
+			# entries for packages at all levels, not just at the
+			# root.  Note that we must save and restore the file
+			# position since hdb_make_rhd accesses the file.
+
+			call flush (fd)
+			savepos = note (fd)
+
+			# Load the database into memory.
+			
+			call seek (fd, HDB_DATAOFFSET(db))
+			call calloc (data, 4*HDB_DATALEN(db), TY_STRUCT)
+			call hdb_getdata (fd, data, HDB_DATALEN(db))
+
+			hp = hdb_make_rhd (db, data, index)
+			pk = HD_NMODULES(hp) + 1
+			call strcpy ("_index", fname, SZ_FNAME)
+			mtime = clktime (long (0))
+			sp = -1
+
+			call mfree (data, TY_STRUCT)
+			call seek (fd, savepos)
+			break
+
+		    } else
+			return			# ALL DONE
+		}
+		call flush (STDOUT)
+	    }
+	}
+end
+	    
+
+# HDB_MAKE_RHD -- Make a dummy root help directory for the database.  This
+#   entry looks just like any other compiled help directory, but serves as an
+#   index to all packages in the database.  Each module in root is a package,
+#   and every package in the system, regardless of its level in the package
+#   hierarchy, has an entry in root.  We could also use the database index
+#   for this purpose, but it is keyed by filename not package name, and the
+#   help code considers the root help directory to be conceptually just another
+#   package help directory (the db index is in the wrong format).  The root
+#   directory is equivalent to the compiled lib$root.hd, except that it
+#   contains entries for all subpackages as well.
+#
+# We purposely do NOT sort the package list, because the list is accessed
+#   sequentially when templates are expanded, hence the order determines the
+#   search order for the database.  Since the order in which packages are
+#   entered into the database is determined by the order in which they are
+#   encountered in a depth first search of the package hierarchy, the order of
+#   the packages in the root help directory determines the search order.
+#   It is desirable to search those packages most visible to the user (e.g.,
+#   clpackage) before the more technical packages (e.g., sys).
+#
+# N.B.: This procedure is functionally similar to HD_OPEN and the HD_CLOSE
+# procedure may be used to close the HD structure returned by either.
+
+pointer procedure hdb_make_rhd (db, data, index)
+
+pointer	db			#I database descriptor
+pointer	data			#I data buffer (compiled help directories)
+pointer	index			#I database index
+
+int	i, j, len_modlist, pos
+pointer	hp, o_hp, mp, ix, sbuf, o_mp, c_modlist, hdfile
+
+bool	streq()
+pointer	coerce()
+int	hd_putstr(), strncmp()
+errchk	hdb_putmodule
+
+begin
+	# Allocate and initialize descriptor and string buffer.  Must init
+	# nextch to 1 because 0 is the null index.
+
+	call calloc (hp, LEN_HDSTRUCT, TY_STRUCT)
+	call calloc (sbuf, SZ_SBUF, TY_CHAR)
+
+	HD_SBUF(hp)   = sbuf
+	HD_DEFDIR(hp) = NULL
+	HD_NEXTCH(hp) = 1
+	HD_SZSBUF(hp) = SZ_SBUF
+	HD_LENHD(hp)  = LEN_HDSTRUCT
+	HD_MAXMODULES(hp) = MAX_MODULES
+
+	# The root help directory is the first module.  Since the root is
+	# not a subpackage of any other package we cannot enter it in the
+	# loop below.  We must handcraft this first entry.
+
+	HD_NMODULES(hp) = 1
+	mp = HD_MODULE(hp,1)
+	call aclri (Memi[mp], LEN_MODSTRUCT)
+	M_NAME(mp) = hd_putstr (hp, "_root")
+	M_PKG(mp)  = hd_putstr (hp, DBI_KEY(index))
+
+	# Examine each compiled helpdir for subpackages.  Add each subpackage
+	# found to the current directory.  Do not add the index entries
+	# themselves because they are already referenced in the help
+	# directories.
+
+	for (i=1;  i <= HDB_NENTRIES(db);  i=i+1) {
+	    ix = index + (i - 1) * LEN_HDBINDEX
+	    if (strncmp (DBI_KEY(ix), "_index", 6) == 0)
+		next
+
+	    o_hp = data + DBI_OFFSET(ix)
+	    HD_SBUF(o_hp) = coerce (o_hp + HD_NEXTCH(o_hp), TY_STRUCT, TY_CHAR)
+
+	    for (j=1;  j <= HD_NMODULES(o_hp);  j=j+1) {
+		mp = HD_MODULE(o_hp,j)
+		if (M_PKG(mp) != NULL)
+		    call hdb_putmodule (hp, o_hp, j)
+	    }
+	}
+
+	# Our procedure for building the _index module list has changed the
+	# ordering of the packages from the depth first order of the database
+	# index.  We want the depth first order to be our search order so
+	# we must reorder the module list to agree with the DBI index.
+	# Make a copy of the modlist, then write a new one, overwriting the
+	# old, looking up each package in the index to determine its order
+	# in the new modlist.
+
+	len_modlist = HD_NMODULES(hp) * LEN_MODSTRUCT
+	call calloc (c_modlist, len_modlist, TY_STRUCT)
+	call amovi (Memi[HD_MODULE(hp,1)], Memi[c_modlist], len_modlist)
+
+	pos = 0
+	do j = 1, HDB_NENTRIES(db) {
+	    # Find next valid index entry.
+	    ix = index + (j - 1) * LEN_HDBINDEX
+	    if (strncmp (DBI_KEY(ix), "_index", 6) == 0)
+		next
+
+	    # Locate corresponding helpdir entry, if any.
+	    do i = 1, HD_NMODULES(hp) {
+		o_mp = c_modlist + (i - 1) * LEN_MODSTRUCT
+		hdfile = HD_SBUF(hp) + M_PKG(o_mp)
+		if (Memc[hdfile] == EOS)
+		    next
+		else if (streq (DBI_KEY(ix), Memc[hdfile])) {
+		    # Append entry to output helpdir.
+		    pos = pos + 1
+		    call amovi (Memi[o_mp], Memi[HD_MODULE(hp,pos)],
+			LEN_MODSTRUCT)
+		}
+	    }
+	}
+
+	call mfree (c_modlist, TY_STRUCT)
+	HD_NMODULES(hp) = pos
+
+	# Return any unused space in string buffer.
+	call realloc (HD_SBUF(hp), HD_NEXTCH(hp), TY_CHAR)
+	HD_SZSBUF(hp) = HD_NEXTCH(hp)
+
+	# Return any unused module descriptors.
+        HD_LENHD(hp) = HD_LENHD(hp) -
+	    LEN_MODSTRUCT * (HD_MAXMODULES(hp) - HD_NMODULES(hp))
+	call realloc (hp, HD_LENHD(hp), TY_STRUCT)
+	HD_MAXMODULES(hp) = HD_NMODULES(hp)
+
+	return (hp)
+end
+
+
+# HDB_OPEN -- Open the help database.  We can either read the precompiled
+# help database or the distributed, raw help database (.hd textfiles).
+# If the precompiled database is to be used access will be faster, but
+# packages and modules added since the database was compiled with not be
+# accessible.  To use the precompiled database we read the database index
+# into memory; the actual help files are accessed via this index at runtime.
+# If multiple compiled databases are specified they are combined to form one
+# large database.
+
+pointer procedure hdb_open (database)
+
+char	database[ARB]		#I name of database to be opened
+
+bool	no_entries_interchanged
+pointer	sp, fname, files, hp, db, d_op, i_op, ix, p1, p2, db_save
+int	nfiles, nints, list, fd, d_len, i_len, i, temp[LEN_HDBINDEX]
+
+long	clktime()
+bool	streq(), strgt()
+pointer	hd_open(), hdb_make_rhd()
+int	open(), mii_readi(), mii_readc()
+int	envgets(), access(), fntopnb(), fntgfnb()
+errchk	calloc, realloc, open, seek, syserrs
+errchk	hd_open, fntopnb, fntgfnb, hdb_make_rhd, hdb_getdata
+define	rejectfile_ 91
+define	readerr_  92
+
+begin
+	call smark (sp)
+	call salloc (files, SZ_HELPDB, TY_CHAR)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+	call salloc (hp, LEN_HDBHEADER, TY_STRUCT)
+	call salloc (db_save, LEN_HDBHEADER, TY_STRUCT)
+
+	# Allocate database descriptor.
+	call calloc (db, LEN_HDBHEADER, TY_STRUCT)
+
+	# If the database name is "helpdir", raw access is desired and the
+	# name of the root help directory file is given by the environment
+	# variable "helpdir".  If the database name is "helpdb", the
+	# precompiled database is to be used and the name of the database
+	# file or files is given by the environment variable "helpdb".
+	# Otherwise the database name is assumed to be the name of a raw or
+	# precompiled database file.
+
+	if (streq (database, "helpdir")) {
+	    HDB_RAW(db) = YES
+	    if (envgets ("helpdir", Memc[files], SZ_HELPDB) <= 0)
+		call syserrs (SYS_ENVNF, "helpdir")
+	} else if (streq (database, "helpdb")) {
+	    HDB_RAW(db) = NO
+	    if (envgets ("helpdb", Memc[files], SZ_HELPDB) <= 0)
+		call syserrs (SYS_ENVNF, "helpdb")
+	} else {
+	    HDB_RAW(db) = access (database, 0, TEXT_FILE)
+	    call strcpy (database, Memc[files], SZ_HELPDB)
+	}
+
+	# We now have the filename or file list; if it is a raw help directory
+	# file, open it with HD_OPEN and we are all done for now.  Otherwise
+	# read the helpdb files and construct the help database index.
+
+	if (HDB_RAW(db) == YES) {
+	    iferr (HDB_RHD(db) = hd_open (Memc[files]))
+		call fatal (1, "cannot open root help directory file")
+
+	    call sfree (sp)
+	    return (db)
+	}
+
+	# Allocate and initialize empty database header and index structures.
+	HDB_MAGIC(db)	   = HDB_MAGICVAL
+	HDB_CRDATE(db)     = clktime (long(0))
+
+	HDB_NENTRIES(db)   = 0
+	HDB_MAXENTRIES(db) = MAX_ENTRIES
+	HDB_NMODULES(db)   = 0
+	HDB_DATAPTR(db)    = NULL
+	HDB_DATALEN(db)    = 0
+	HDB_INDEXPTR(db)   = NULL
+	HDB_INDEXLEN(db)   = 0
+	HDB_MAXENTRIES(db) = HDB_NENTRIES(db)
+
+	# Link a binary help database; open each precompiled database and
+	# link it into the full runtime database.
+
+	list = fntopnb (Memc[files], YES)
+	nfiles = 0
+
+	while (fntgfnb (list, Memc[fname], SZ_FNAME) != EOF) {
+	    iferr (fd = open (Memc[fname], READ_ONLY, BINARY_FILE)) {
+		call eprintf ("Cannot open help database file %s\n")
+		    call pargstr (Memc[fname])
+		next
+	    }
+
+	    # Save descriptor in case we cannot read this file.
+	    call amovi (Memi[db], Memi[db_save], LEN_HDBHEADER)
+
+	    # Read the database file header.
+	    nints = LEN_HDBHEADER
+	    if (mii_readi (fd, Memi[hp], nints) < nints) {
+		call eprintf ("Cannot read help database file header (%s)\n")
+		    call pargstr (Memc[fname])
+		goto rejectfile_
+	    }
+
+	    # Verify the file type.
+	    if (HDB_MAGIC(hp) != HDB_MAGICVAL) {
+		call eprintf ("Not a help database file (%s)\n")
+		    call pargstr (Memc[fname])
+		goto rejectfile_
+	    }
+
+	    # Merge the headers.
+	    HDB_NENTRIES(db) = HDB_NENTRIES(db) + HDB_NENTRIES(hp)
+	    HDB_NMODULES(db) = HDB_NMODULES(db) + HDB_NMODULES(hp)
+	    HDB_MAXENTRIES(db) = HDB_NENTRIES(db)
+
+	    d_len = HDB_DATALEN(db)
+	    i_len = HDB_INDEXLEN(db)
+
+	    # Make room for the new data and index entries.
+	    iferr {
+		HDB_DATALEN(db) = HDB_DATALEN(db) + HDB_DATALEN(hp)
+		call realloc (HDB_DATAPTR(db), HDB_DATALEN(db), TY_STRUCT)
+		HDB_INDEXLEN(db) = HDB_INDEXLEN(db) + HDB_INDEXLEN(hp)
+		call realloc (HDB_INDEXPTR(db), HDB_INDEXLEN(db), TY_STRUCT)
+	    } then
+		call erract (EA_WARN)
+
+	    d_op = HDB_DATAPTR(db) + d_len
+	    i_op = HDB_INDEXPTR(db) + i_len
+
+	    # Append the data area of the new database file to the end of
+	    # the data buffer.
+
+	    call seek (fd, HDB_DATAOFFSET(hp))
+	    call hdb_getdata (fd, d_op, HDB_DATALEN(hp))
+
+	    # Append the index area of the new database file to the end of
+	    # the index buffer.
+
+	    nints = HDB_INDEXLEN(hp)
+	    call seek (fd, HDB_INDEXOFFSET(hp))
+
+	    do i = 1, HDB_NENTRIES(hp) {
+		ix = i_op + (i - 1) * LEN_HDBINDEX
+		if (mii_readc (fd, DBI_KEY(ix), SZ_DBIKEY + 1) < SZ_DBIKEY + 1)
+		    goto readerr_
+		if (mii_readi (fd, DBI_OFFSET(ix), LEN_DBIDATA) < LEN_DBIDATA) {
+readerr_	    call eprintf ("Cannot read database index (%s)\n")
+			call pargstr (Memc[fname])
+		    goto rejectfile_
+		}
+
+		# Patch the index entry to reflect the new offset of the
+		# directory entry in the data buffer.  Rename the _index
+		# entries in the old (input) databases, since we will be
+		# building a new _index for the final composite database.
+
+		DBI_OFFSET(ix) = DBI_OFFSET(ix) + d_len
+		if (streq (DBI_KEY(ix), "_index")) {
+		    call sprintf (DBI_KEY(ix), SZ_DBIKEY, "_index.%s")
+			call pargstr (Memc[fname])
+		}
+	    }
+
+	    nfiles = nfiles + 1
+	    call close (fd)
+	    next
+rejectfile_
+	    # Could not read file; restore the descriptor to the state it
+	    # was in before we tried to read the file, to repair any damage.
+
+	    d_op = HDB_DATAPTR(db);  i_op = HDB_INDEXPTR(db)
+	    call amovi (Memi[db_save], Memi[db], LEN_HDBHEADER)
+	    HDB_DATAPTR(db) = d_op;  HDB_INDEXPTR(db) = i_op
+
+	    call close (fd)
+	}
+
+	# Verify that there was at least one valid file in the list.
+	if (nfiles <= 0)
+	    call error (5, "invalid help database file list")
+
+	# Build the package name index (root helpdir) for the new database.
+	hp = hdb_make_rhd (db, HDB_DATAPTR(db), HDB_INDEXPTR(db))
+
+	# Append the compiled package help directory to the database.
+	HDB_NMODULES(db) = HDB_NMODULES(db) + HD_NMODULES(hp)
+	HDB_NENTRIES(db) = HDB_NENTRIES(db) + 1
+
+	# Append the compiled _index helpdir and associated string buffer
+	# to the database data buffer, as if these data structures had been
+	# read from the help database file (all helpdir access codes assume
+	# this structure).
+
+	d_len = HDB_DATALEN(db)
+	nints = HD_LENHD(hp) + (HD_SZSBUF(hp) + SZ_STRUCT32-1) / SZ_STRUCT32
+	HDB_DATALEN(db) = HDB_DATALEN(db) + nints
+	call realloc (HDB_DATAPTR(db), HDB_DATALEN(db), TY_STRUCT)
+	d_op = HDB_DATAPTR(db) + d_len
+
+	HD_NEXTCH(hp) = HD_LENHD(hp)
+	call amovi (Memi[hp], Memi[d_op], HD_LENHD(hp))
+	call amovc (Memc[HD_SBUF(hp)], Memi[d_op+HD_LENHD(hp)], HD_SZSBUF(hp))
+
+	# Add an index entry for the _index helpdir.
+	if (HDB_NENTRIES(db) > HDB_MAXENTRIES(db)) {
+	    HDB_MAXENTRIES(db) = HDB_MAXENTRIES(db) + 1
+	    nints = HDB_MAXENTRIES(db) * LEN_HDBINDEX
+	    iferr (call realloc (HDB_INDEXPTR(db), nints, TY_STRUCT))
+		call fatal (1, "cannot reallocate index buffer")
+	}
+
+	ix = HDB_INDEXPTR(db) + (HDB_NENTRIES(db) - 1) * LEN_HDBINDEX
+	call strcpy ("_index", DBI_KEY(ix), SZ_DBIKEY)
+	DBI_MTIME(ix)  = clktime (long(0))
+	DBI_OFFSET(ix) = d_op - HDB_DATAPTR(db)
+
+	# Free dedicated hp/sbuf, since descriptor is in data buffer now.
+	call hd_close (hp)
+
+	# Sort the index; a crude sort is adequate here.
+	if (HDB_NENTRIES(db) > 1) {
+	    repeat {
+		no_entries_interchanged = true
+		do i = 1, HDB_NENTRIES(db) - 1 {
+		    p1 = HDB_INDEXPTR(db) + (i - 1) * LEN_HDBINDEX
+		    p2 = p1 + LEN_HDBINDEX
+		    if (strgt (DBI_KEY(p1), DBI_KEY(p2))) {
+			call amovi (Memi[p1], temp,     LEN_HDBINDEX)
+			call amovi (Memi[p2], Memi[p1], LEN_HDBINDEX)
+			call amovi (temp,     Memi[p2], LEN_HDBINDEX)
+			no_entries_interchanged = false
+		    }
+		}
+	    } until (no_entries_interchanged)
+	}
+
+	# Get the index offset of the NEW "_index" help directory.
+	HDB_INDEX(db) = 0
+	do i = 1, HDB_NENTRIES(db) {
+	    ix = HDB_INDEXPTR(db) + (i - 1) * LEN_HDBINDEX
+	    if (streq (DBI_KEY(ix), "_index")) {
+		HDB_INDEX(db) = i
+		break
+	    }
+	}
+	if (HDB_INDEX(db) == 0)
+	    call eprintf ("Help warning: cannot find _index")
+
+	call fntclsb (list)
+	call sfree (sp)
+
+	return (db)
+end
+
+
+# HDB_CLOSE -- Close the help database.  If raw database is being accessed,
+# this means close root help directory and free DB struct.  If precompiled,
+# deallocate all buffers used by the database.
+
+procedure hdb_close (db)
+
+pointer	db			# database descriptor
+
+begin
+	if (HDB_RAW(db) == YES)
+	    call hd_close (HDB_RHD(db))
+	else {
+	    call mfree (HDB_DATAPTR(db), TY_STRUCT)
+	    call mfree (HDB_INDEXPTR(db), TY_STRUCT)
+	}
+	call mfree (db, TY_STRUCT)
+end
+
+
+# HDB_LOAD -- Load the named help directory.  Help directories are referred
+# to by the name of the directory file, and the database is keyed by the name
+# of the file.  If the database is being accessed raw, we open and compile
+# the named file directly, otherwise we look up the compiled directory in the
+# database.  In either case we return a HD pointer to the compiled directory.
+# The directory "_index" is special, being the package index directory for the
+# entire database.
+
+pointer procedure hdb_load (db, helpdir)
+
+pointer	db			# database descriptor
+char	helpdir[ARB]		# help directory to be accessed.
+
+bool	index
+pointer	hp, ix, sp, errmsg
+
+bool	streq()
+int	hdb_search()
+pointer	hd_open(), coerce()
+errchk	hdb_open
+
+begin
+	call smark (sp)
+	call salloc (errmsg, SZ_LINE, TY_CHAR)
+
+	index = (streq (helpdir, "_index") || streq (helpdir, "_root"))
+
+	if (HDB_RAW(db) == YES) {
+	    if (index)
+		hp = HDB_RHD(db)
+	    else
+		hp = hd_open (helpdir)
+
+	} else {
+	    # Compute and return pointer to compiled HD.  Fix up pointer to
+	    # the string buffer sbuf, since the pointer value depends on the
+	    # value of the pointer to the newly allocated data area.  If not
+	    # found, return NULL pointer.
+
+	    if (index) {
+		ix = HDB_INDEXPTR(db) + (HDB_INDEX(db) - 1) * LEN_HDBINDEX
+
+	    } else if (hdb_search (db, helpdir, ix) == ERR) {
+		# There should be a better way to do this...  Format error
+		# message and pass to the error handling code, then restore
+		# stack before taking the error action.
+
+		call sprintf (Memc[errmsg], SZ_LINE,
+		"help directory `%s' not found")
+		    call pargstr (helpdir)
+	        iferr (call error (6, Memc[errmsg])) {
+		    call sfree (sp)
+		    call erract (EA_ERROR)
+		    return (NULL)
+		}
+	    }
+
+	    hp = HDB_DATAPTR(db) + DBI_OFFSET(ix)
+	    HD_SBUF(hp) = coerce (hp + HD_NEXTCH(hp), TY_STRUCT, TY_CHAR)
+	}
+
+	call sfree (sp)
+	return (hp)
+end
+
+
+# HDB_FREE -- Free space for a help directory loaded with HDB_LOAD.  If we are
+# using raw access, we let the helpdir package free what ever it wants to since
+# it did the allocating.  If we are accessing the compiled database then there
+# is nothing to free, since everything is maintained in memory.
+
+procedure hdb_free (db, hp)
+
+pointer	db			# database descriptor
+pointer	hp			# help directory
+
+begin
+	if (HDB_RAW(db) == YES)
+	    call hd_close (hp)
+end
+
+
+# HDB_SEARCH -- Search the database index for the given key.  Since the index
+# has been sorted we can use a binary search.  If the key is found we return
+# a pointer to the associated index as an output argument, and OK as the
+# function value.
+
+int procedure hdb_search (db, key, ix)
+
+pointer	db			# database descriptor
+char	key[ARB]		# filename key to be located
+pointer	ix			# pointer to index entry (output)
+
+int	low, high, pos
+pointer	ixoff
+bool	strle(), streq()
+
+begin
+	ixoff = HDB_INDEXPTR(db)
+	low   = 1
+	high  = HDB_NENTRIES(db)
+
+	# Cut range of search in half until range is narrowed to two values (if
+	# we go until HIGH-LOW >= 1 an infinite loop can occur).
+
+	while (high - low > 1) {
+	    pos = (high + low) / 2
+	    if (strle (key, DBI_KEY(ixoff + (pos-1) * LEN_HDBINDEX)))
+		high = pos
+	    else
+		low  = pos
+	} 
+
+	ix = ixoff + (high - 1) * LEN_HDBINDEX
+	if (streq (key, DBI_KEY(ix)))
+	    return (high)
+	ix = ixoff + (low - 1) * LEN_HDBINDEX
+	if (streq (key, DBI_KEY(ix)))
+	    return (low)
+
+	return (ERR)
+end
+
+
+# HDB_EXAMINE -- Examine the structure of the compiled database.  A description
+# of the contents is printed on the output file.
+
+procedure hdb_examine (fd, helpdb, verbose)
+
+int	fd			# output file
+char	helpdb[ARB]		# filename of database to be examined
+bool	verbose			# print menus as well
+
+int	list, i
+long	fi[LEN_FINFO]
+pointer	sp, fname, date, db, ixoff, ix
+
+bool	strne()
+pointer	hdb_open()
+int	finfo(), fntopnb(), fntgfnb()
+errchk	hdb_open, hdb_printpack, fntopnb, fntgfnb
+
+begin
+	call smark (sp)
+	call salloc (date, SZ_DATE, TY_CHAR)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+
+	db = hdb_open (helpdb)
+	ixoff = HDB_INDEXPTR(db)
+
+	call cnvdate (HDB_CRDATE(db), Memc[date], SZ_DATE)
+
+	list = fntopnb (helpdb, YES)
+	while (fntgfnb (list, Memc[fname], SZ_FNAME) != EOF) {
+	    if (finfo (Memc[fname], fi) == ERR) {
+		call eprintf ("Cannot get info on file `%s'\n")
+		    call pargstr (Memc[fname])
+		next
+	    }
+
+	    call fprintf (fd, "Help database %s created %s by %s, size=%d\n")
+		call pargstr (Memc[fname])
+		call pargstr (Memc[date])
+		call pargstr (FI_OWNER(fi))
+		call pargl (FI_SIZE(fi))
+	}
+	call fntclsb (list)
+
+	call fprintf (fd, "Total of %d modules in %d packages\n")
+	    call pargi (HDB_NMODULES(db))
+	    call pargi (HDB_NENTRIES(db) - 1)
+	
+	do i = 1, HDB_NENTRIES(db) {
+	    ix = ixoff + (i - 1) * LEN_HDBINDEX
+	    if (strne (DBI_KEY(ix), "_index"))
+		call hdb_printpack (fd, db, ix, verbose)
+	}
+
+	call hdb_close (db)
+	call sfree (sp)
+end
+
+
+# HDB_PRINTPACK -- Print a description of a single package on the output
+# file.
+
+procedure hdb_printpack (fd, db, ix, verbose)
+
+int	fd			# output file
+pointer	db			# database descriptor
+pointer	ix			# database index descriptor of package
+bool	verbose			# print menus
+
+int	m
+pointer	sp, hp, paknames, date
+long	fi[LEN_FINFO]
+int	hd_getname(), envgeti(), finfo()
+pointer	hdb_load()
+errchk	hd_getname
+
+begin
+	call smark (sp)
+	call salloc (paknames, MAX_MENUSIZE, TY_POINTER)
+	call salloc (date, SZ_DATE, TY_CHAR)
+
+	iferr (hp = hdb_load (db, DBI_KEY(ix))) {
+	    call erract (EA_WARN)
+	    call sfree (sp)
+	    return
+	}
+
+	call cnvdate (DBI_MTIME(ix), Memc[date], SZ_DATE)
+	if (finfo (DBI_KEY(ix), fi) == ERR)
+	    FI_OWNER(fi) = EOS
+
+	if (verbose)
+	    call fprintf (fd, "\n%s %s %s %s\n")
+	else
+	    call fprintf (fd, "%-12s %s %-8s %s\n")
+
+	if (HD_PAKNAME(hp) != 0)
+	    call pargstr (Memc[HD_SBUF(hp) + HD_PAKNAME(hp)])
+	else
+	    call pargstr ("")
+	call pargstr (Memc[date])
+	call pargstr (FI_OWNER(fi))
+	call pargstr (DBI_KEY(ix))
+
+	if (verbose) {
+	    # Extract the names of the modules in the package.  Save the
+	    # pointers in an array for the table print routine.
+
+	    for (m=0;  m < MAX_MENUSIZE;  m=m+1) {
+		call salloc (Memi[paknames+m], MAX_NAMELEN, TY_CHAR)
+		if (hd_getname (hp, m+1, TY_MODNAME, Memc[Memi[paknames+m]],
+		    MAX_NAMELEN) <= 0)
+			break
+	    }
+
+	    # Now print the table.  It is not necessary to sort the table,
+	    # because the "helpdir" code (which reads the help directory) has
+	    # already done so.
+
+	    call strtbl (fd, Memc, Memi[paknames], m, FIRST_COL,
+		envgeti ("ttyncols"), MAX_NAMELEN, 0)
+	}
+
+	call hdb_free (db, hp)
+	call sfree (sp)
+end
+
+
+# HDB_PUTMODULE -- Put a module (subpackage) into the root help directory.
+# Add new entry, increasing space if necessary.  Expand all filenames to
+# remove help ldir references and place filenames in our string buffer,
+# sbuf offsets into module descriptor.  Increase size of sbuf if it fills.
+
+procedure hdb_putmodule (hp, o_hp, pk)
+
+pointer	hp		# new help directory being extended
+pointer	o_hp		# old help directory
+int	pk		# module number in old directory
+
+int	firstch, m
+pointer	sp, fname, sbuf, o_sbuf, mp, o_mp, pakname
+int	hd_getname(), hd_putstr()
+bool	streq()
+
+begin
+	call smark (sp)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+
+	sbuf    = HD_SBUF(hp)
+	o_sbuf  = HD_SBUF(o_hp)
+	o_mp    = HD_MODULE(o_hp,pk)
+	pakname = o_sbuf + M_NAME(o_mp)
+
+	# Check if this is a redefinition of a module already defined.
+	# If so, warn user that new package does not have a unique name,
+	# and omit package.
+
+	firstch = Memc[pakname]
+	for (m=1;  m <= HD_NMODULES(hp);  m=m+1) {
+	    mp = HD_MODULE(hp,m)
+	    if (Memc[sbuf+M_NAME(mp)] == firstch)
+		if (streq (Memc[sbuf+M_NAME(mp)], Memc[pakname])) {
+		    call eprintf ("package name `%s' (hd=%s) is not unique\n")
+			call pargstr (Memc[pakname])
+			call pargstr (Memc[o_sbuf+M_PKG(o_mp)])
+		    call sfree (sp)
+		    return
+		}
+	}
+
+	# If we are out of space for modules, increase the descriptor
+	# structure size to allow more module descriptors.
+
+	if (m > HD_NMODULES(hp)) {
+	    if (m > HD_MAXMODULES(hp)) {
+		HD_LENHD(hp) = HD_LENHD(hp) + (INC_MODULES * LEN_MODSTRUCT)
+		call realloc (hp, HD_LENHD(hp), TY_STRUCT)
+		HD_MAXMODULES(hp) = HD_MAXMODULES(hp) + INC_MODULES
+	    }
+	    HD_NMODULES(hp) = m
+	}
+
+	mp = HD_MODULE(hp,m)
+	call aclri (Memi[mp], LEN_MODSTRUCT)
+
+	# Put module name in string buffer and save index of string in module
+	# descriptor.
+
+	M_NAME(mp) = hd_putstr (hp, Memc[pakname])
+
+	# Extract all filenames and move into string buffer.  Call hd_getname to
+	# extract filenames from old directory, so that help-ldir references
+	# are expanded.
+
+	if (hd_getname (o_hp, pk, TY_HLP, Memc[fname], SZ_FNAME) > 0)
+	    M_HLP(mp) = hd_putstr (hp, Memc[fname])
+	if (hd_getname (o_hp, pk, TY_SYS, Memc[fname], SZ_FNAME) > 0)
+	    M_SYS(mp) = hd_putstr (hp, Memc[fname])
+	if (hd_getname (o_hp, pk, TY_SRC, Memc[fname], SZ_FNAME) > 0)
+	    M_SRC(mp) = hd_putstr (hp, Memc[fname])
+	if (hd_getname (o_hp, pk, TY_PKG, Memc[fname], SZ_FNAME) > 0)
+	    M_PKG(mp) = hd_putstr (hp, Memc[fname])
+	if (hd_getname (o_hp, pk, TY_MEN, Memc[fname], SZ_FNAME) > 0)
+	    M_MEN(mp) = hd_putstr (hp, Memc[fname])
+
+	call sfree (sp)
+end
+
+
+# HDB_GETDATA -- Read a stored series of compiled help directories, stored
+# externally in a machine independent format, into the given data buffer.
+# Each stored help directory consists of a fixed sized MII-32 header followed
+# by a MII byte packed string buffer of arbitrary length.  Reading begins
+# at the current file position.
+
+procedure hdb_getdata (fd, obuf, buflen)
+
+int	fd			#I input file
+pointer	obuf			#O receives unpacked helpdir data
+int	buflen			#O max su out
+
+int	i, nelem, nr, sz_mii_struct
+pointer	op, hp
+int	mii_readi(), mii_readc()
+errchk	mii_readi, mii_readc
+define	readerr_ 91
+
+begin
+	nr = 0
+	for (op=obuf;  nr < buflen;  ) {
+	    hp = op
+
+	    # Get fixed size helpdir header.
+	    if (mii_readi (fd, Memi[op], LEN_BASEHD) < LEN_BASEHD)
+		goto readerr_
+	    nr = nr + LEN_BASEHD
+
+	    # Get module entries.
+	    op = op + LEN_BASEHD
+	    nelem = HD_LENHD(hp) - LEN_BASEHD
+	    if (mii_readi (fd, Memi[op], nelem) < nelem)
+		goto readerr_
+	    nr = nr + nelem
+
+	    # Get string buffer.
+	    op = op + nelem
+	    nelem = HD_SZSBUF(hp) # / (SZ_INT / SZ_INT32)
+	    if (mii_readc (fd, Memi[op], nelem) < nelem)
+		goto readerr_
+
+	    nr = nr + ((nelem + SZ_STRUCT32-1) / SZ_STRUCT32)
+            sz_mii_struct = MII_INT / NBITS_BYTE / SZB_CHAR
+            op = op + ((nelem + sz_mii_struct-1) / sz_mii_struct)
+	}
+
+	return
+
+readerr_
+	# Common read error code.
+	call error (1, "Cannot read help database data\n")
+end
diff --git a/pkg/system/help/helpdir.h b/pkg/system/help/helpdir.h
new file mode 100644
index 00000000..89b851a7
--- /dev/null
+++ b/pkg/system/help/helpdir.h
@@ -0,0 +1,34 @@
+# Size limiting definitions.
+
+define	MAX_LDIRS	50			# maximum ldirs for package
+define	MAX_MODULES	100			# initial maximum no. modules
+define	INC_MODULES	50			# increment if overflow
+define	SZ_SBUF		2048			# initial size string buffer
+define	INC_SZSBUF	1024			# increment if overflow
+
+define	LEN_HDSTRUCT	(10+MAX_LDIRS+MAX_MODULES*LEN_MODSTRUCT)
+define	LEN_BASEHD	10
+define	LEN_MODSTRUCT	6
+
+# Helpdir descriptor structure.
+
+define	HD_SBUF		Memi[$1]		# string buffer
+define	HD_NEXTCH	Memi[$1+1]		# index of next char in sbuf
+define	HD_SZSBUF	Memi[$1+2]		# size of string buffer
+define	HD_DEFDIR	Memi[$1+3]		# offset to defdir
+define	HD_NLDIRS	Memi[$1+4]		# number of logical directories
+define	HD_NMODULES	Memi[$1+5]		# number of modules in package
+define	HD_MAXMODULES	Memi[$1+6]		# max no. of modules in package
+define	HD_LENHD	Memi[$1+7]		# length of this structure
+define	HD_PAKNAME	Memi[$1+8]		# offset of package name string
+define	HD_LDIR		Memi[$1+10+$2-1]	# indices of ldir strings
+define	HD_MODULE	($1+60+($2-1)*6)	# module descriptors
+
+# Module descriptor structure.
+
+define	M_NAME		Memi[$1]		# module name index
+define	M_HLP		Memi[$1+1]		# help file index
+define	M_SYS		Memi[$1+2]		# system docs file index
+define	M_SRC		Memi[$1+3]		# source code file index
+define	M_PKG		Memi[$1+4]		# pkg helpdir file index
+define	M_MEN		Memi[$1+5]		# package menu file index
diff --git a/pkg/system/help/helpdir.x b/pkg/system/help/helpdir.x
new file mode 100644
index 00000000..38a2e12c
--- /dev/null
+++ b/pkg/system/help/helpdir.x
@@ -0,0 +1,775 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	<error.h>
+include	"help.h"
+include	"helpdir.h"
+
+.help helpdir
+.nf ___________________________________________________________________________
+HELPDIR -- Routines for accessing helpdir files.  A helpdir lists the
+modules in a package, as well as the files containing the help text and
+source for each module.  There are two kinds of helpdir files: the 'packhelp'
+file lists the packages, and each package has its own helpdir file listing
+the modules within that package.  The following functions are required to
+access help module lists:
+
+	hp =        hd_open (fname)
+		   hd_close (hp)
+
+	modnum = hd_findmod (hp, modname)
+	nchars = hd_getname (hp, modnum, field, outstr, maxch)
+		   hd_debug (hp, out)
+
+The HD_OPEN function opens the helpdir file and decodes the contents, producing
+a binary structure pointed to by hp.  HD_FINDMOD searches for a module by name,
+returning the module number within the helpdir.  HD_GETNAME fetches either
+the module name (field "mod") or a file name (fields "hlp", "sys", etc.).
+File names are returned as OS pathnames, with logical directory expansion taking
+place within the helpdir package.
+
+Helpdir files have the following structure:
+
+	$defdir = pathname|ldir
+	$ldir1  = pathname|ldir
+	$ldir2  = pathname|ldir
+			...
+	$ldirN  = pathname|ldir
+
+	module1 hlp=file, sys=file, src=file, pkg=file, men=file
+	module2 hlp=file, etc.
+			...
+	moduleN hlp=file, etc.
+
+The dollar signs are required to make it easy to distinguish logical directory
+declarations from module entries.  Logical directories defined local to the
+help directory file are not expanded recursively.  If the last nonwhite char
+on a line is a comma, the file list is assumed to be continued on the next
+line.  If two or more files in a module list are the same, all but the first
+may be set to ".." and the last file name given will be used.  Quotes are
+optional.  All file assignments are optional, and they may occur in any order.
+.endhelp _______________________________________________________________________
+
+
+# HD_OPEN -- Open the helpdir file, allocate descriptor and decode file
+# into descriptor.  Sort module list when done, unless sorting is disabled
+# by inclusion of the directive ".nosort" in the text. Ignore comment lines and
+# blank lines.  Ldir declarations and module entries may be given in any
+# order.
+
+pointer procedure hd_open (helpdir_file)
+
+char	helpdir_file[ARB]
+bool	sort_modules
+pointer	sp, hp, ip, lbuf, defdir, word
+int	fd, junk
+
+bool	streq()
+int	open(), getline(), fnroot(), fnldir(), ctowrd()
+int	hd_putstr()
+errchk	salloc, open, calloc, getline, hd_putstr
+errchk	hd_putldiry, hd_putmodule, hd_sort_modules
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+	call salloc (defdir, SZ_PATHNAME, TY_CHAR)
+	call salloc (word, SZ_FNAME, TY_CHAR)
+
+	# If helpdir file is not yet installed, print warning message
+	# and return the NULL pointer, indicating that the help directory
+	# could not be loaded.
+
+	iferr (fd = open (helpdir_file, READ_ONLY, TEXT_FILE)) {
+	    call sfree (sp)
+	    call erract (EA_ERROR)
+	    return (NULL)
+	}
+
+	# Allocate and initialize descriptor and string buffer.  Must init
+	# nextch to 1 because 0 is the null index.
+
+	call calloc (hp, LEN_HDSTRUCT, TY_STRUCT)
+	call calloc (HD_SBUF(hp), SZ_SBUF, TY_CHAR)
+	HD_DEFDIR(hp) = NULL
+	HD_NEXTCH(hp) = 1
+	HD_SZSBUF(hp) = SZ_SBUF
+	HD_MAXMODULES(hp) = MAX_MODULES
+	HD_LENHD(hp) = LEN_HDSTRUCT
+	sort_modules = true
+
+	# Extract package name into string buffer.  The package name is assumed
+	# to be the root of the filename.
+
+	junk = fnroot (helpdir_file, Memc[lbuf], SZ_LINE)
+	HD_PAKNAME(hp) = hd_putstr (hp, Memc[lbuf])
+
+	# Extract directory prefix of the package help directory file.  Make
+	# it the default directory prefix for all filenames in the directory.
+
+	if (fnldir (helpdir_file, Memc[defdir], SZ_PATHNAME) == 0) {
+	    call fpathname (helpdir_file, Memc[lbuf], SZ_LINE)
+	    junk = fnldir (Memc[lbuf], Memc[defdir], SZ_PATHNAME)
+	}
+	call sprintf (Memc[lbuf], SZ_LINE, "defdir = %s\n")
+	    call pargstr (Memc[defdir])
+	call hd_putldiry (hp, Memc[lbuf])
+
+	# Compile the file, processing all logical directory definitions,
+	# set option statments, and module declarations. Ignore blank lines
+	# and comment lines.
+
+	while (getline (fd, Memc[lbuf]) != EOF) {
+	    for (ip=lbuf;  IS_WHITE (Memc[ip]);  ip=ip+1)
+		;
+	    if (Memc[ip] == '#' || Memc[ip] == '\n' || Memc[ip] == EOS)
+		next
+
+	    # A leading dollar sign denotes a logical directory declaration.
+	    # A leading period denotes a compiler directive.
+	    # Anything else is a module entry.
+
+	    if (Memc[ip] == '$') {
+		call hd_putldiry (hp, Memc[ip+1])
+	    } else if (Memc[ip] == '.') {
+		ip = ip + 1
+		if (ctowrd (Memc, ip, Memc[word], SZ_FNAME) <= 0)
+		    next
+		# The directive ".nosort" disables sorting of the module list.
+		# There are no other such directives at present, but they are
+		# easy to add.
+		if (streq (Memc[word], "nosort"))
+		    sort_modules = false
+		else {
+		    call eprintf ("Warning: unknown directive %s in helpdir\n")
+			call pargstr (Memc[word])
+		}
+		next
+	    } else
+		call hd_putmodule (hp, fd, Memc[lbuf])
+	}
+
+	# We are all done with the helpdir file, so close it.  Sort module
+	# list alphabetically by name.
+
+	call close (fd)
+	if (sort_modules && HD_NMODULES(hp) > 1)
+	    call hd_sort_modules (hp)
+
+	# Return any unused space in string buffer.
+	call realloc (HD_SBUF(hp), HD_NEXTCH(hp), TY_CHAR)
+	HD_SZSBUF(hp) = HD_NEXTCH(hp)
+
+	# Return any unused module descriptors.
+	HD_LENHD(hp) = HD_LENHD(hp) -
+	    LEN_MODSTRUCT * (HD_MAXMODULES(hp) - HD_NMODULES(hp))
+	HD_MAXMODULES(hp) = HD_NMODULES(hp)
+	call realloc (hp, HD_LENHD(hp), TY_STRUCT)
+
+	call sfree (sp)
+	return (hp)
+end
+
+
+# HD_CLOSE -- Close the helpdir descriptor.  The helpdir file has already
+# been closed; all we need do is return the string buffer and the helpdir
+# descriptor structure.
+
+procedure hd_close (hp)
+
+pointer	hp
+
+begin
+	if (hp != NULL) {
+	    call mfree (HD_SBUF(hp), TY_CHAR)
+	    call mfree (hp, TY_STRUCT)
+	}
+end
+
+
+# HD_PUTLDIRY -- Decode a logical directory declaration and store it away
+# in the descriptor.  We are passed the declaration minus the leading
+# dollar sign.  Format: "variable = string".
+
+procedure hd_putldiry (hp, decl)
+
+pointer	hp
+char	decl[ARB]
+
+int	ip, nldir, strp
+pointer	sp, buf, op
+char	hd_getc()
+int	strncmp(), hd_putstr()
+errchk	salloc, hd_getc, hd_putstr
+
+begin
+	call smark (sp)
+	call salloc (buf, SZ_LINE, TY_CHAR)
+
+	# Do nothing if null declaration.
+	for (ip=1;  IS_WHITE(decl[ip]);  ip=ip+1)
+	    ;
+	if (decl[ip] == '\n' || decl[ip] == EOS || decl[ip] == '#') {
+	    call sfree (sp)
+	    return
+	}
+
+	# Extract "ldir=string", eliminating all whitespace, and deleting
+	# the newline at the end of the statement.  Quotes around the string
+	# are optional and are deleted in hl_getc.
+
+	for (op=buf;  hd_getc (decl, ip, Memc[op]) != EOS;  op=op+1)
+	    if (Memc[op] == '\n')
+		op = op - 1
+
+	# Deposit the "ldir=string" in the string buffer, and set the
+	# appropriate pointers for either defdir or a new ldir.
+
+	strp = hd_putstr (hp, Memc[buf])
+	if (strncmp (Memc[buf], "defdir", 6) == 0)
+	    HD_DEFDIR(hp) = strp
+	else {
+	    nldir = HD_NLDIRS(hp) + 1
+	    if (nldir > MAX_LDIRS)
+		call error (8, "Too many ldir declarations in helpdir")
+	    HD_NLDIRS(hp) = nldir
+	    HD_LDIR(hp,nldir) = strp
+	}
+
+	call sfree (sp)
+end
+
+
+# HD_PUTMODULE -- Put a module declaration into the helpdir descriptor.
+# Start new entry; save module name; process file names.  If line ends
+# in a comma, get a new line.  Print warning if an unknown file type
+# keyword is encountered.
+
+procedure hd_putmodule (hp, fd, lbuf)
+
+pointer	hp
+int	fd
+char	lbuf[ARB]
+
+char	ch
+int	ip, junk, m, ftype, strp
+pointer	sp, buf, op, sbuf, mp
+
+bool	streq()
+char	hd_getc(), hd_peekc()
+int	getline(), ctowrd(), hd_putstr()
+errchk	salloc, getline, hd_putstr, hd_getc
+
+begin
+	call smark (sp)
+	call salloc (buf, SZ_LINE, TY_CHAR)
+	sbuf = HD_SBUF(hp)
+
+	# Fetch module name.  Cannot be null or line is blank and we would
+	# not have been called.
+	ip = 1
+	junk = ctowrd (lbuf, ip, Memc[buf], SZ_LINE)
+
+	# Check if this is a redefinition of a module already defined.
+	# If so, overwrite descriptor of earlier module, else allocate
+	# a new descriptor.
+
+	for (m=1;  m <= HD_NMODULES(hp);  m=m+1) {
+	    mp = HD_MODULE(hp,m)
+	    if (streq (Memc[buf], Memc[sbuf+M_NAME(mp)]))
+		break
+	}
+
+	if (m > HD_NMODULES(hp)) {
+	    # If we are out of space for modules, increase the descriptor
+	    # structure size to allow more module descriptors.
+	    if (m > HD_MAXMODULES(hp)) {
+		HD_LENHD(hp) = HD_LENHD(hp) + (INC_MODULES * LEN_MODSTRUCT)
+		call realloc (hp, HD_LENHD(hp), TY_STRUCT)
+		HD_MAXMODULES(hp) = HD_MAXMODULES(hp) + INC_MODULES
+	    }
+	    HD_NMODULES(hp) = m
+	}
+
+	mp = HD_MODULE(hp,m)
+	call aclri (Memi[mp], LEN_MODSTRUCT)
+
+	# Put module name in string buffer and save index of string in descr.
+	M_NAME(mp) = hd_putstr (hp, Memc[buf])
+
+	# Process file name fields, if any.  Unrecognized file type keywords
+	# cause a warning to be issued.  Redundant entires overwrite old
+	# entries.  Order makes no diff, absence is ok.
+
+	op = buf
+	ftype = TY_UNKNOWN
+	strp = 0
+
+	repeat {
+	    ch = hd_getc (lbuf, ip, ch)
+
+	    switch (ch) {
+	    case '=':
+		# Buffer contains the code word for the file being set.
+		Memc[op] = EOS
+		call strlwr (Memc[buf])
+		if (     streq (Memc[buf], "hlp"))
+		    ftype = TY_HLP
+		else if (streq (Memc[buf], "sys"))
+		    ftype = TY_SYS
+		else if (streq (Memc[buf], "src"))
+		    ftype = TY_SRC
+		else if (streq (Memc[buf], "pkg"))
+		    ftype = TY_PKG
+		else if (streq (Memc[buf], "men"))
+		    ftype = TY_MEN
+		else {
+		    ftype = TY_UNKNOWN
+		    call eprintf ("Warning: bad file type `%s' in helpdir\n")
+			call pargstr (Memc[buf])
+		}
+		op = buf
+
+	    case ',', '\n', EOS:
+		# Buffer contains the file name string.  Put it in the string
+		# buffer and save pointer in appropriate field of module
+		# descriptor.
+
+		Memc[op] = EOS
+
+		# If filename is "..", i.e., "sys=..", the filename is identical
+		# to that last specified.  Use the file name pointer.
+
+		if (streq (Memc[buf], "..")) {
+		    if (strp == 0)
+			call error (9, "helpdir: `..' reference, no prev file")
+		} else
+		    strp = hd_putstr (hp, Memc[buf])
+
+		switch (ftype) {
+		case TY_HLP:
+		    M_HLP(mp) = strp
+		case TY_SYS:
+		    M_SYS(mp) = strp
+		case TY_SRC:
+		    M_SRC(mp) = strp
+		case TY_PKG:
+		    M_PKG(mp) = strp
+		case TY_MEN:
+		    M_MEN(mp) = strp
+		}
+		op = buf
+
+		if (ch == '\n' || ch == EOS)
+		    break				# end of statement
+		else {
+		    # Check for continuation on next line.  Read a new line into
+		    # the buffer if end of line follows comma.
+
+		    if (hd_peekc(lbuf,ip) == '\n' || hd_peekc(lbuf,ip) == EOS) {
+			if (getline (fd, lbuf) == EOF) {
+			    call eprintf ("Unexpected EOF in helpdir file\n")
+			    call sfree (sp)
+			    return
+			}
+			ip = 1
+		    }
+		}
+
+	    default:
+		# Regular character.  Deposit in buffer.
+		if (op >= buf+SZ_LINE)
+		    call error (10, "helpdir: buffer overflow reading modspec")
+		Memc[op] = ch
+		op = op + 1
+	    }
+	}
+
+	call sfree (sp)
+end
+
+
+# HD_GETC -- Get next nonwhite character from the input line.  Leave pointer
+# pointing to next character.  Ignore quotes, so that file name strings can
+# be quoted without penalty.
+
+char procedure hd_getc (lbuf, ip, ch)
+
+char	lbuf[ARB]
+int	ip
+char	ch
+char	hd_peekc()
+
+begin
+	ch = hd_peekc (lbuf, ip)
+	if (ch != EOS)
+	    ip = ip + 1
+	
+	return (ch)
+end
+
+
+# HD_PEEKC -- Peek at the next nonwhite character on a line, but do not
+# advance the input pointer past the character.  Ignore quote characters
+# and comments.
+
+char procedure hd_peekc (lbuf, ip)
+
+char	lbuf[ARB]
+int	ip
+char	ch
+
+begin
+	for (ch=lbuf[ip];  ch != EOS;  ch=lbuf[ip])
+	    if (IS_WHITE(ch) || ch == '\'' || ch == '"') {
+		ip = ip + 1
+	    } else if (ch == '#') {
+		while (lbuf[ip] != '\n' && lbuf[ip] != EOS)
+		    ip = ip + 1
+	    } else
+		break
+	
+	return (ch)
+end
+
+
+# HD_PUTSTR -- Put a string (incl EOS) in the string buffer at nextch.
+# If there is not enough space in the buffer, reallocate a larger buffer.
+# Return the index of the string in the string buffer.
+
+int procedure hd_putstr (hp, str)
+
+pointer	hp
+char	str[ARB]
+int	nextch, nchars, strlen()
+errchk	realloc
+
+begin
+	# Null strings are not stored and cause a null index to be returned.
+	nchars = strlen (str)
+	if (nchars == 0)
+	    return (0)
+
+	nextch = HD_NEXTCH(hp)
+	if (nextch + nchars + 1 > HD_SZSBUF(hp)) {
+	    HD_SZSBUF(hp) = HD_SZSBUF(hp) + INC_SZSBUF
+	    call realloc (HD_SBUF(hp), HD_SZSBUF(hp), TY_CHAR)
+	}
+
+	call strcpy (str, Memc[HD_SBUF(hp) + nextch], ARB)
+	HD_NEXTCH(hp) = nextch + nchars + 1
+
+	return (nextch)
+end
+
+
+# HD_SORT_MODULES -- Sort the module list alphabetically by name.
+# A simple exchange sort is ok because the sort time is negligible
+# compared to all the file accesses, Lroff etc.
+
+procedure hd_sort_modules (hp)
+
+pointer	hp
+
+bool	sorted
+int	nmodules, m, mlen, i, temp
+pointer	sbuf, mp1, mp2
+bool	strgt()
+
+begin
+	nmodules = HD_NMODULES(hp)
+	sbuf = HD_SBUF(hp)
+	mlen = LEN_MODSTRUCT
+	if (nmodules < 2)
+	    return
+
+	repeat {
+	    sorted = true
+	    do m = 1, nmodules-1 {
+		mp1 = HD_MODULE(hp,m)
+		mp2 = mp1 + mlen
+		if (strgt (Memc[sbuf+M_NAME(mp1)], Memc[sbuf+M_NAME(mp2)])) {
+		    do i = 0, mlen-1 {
+			temp = Memi[mp1+i]
+			Memi[mp1+i] = Memi[mp2+i]
+			Memi[mp2+i] = temp
+		    }
+		    sorted = false
+		}
+	    }
+	} until (sorted)
+end
+
+
+# HD_FINDMOD -- Search for the named module and return the module number
+# if found.  Abbreviations are permitted.  An ambiguous abbreviation is
+# an error.
+
+int procedure hd_findmod (hp, modname)
+
+pointer	hp
+char	modname[ARB]
+int	m, namelen, module
+pointer	mp, sbuf
+int	strlen(), strncmp()
+
+begin
+	namelen = strlen (modname)
+	if (namelen == 0)
+	    return (0)
+	module = 0
+	sbuf = HD_SBUF(hp)
+
+	for (m=1;  m <= HD_NMODULES(hp);  m=m+1) {
+	    mp = HD_MODULE(hp,m)
+
+	    if (strncmp (Memc[sbuf+M_NAME(mp)], modname, namelen) == 0) {
+		if (strlen (Memc[sbuf+M_NAME(mp)]) == namelen) {
+		    return (m)				# exact match
+		} else if (module != 0) {
+		    call eprintf ("\n--> %s <--\n")
+			call pargstr (modname)
+		    call error (11, "Ambiguous module name abbreviation")
+		} else
+		    module = m
+	    }
+	}
+
+	return (module)
+end
+
+
+# HD_GETNAME -- Get the module name or a filename.  If a filename is requested,
+# check if filename contains a logical directory reference.  If yes, try to
+# satisfy the reference from the local list of ldirs (but no recursive refs
+# permitted here).  If no, prepend the defdir string.
+
+define	nullstr_	91
+define	nodefdir_	92
+
+int procedure hd_getname (hp, m, field, outstr, maxch)
+
+pointer	hp
+int	m				# module number
+int	field				# field code
+char	outstr[ARB]
+int	maxch
+
+int	len_ldir, op
+pointer	mp, sp, ldir, sbuf, fname_ptr, ip, subdir
+int	strncmp(), gstrcpy(), hd_getldir(), fnldir()
+errchk	salloc, hd_getldir
+
+begin
+	call smark (sp)
+	call salloc (ldir, SZ_PATHNAME, TY_CHAR)
+	call salloc (subdir, SZ_FNAME, TY_CHAR)
+
+	if (hp == NULL)
+	    call error (12, "hd_getname: bad helpdir descriptor")
+
+	if (m < 1 || m > HD_NMODULES(hp)) {
+nullstr_    call sfree (sp)
+	    outstr[1] = EOS
+	    return (0)
+	}
+
+	mp = HD_MODULE(hp,m)
+	sbuf = HD_SBUF(hp)
+
+	switch (field) {
+	case TY_MODNAME:
+	    call sfree (sp)
+	    return (gstrcpy (Memc[sbuf+M_NAME(mp)], outstr, maxch))
+	case TY_HLP:
+	    fname_ptr = M_HLP(mp)
+	case TY_SYS:
+	    fname_ptr = M_SYS(mp)
+	case TY_SRC:
+	    fname_ptr = M_SRC(mp)
+	case TY_PKG:
+	    fname_ptr = M_PKG(mp)
+	case TY_MEN:
+	    fname_ptr = M_MEN(mp)
+	default:
+	    goto nullstr_
+	}
+
+	# If index is zero, no filename was given.
+	if (fname_ptr == 0)
+	    goto nullstr_
+
+	# Get ldir substring, if any, from filename.  If no ldir, prepend
+	# defdir and quit.  Otherwise lookup ldir in local list.  If found,
+	# prepend value.  Otherwise the ldir is a CL global one, and return
+	# filename without modification.  If the given ldir string begins
+	# with "./", substitute the value of defdir for the ".".
+
+	len_ldir = fnldir (Memc[sbuf+fname_ptr], Memc[ldir], SZ_PATHNAME)
+
+	if (len_ldir == 0) {
+	    if (HD_DEFDIR(hp) == 0)
+		goto nodefdir_
+	    for (ip = sbuf + HD_DEFDIR(hp);  Memc[ip] != '=';  ip=ip+1)
+		;
+	    op = gstrcpy (Memc[ip+1], outstr, maxch) + 1
+	    ip = sbuf + fname_ptr
+	} else if (hd_getldir (hp, Memc[ldir], Memc[subdir], SZ_FNAME) == 0) {
+	    op = 1
+	    ip = sbuf + fname_ptr
+	} else {
+	    if (strncmp (Memc[subdir], "./", 2) == 0) {
+		if (HD_DEFDIR(hp) == 0)
+nodefdir_	    call error (13, "Default directory omitted in helpdir")
+		for (ip = sbuf + HD_DEFDIR(hp);  Memc[ip] != '=';  ip=ip+1)
+		    ;
+		op = gstrcpy (Memc[ip+1], outstr, maxch) + 1
+		ip = subdir + 2
+	    } else {
+		op = 1
+		ip = subdir
+	    }
+	    op = op + gstrcpy (Memc[ip], outstr[op], maxch - op + 1)
+	    ip = sbuf + fname_ptr + len_ldir
+	}
+
+	call sfree (sp)
+	return (gstrcpy (Memc[ip], outstr[op], maxch - op + 1))
+end
+
+
+# HD_GETLDIR -- Search the logical directory list for the package helpdir
+# file for the given ldir name.  Return value in output string if found.
+
+int procedure hd_getldir (hp, ldir, outstr, maxch)
+
+pointer	hp
+char	ldir[ARB]
+char	outstr[ARB]
+int	maxch
+
+int	i
+pointer	sp, ip, op, sbuf, envvar, filvar
+bool	streq()
+int	gstrcpy()
+
+begin
+	call smark (sp)
+	call salloc (envvar, SZ_FNAME, TY_CHAR)
+	call salloc (filvar, SZ_FNAME, TY_CHAR)
+
+	sbuf = HD_SBUF(hp)
+
+	# Ldir string has the form "ldir$".  Extract the ldir name into the
+	# filvar buffer, omitting the $ delimiter.
+	i = 1
+	for (op=filvar;  ldir[i] != '$' && ldir[i] != EOS;  op=op+1) {
+	    Memc[op] = ldir[i]
+	    i = i + 1
+	}
+	Memc[op] = EOS
+
+	for (i=1;  i <= HD_NLDIRS(hp);  i=i+1) {
+	    # Extract the helpdir env variable name into the envvar buffer,
+	    # stopping when the '=' is reached.
+
+	    ip = sbuf + HD_LDIR(hp,i)
+	    for (op=envvar;  Memc[ip] != '=';  op=op+1) {
+		if (Memc[ip] == EOS)
+		    call error (14, "helpdir: missing '=' in ldir declaration")
+		Memc[op] = Memc[ip]
+		ip = ip + 1
+	    }
+	    Memc[op] = EOS
+
+	    # Return whatever follows the '=' if we have a match.  The input
+	    # pointer is left pointing at the '='.
+
+	    if (streq (Memc[filvar], Memc[envvar])) {
+		call sfree (sp)
+		return (gstrcpy (Memc[ip+1], outstr, maxch))
+	    }
+	}
+
+	outstr[1] = EOS
+	call sfree (sp)
+	return (0)
+end
+
+
+# HD_DEBUG -- Dump a compiled helpdir structure to the given output file.
+
+procedure hd_debug (hd, out)
+
+pointer	hd				#I compiled helpdir
+int	out				#O output file
+
+int	i
+pointer	sbuf, mp
+
+begin
+	sbuf = HD_SBUF(hd)
+
+	# General header stuff.
+	call fprintf (out,
+	    "pakname=%d(%s), nmodules=%d, maxmodules=%d, lenhd=%d\n")
+	    call pargi (HD_PAKNAME(hd))
+	    call pargstr (Memc[sbuf+HD_PAKNAME(hd)])
+	    call pargi (HD_NMODULES(hd))
+	    call pargi (HD_MAXMODULES(hd))
+	    call pargi (HD_LENHD(hd))
+	call fprintf (out, "defdir=%d(%s)\n")
+	    call pargi (HD_DEFDIR(hd))
+	    call pargstr (Memc[sbuf+HD_DEFDIR(hd)])
+	call fprintf (out, "sbuf=%x, szsbuf=%d, nextch=%d\n")
+	    call pargi (HD_SBUF(hd))
+	    call pargi (HD_SZSBUF(hd))
+	    call pargi (HD_NEXTCH(hd))
+
+	# List of defined logical directories.
+	if (HD_NLDIRS(hd) > 0) {
+	    call fprintf (out, "nldirs=%d:\n")
+		call pargi (HD_NLDIRS(hd))
+	    do i = 1, HD_NLDIRS(hd) {
+		call fprintf (out, "%7d %s\n")
+		    call pargi (i)
+		    call pargstr (Memc[sbuf+HD_LDIR(hd,i)])
+	    }
+	}
+
+	# List the modules.
+	if (HD_NMODULES(hd) > 0) {
+	    call fprintf (out, "modules:\n")
+	    do i = 1, HD_NMODULES(hd) {
+		mp = HD_MODULE(hd,i)
+		call fprintf (out, "%7d %20s\n")
+		    call pargi (i)
+		    call pargstr (Memc[sbuf+M_NAME(mp)])
+		if (M_HLP(mp) > 0) {
+		    call fprintf (out, "\t\t\t\tHLP=%d(%s)\n")
+			call pargi (M_HLP(mp))
+			call pargstr (Memc[sbuf+M_HLP(mp)])
+		}
+		if (M_SYS(mp) > 0) {
+		    call fprintf (out, "\t\t\t\tSYS=%d(%s)\n")
+			call pargi (M_SYS(mp))
+			call pargstr (Memc[sbuf+M_SYS(mp)])
+		}
+		if (M_SRC(mp) > 0) {
+		    call fprintf (out, "\t\t\t\tSRC=%d(%s)\n")
+			call pargi (M_SRC(mp))
+			call pargstr (Memc[sbuf+M_SRC(mp)])
+		}
+		if (M_PKG(mp) > 0) {
+		    call fprintf (out, "\t\t\t\tPKG=%d(%s)\n")
+			call pargi (M_PKG(mp))
+			call pargstr (Memc[sbuf+M_PKG(mp)])
+		}
+		if (M_MEN(mp) > 0) {
+		    call fprintf (out, "\t\t\t\tMEN=%d(%s)\n")
+			call pargi (M_MEN(mp))
+			call pargstr (Memc[sbuf+M_MEN(mp)])
+		}
+	    }
+	}
+end
diff --git a/pkg/system/help/hinput.x b/pkg/system/help/hinput.x
new file mode 100644
index 00000000..c47716e8
--- /dev/null
+++ b/pkg/system/help/hinput.x
@@ -0,0 +1,274 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	"help.h"
+
+.help hinput
+.nf __________________________________________________________________________
+HINPUT -- Lroff line input procedure.  Called by Lroff to get lines of
+input text.  Function varies slightly depending on the Help option.
+If printing only single param or single section, our job is to eat all
+input which is not part of the indicated section of the help block.
+A parameter block begins with ".ls paramname" and ends with a matching
+".le" or ".ih", if the text is formatted.  Similarly, the single section
+begins with a ".ih" followed by the section name on the next line.
+.endhelp ______________________________________________________________________
+
+define	SEARCHING	1
+define	IN_BLOCK	2
+define	MAXPAT		10
+
+
+# HINPUT -- Get next input line from the help block.
+
+int procedure hinput (ctrl, lbuf)
+
+pointer	ctrl
+char	lbuf[ARB]
+
+bool	formatted
+int	fd, level, status, ip
+int	getline(), strmatch(), gstrcpy(), stridxs()
+int	h_findparam(), h_findsection()
+errchk	getline, h_findparam, h_findsection
+define	findsec_ 91
+
+begin
+	fd = H_IN(ctrl)
+
+	# EOF flag is set after param block has been fully input,
+	# but normally before the actual end of file.
+
+	if (H_EOF(ctrl) == YES || H_QUIT(ctrl) == YES)
+	    return (EOF)
+	else if (H_RAWIN(ctrl) == YES || H_FILTER_INPUT(ctrl) == NO)
+	    return (getline (fd, lbuf))
+
+	# We get here only if special processing is required to
+	# filter out all but a section of the help text.
+
+	switch (H_STATE(ctrl)) {
+	case BOF:
+	    # Determine whether or not the help block is formatted.
+	    # Read in first line but save it for later and return a
+	    # dummy line to Lroff, telling it whether or not the help
+	    # block is formatted.
+
+	    repeat {
+		if (getline (fd, lbuf) == EOF)
+		    return (EOF)
+		for (ip=1;  IS_WHITE(lbuf[ip]);  ip=ip+1)
+		    ;
+	    } until (lbuf[ip] != '\n')
+	    call ungetline (fd, lbuf)
+
+	    if (lbuf[1] == '.') {
+		formatted = true
+		status = gstrcpy (".in 4\n", lbuf, SZ_LINE)
+	    } else {
+		formatted = false
+		status = gstrcpy (".nf\n", lbuf, SZ_LINE)
+	    }
+	    H_STATE(ctrl) = SEARCHING
+
+	case SEARCHING:
+	    status = getline (fd, lbuf)
+	    if (H_PARNAME(ctrl) != EOS) {
+		status = h_findparam (fd, lbuf, formatted, H_PARNAME(ctrl))
+		level = 1
+	    } else if (H_SECNAME(ctrl) != EOS)
+findsec_	status = h_findsection (fd, lbuf, formatted, H_SECNAME(ctrl))
+	    H_STATE(ctrl) = IN_BLOCK
+
+	case IN_BLOCK:
+	    # By the time we get here we are in the parameter or single
+	    # section.
+	    status = getline (fd, lbuf)
+
+	    if (lbuf[1] == '.') {
+		if (strmatch (lbuf, "^.{ih}") > 0) {
+		    if (stridxs ("|", H_SECNAME(ctrl)) > 0)
+			goto findsec_
+		    status = EOF
+		} else if (H_PARNAME(ctrl) != EOS) {
+		    if (strmatch (lbuf, "^.{ls}") > 0)
+			level = level + 1
+		    else if (strmatch (lbuf, "^.{le}") > 0) {
+			level = level - 1
+			if (level == 0)
+			    status = EOF
+		    }
+		}
+	    }
+
+	default:
+	    call error (15, "hinput: unknown input state encountered")
+	}
+
+	return (status)
+end
+
+
+# H_FINDPARAM -- If text contains format directives, eat input lines until
+# a ".ls" directive is found which contains the param name as a substring.
+# If the text is not formatted, search for a line beginning with the pattern.
+# We are called with the first line of the file in lbuf.
+
+int procedure h_findparam (fd, lbuf, formatted, param)
+
+int	fd
+char	lbuf
+bool	formatted
+char	param[ARB]
+
+bool	match_found
+pointer	sp, pattern
+int	getline(), strmatch(), strlen()
+errchk	getline
+
+begin
+	call smark (sp)
+	call salloc (pattern, SZ_FNAME, TY_CHAR)
+
+	match_found = false
+
+	if (formatted) {
+	    call sprintf (Memc[pattern], SZ_FNAME, "{%s}")
+		call pargstr (param)
+	    repeat {
+		if (strmatch (lbuf, "^.{ls}") > 0)
+		    if (strmatch (lbuf, Memc[pattern]) > 0) {
+			match_found = true
+			break
+		    }
+	    } until (getline (fd, lbuf) == EOF)
+
+	} else {
+	    call sprintf (Memc[pattern], SZ_FNAME, "^#{%s}")
+		call pargstr (param)
+	    repeat {
+		if (strmatch (lbuf, Memc[pattern]) > 0) {
+		    match_found = true
+		    break
+		}
+	    } until (getline (fd, lbuf) == EOF)
+	}
+
+	call sfree (sp)
+	if (match_found)
+	    return (strlen (lbuf))
+	else
+	    return (EOF)
+end
+
+
+# H_FINDSECTION -- If text contains format directives, eat input lines until
+# a ".ih" directive is found for the named section.  If the text is not
+# formatted, search for a line beginning with the section name.
+# We are called with the first line of the file in lbuf.
+
+int procedure h_findsection (fd, lbuf, formatted, sections)
+
+int	fd			# input file
+char	lbuf			# line buffer
+bool	formatted		# is help block formatted
+char	sections[ARB]		# list of sections "a|b|c"
+
+bool	match_found
+int	npat, ip
+pointer	sp, patbuf, patoff[MAXPAT], op
+bool	h_match()
+int	getline(), strmatch(), gstrcpy()
+errchk	getline
+
+begin
+	call smark (sp)
+	call salloc (patbuf, SZ_LINE, TY_CHAR)
+
+	# Process the list of sections into patbuf and patoff, i.e., into a
+	# list of EOS delimited strings in the string buffer patbuf.  Each
+	# section name or abbreviation is delimited by '|' (or).
+
+	npat = 1
+	op = patbuf
+	patoff[1] = op
+
+	for (ip=1;  sections[ip] != EOS;  ip=ip+1)
+	    switch (sections[ip]) {
+	    case '|':
+		Memc[op] = EOS
+		op = op + 1
+		npat = min (MAXPAT, npat + 1)
+		patoff[npat] = op
+	    default:
+		Memc[op] = sections[ip]
+		op = op + 1
+	    }
+	Memc[op] = EOS
+
+	match_found = false
+
+	if (formatted) {
+	    repeat {
+		if (strmatch (lbuf, "^.{ih}") > 0)
+		    if (getline (fd, lbuf) != EOF) {
+			match_found = h_match (lbuf, patoff, npat)
+			if (match_found)
+			    break
+		    }
+	    } until (getline (fd, lbuf) == EOF)
+
+	} else {
+	    repeat {
+		match_found = h_match (lbuf, patoff, npat)
+		if (match_found)
+		    break
+	    } until (getline (fd, lbuf) == EOF)
+	}
+
+
+	# If only one section is to be printed, skip the section name, otherwise
+	# pass the .ih secname on to the text formatter.
+
+	call sfree (sp)
+	if (match_found) {
+	    if (npat == 1)
+		return (getline (fd, lbuf))
+	    else {
+		call ungetline (fd, lbuf)
+		return (gstrcpy (".ih\n", lbuf, SZ_LINE))
+	    }
+	} else
+	    return (EOF)
+end
+
+
+# H_MATCH -- Match a set of patterns against a line of test, matching only
+# at the beginning of line in either case.
+
+bool procedure h_match (lbuf, patoff, npat)
+
+char	lbuf[ARB]		# line of text
+pointer	patoff[npat]		# pointers to pattern strings
+int	npat			# number of patterns
+
+int	pat
+pointer	sp, pattern
+int	strmatch()
+
+begin
+	call smark (sp)
+	call salloc (pattern, SZ_FNAME, TY_CHAR)
+
+	for (pat=1;  pat <= npat;  pat=pat+1) {
+	    call sprintf (Memc[pattern], SZ_FNAME, "^{%s}")
+		call pargstr (Memc[patoff[pat]])
+	    if (strmatch (lbuf, Memc[pattern]) > 0) {
+		call sfree (sp)
+		return (true)
+	    }
+	}
+
+	call sfree (sp)
+	return (false)
+end
diff --git a/pkg/system/help/houtput.x b/pkg/system/help/houtput.x
new file mode 100644
index 00000000..499e4292
--- /dev/null
+++ b/pkg/system/help/houtput.x
@@ -0,0 +1,147 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ttyset.h>
+include	<chars.h>
+include	<fset.h>
+include	"help.h"
+
+define	HELP	"[q=quit,d=downhalf,f|sp=downfull,j|cr=downline,N=next]"
+define	QUIT		'q'
+define	FWD_SCREEN	'f'
+define	SCROLL_DOWN	'd'
+define	NEXT_LINE	'j'
+define	NEXT_FILE	'N'
+define	NEXT_FILE_ALT	'n'
+define	REDRAW		'\014'
+
+
+# HOUTPUT -- The help line output procedure.  Count output lines and paginate
+# output if enabled.  If manpage format is desired pass output lines to manpage
+# filter.
+
+procedure houtput (ctrl, lbuf)
+
+pointer	ctrl
+char	lbuf[ARB]
+bool	formfeed, query_enabled
+int	map_cc, maxlines, nlines, ip, cmd
+int	envgeti(), ho_getcmd()
+data	map_cc /YES/
+
+begin
+	if (H_RAWOUT(ctrl) == YES) {
+	    call putline (H_OUT(ctrl), lbuf)
+	    return
+	} else if (H_MANPAGE(ctrl) == YES) {
+	    call man_output (H_OUT(ctrl), lbuf, H_NLPP(ctrl),
+		H_LMARGIN(ctrl), H_RMARGIN(ctrl))
+	    return
+	} else if (H_EOF(ctrl) == YES)
+	    return
+
+	# Check for end of page (either full or formfeed in line)
+	# before processing the output line, because we want to
+	# pause to let the user read the screen before paging.
+
+	ip = 1
+	formfeed = (lbuf[1] == '\f')
+	if (formfeed)
+	    ip = 2
+
+	iferr (maxlines = envgeti ("ttynlines") - 1)
+	    maxlines = 24
+
+	# Help blocks and files are preceded by a form feed and a header.
+	# The "more" query is issued between blocks and files to give the
+	# user a chance to read the previous page before the screen is cleared.
+	# The query is disabled for the first block or file by initialization
+	# of nlines to -1 in the main help routine.
+
+	query_enabled = (H_NLINES(ctrl) > 0)
+	nlines = max (0, H_NLINES(ctrl))
+
+	if (H_PAGINATE(ctrl) == YES) {
+	    if (formfeed || nlines >= maxlines) {
+		if (query_enabled) {
+		    # Pause to give the user a chance to read the output, and
+		    # to indicate whether they wish to continue.
+
+		    repeat {
+			cmd = ho_getcmd (H_TTY(ctrl))
+
+			switch (cmd) {
+			case FWD_SCREEN, BLANK, REDRAW:
+			    nlines = 0
+			case SCROLL_DOWN:
+			    nlines = (maxlines + 1) / 2
+			case NEXT_LINE, CR, LF:
+			    nlines = maxlines - 1
+			case NEXT_FILE, NEXT_FILE_ALT:
+			    H_EOF(ctrl) = YES
+			    nlines = 0
+			case QUIT:
+			    H_QUIT(ctrl) = YES
+			    H_EOF(ctrl) = YES
+			default:
+			    call eprintf ("\07")
+			    call flush (STDERR)
+			    cmd = ERR
+			}
+		    } until (cmd > 0)
+		}
+
+		if (formfeed && H_QUIT(ctrl) == NO) {
+		    call ttyclear (H_OUT(ctrl), H_TTY(ctrl))
+		    nlines = 0
+		}
+	    }
+	}
+
+	# Do not output the line if the user just said to quit.
+	if (H_EOF(ctrl) == NO)
+	    if (lbuf[ip] != EOS) {
+		call ttyputline (H_OUT(ctrl), H_TTY(ctrl), lbuf[ip], map_cc)
+		nlines = nlines + 1
+	    }
+
+	H_NLINES(ctrl) = nlines
+end
+
+
+# HO_GETCMD -- Query the user for a single character command keystroke.
+
+int procedure ho_getcmd (tty)
+
+pointer	tty			# tty descriptor
+
+int	key
+char	strval[1]
+int	clgkey()
+
+begin
+	# Ensure synchronization with the standard output.
+	call flush (STDOUT)
+
+	# Print query in standout mode.
+	call ttyso (STDERR, tty, YES)
+	call eprintf (HELP)
+	call ttyso (STDERR, tty, NO)
+	call flush (STDERR)
+
+	call fseti (STDIN, F_SETREDRAW, REDRAW)
+
+	# Read the command keystroke in raw mode.
+	if (clgkey ("cl.ukey", key, strval, 1) == EOF)
+	    key = QUIT
+	else if (key == INTCHAR)
+	    key = QUIT
+
+	call fseti (STDIN, F_SETREDRAW, 0)
+
+	# Erase the prompt and return.
+	call eprintf ("\r")
+	call ttyclearln (STDERR, tty)
+	call flush (STDERR)
+
+	return (key)
+end
diff --git a/pkg/system/help/lroff/breakline.o b/pkg/system/help/lroff/breakline.o
new file mode 100644
index 00000000..d0dceee2
--- /dev/null
+++ b/pkg/system/help/lroff/breakline.o
diff --git a/pkg/system/help/lroff/breakline.x b/pkg/system/help/lroff/breakline.x
new file mode 100644
index 00000000..2be9f5a1
--- /dev/null
+++ b/pkg/system/help/lroff/breakline.x
@@ -0,0 +1,99 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	"lroff.h"
+
+define	LEFT	1
+define	RIGHT	2
+
+# BREAKLINE -- Break the current output line and output with or without
+# justification.  The "current output line" is the set of NWORDS words
+# pointed to by the WORDS array of word pointers.  Copy the NWORDS words
+# into the output buffer; do nothing if the word buffer is empty.  If the
+# OUTPUT buffer is not empty, append a newline, causing the buffer to be
+# flushed.  Breakline is called by textout() when enough words have been
+# collected to fill an output line, or whenever it is desired to flush
+# the output buffer.  If both the word buffer and the output buffer are
+# empty, breakline is essentially a nop.
+
+procedure breakline (out, justify_flag)
+
+extern	out()
+int	justify_flag
+
+int	w, i, end_to_fill_from, next_output_column
+int	nholes, nfill, n_per_hole, nextra, hole1, hole2
+errchk	outstr, outc
+include	"lroff.com"
+include	"words.com"
+
+begin
+	if (wbuf == NULL || words == NULL)
+	    call error (1, "No Lroff word buffer allocated")
+
+	# First we flush the word buffer.
+	if (nwords > 0) {
+	    # Strip any trailing whitespace from the line.
+	    for (wp=wp-2;  Memc[wp] == BLANK && wp > wbuf;  wp=wp-1)
+		wcols = wcols - 1
+	    wp = wp + 1
+	    Memc[wp] = EOS
+	    wp = wp + 1
+
+	    # If justification is disabled or if there is only one word on
+	    # the line, do not add spaces to right justify.
+
+	    if (justify_flag == NJ || justify == NO || nwords <= 1) {
+		for (w=1;  w <= nwords;  w=w+1)
+		    call outstr (out, Memc[Memi[words+w-1]])
+
+	    } else {
+		# To justify the line, determine the number of extra spaces
+		# needed to right justify the last character on the line.
+		# Determine the number of holes between words, and how many
+		# spaces to add to each.
+
+		nholes = nwords - 1
+		nfill = max (0, right_margin - left_margin + 1 - wcols)
+		n_per_hole = nfill / nholes
+		nextra = nfill - (n_per_hole * nholes)
+
+		# Determine where the extra spaces need to be added.  Add
+		# extra spaces from the left and then the right on succesive
+		# lines.
+		if (end_to_fill_from == LEFT) {
+		    hole1 = 1
+		    hole2 = nextra
+		    end_to_fill_from = RIGHT
+		} else {
+		    hole1 = nwords - nextra
+		    hole2 = nholes
+		    end_to_fill_from = LEFT
+		}
+
+		# Fill the output line.  Move the word and then add the
+		# requisite number of blanks per hole, plus an extra if in
+		# the range hole1 to hole2 (at left or right).
+
+		do w = 1, nwords {
+		    call outstr (out, Memc[Memi[words+w-1]])
+		    do i = 1, n_per_hole
+			call outc (out, BLANK)
+		    if (w >= hole1 && w <= hole2)
+			call outc (out, BLANK)
+		}
+	    }
+	}
+
+
+	# If there is anything in the output buffer, append a newline, flushing
+	# the buffer.
+
+	call getoutcol (next_output_column)
+	if (next_output_column > left_margin)
+	    call outc (out, '\n')
+
+	wp = wbuf				# clear the word buffer
+	nwords = 0
+	wcols = 0
+end
diff --git a/pkg/system/help/lroff/center.o b/pkg/system/help/lroff/center.o
new file mode 100644
index 00000000..6e3442d3
--- /dev/null
+++ b/pkg/system/help/lroff/center.o
diff --git a/pkg/system/help/lroff/center.x b/pkg/system/help/lroff/center.x
new file mode 100644
index 00000000..b3581f3a
--- /dev/null
+++ b/pkg/system/help/lroff/center.x
@@ -0,0 +1,32 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	"lroff.h"
+
+# CENTER_TEXT -- Center and output the next input line within the current
+# left and right margins.  The optional argument specifies the column
+# (measured from the left margin) at which the text is to be centered.
+
+procedure center_text (in, out, linebuf, ip)
+
+extern	in(), out()
+char	linebuf[ARB]
+int	ip
+
+int	len_inputline, center_column, nblanks, i
+int	in(), input(), lgetarg()
+errchk	breakline, input, outc, outline
+include	"lroff.com"
+
+begin
+	call breakline (out, NJ)
+	center_column = lgetarg (linebuf, ip, (left_margin + right_margin) / 2)
+	len_inputline = input (in, linebuf) - 1
+
+	if (len_inputline != EOF) {
+	    nblanks = center_column - (len_inputline / 2) - left_margin
+	    for (i=1;  i <= nblanks;  i=i+1)
+		call outc (out, BLANK)
+	    call outline (out, linebuf)
+	}
+end
diff --git a/pkg/system/help/lroff/dols.o b/pkg/system/help/lroff/dols.o
new file mode 100644
index 00000000..55b5fbcd
--- /dev/null
+++ b/pkg/system/help/lroff/dols.o
diff --git a/pkg/system/help/lroff/dols.x b/pkg/system/help/lroff/dols.x
new file mode 100644
index 00000000..26ba165b
--- /dev/null
+++ b/pkg/system/help/lroff/dols.x
@@ -0,0 +1,108 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	<error.h>
+include	"lroff.h"
+
+# DO_LS -- Depending on the action, push an LS block (print label and indent
+# one level), or pop an LS block (restore the previous indent level).  INIT_LS
+# clears the LS stack.  Called with the LS command line in "lbuf", minus
+# the ".ls".  The command line may consist of an argument specifying the number
+# of spaces to indent, the label string, both, or neither.
+# If the label string is shorter than the amount by which the block is indented,
+# the text block will begin on the same line as the label, otherwise the line
+# is broken and the text block begins on the following line.
+
+procedure do_LS (out, lbuf, action, last_command)
+
+extern	out()
+char	lbuf[ARB]			# ".ls [arg] [text]"
+int	action				# LS or LE
+int	last_command
+
+int	indent[MAX_NLS]
+int	n, ip, next_output_column
+int	lgetarg(), strlen()
+errchk	skiplines, outstr, outc, breakline
+include	"lroff.com"
+
+begin
+	switch (action) {
+	case LS:
+	    # We normally skip a line when beginning an LS block.  If two or
+	    # more LS directives are given in a row, however, only skip a
+	    # single line.
+
+	    call breakline (out, NJ)
+	    if (last_command != LS)
+		call skiplines (out, 1)
+	    call testpage (out, 3)
+
+	    # Push new LS block on stack.
+	    nls = nls + 1
+	    if (nls > MAX_NLS) {
+		iferr (call error (1, "LS blocks nested too deep"))
+		    call erract (EA_WARN)
+		nls = MAX_NLS
+	    }
+
+	    # Get number of spaces to indent, if given.  If arg is negative,
+	    # do not remember the argument, otherwise make it the new default.
+	    ip = 1
+	    n = lgetarg (lbuf, ip, ls_indent)
+	    if (n < 0)
+		indent[nls] = -n
+	    else {
+		ls_indent = n
+		indent[nls] = ls_indent
+	    }
+
+	    # Copy the label, if any, into the output buffer.  We must do this
+	    # before we change the left margin since the label is not indented.
+
+	    call outstr (out, lbuf[ip])
+
+	    # Try to adjust the left margin by the indicated amount.  Save the
+	    # actual indentation level for restoration by LE.
+
+	    indent[nls] = max (perm_left_margin, min (right_margin,
+		left_margin + indent[nls])) - left_margin
+	    left_margin = left_margin + indent[nls]
+
+	    # If the length of the label string plus one blank does not leave
+	    # space to start the first line of the text block on the same line,
+	    # we must break the line and start the block on the next line.
+	    # Otherwise, output spaces until the new left margin is reached.
+
+	    if (strlen (lbuf[ip]) >= indent[nls])
+		call outc (out, '\n')
+	    else {
+		call getoutcol (next_output_column)
+		while (next_output_column < left_margin) {
+		    call outc (out, BLANK)
+		    call getoutcol (next_output_column)
+		}
+	    }
+
+	case LE:					# end LS block
+	    call breakline (out, NJ)
+	    if (nls >= 1) {
+		left_margin = left_margin - indent[nls]
+		nls = nls - 1
+	    }
+
+	default:
+	    call error (1, "do_LS")
+	}
+end
+
+
+# INIT_LS -- Set or clear any LS indentation.
+
+procedure init_ls()
+
+include	"lroff.com"
+
+begin
+	nls = 0
+end
diff --git a/pkg/system/help/lroff/getarg.o b/pkg/system/help/lroff/getarg.o
new file mode 100644
index 00000000..fcfc8fd9
--- /dev/null
+++ b/pkg/system/help/lroff/getarg.o
diff --git a/pkg/system/help/lroff/getarg.x b/pkg/system/help/lroff/getarg.x
new file mode 100644
index 00000000..b85c5462
--- /dev/null
+++ b/pkg/system/help/lroff/getarg.x
@@ -0,0 +1,35 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	"lroff.h"
+
+.help lgetarg
+.nf _________________________________________________________________________
+LGETARG -- Get an integer argument to a directive.  If no argument is found,
+return the default value.  We are called with IP pointing to the start of
+the argument field to be searched.  Leave IP pointing to the next argument
+field.
+.endhelp ____________________________________________________________________
+
+int procedure lgetarg (input_line, ip, default_value)
+
+char	input_line[ARB]
+int	ip, default_value
+int	argument
+int	ctoi()
+
+begin
+	if (ctoi (input_line, ip, argument) == 0)
+	    argument = default_value
+
+	# Eat comma argument delimiter, if multiple arguments.  Also eat
+	# trailing whitespace, in case a string argument follows.
+	while (input_line[ip] == BLANK)
+	    ip = ip + 1
+	if (input_line[ip] == ',')
+	    ip = ip + 1
+	while (input_line[ip] == BLANK)
+	    ip = ip + 1
+
+	return (argument)
+end
diff --git a/pkg/system/help/lroff/indent.o b/pkg/system/help/lroff/indent.o
new file mode 100644
index 00000000..b2f75a10
--- /dev/null
+++ b/pkg/system/help/lroff/indent.o
diff --git a/pkg/system/help/lroff/indent.x b/pkg/system/help/lroff/indent.x
new file mode 100644
index 00000000..6857e315
--- /dev/null
+++ b/pkg/system/help/lroff/indent.x
@@ -0,0 +1,17 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"lroff.h"
+
+# INDENT_LEFT_MARGIN -- Execute a relative indent of the left margin.
+
+procedure indent_left_margin (in, out, number_of_spaces)
+
+extern	in(), out()
+int	in(), number_of_spaces
+include	"lroff.com"
+
+begin
+	call breakline (out, NJ)
+	left_margin = max (perm_left_margin, min (right_margin,
+	    left_margin + number_of_spaces))
+end
diff --git a/pkg/system/help/lroff/input.o b/pkg/system/help/lroff/input.o
new file mode 100644
index 00000000..80ef4865
--- /dev/null
+++ b/pkg/system/help/lroff/input.o
diff --git a/pkg/system/help/lroff/input.x b/pkg/system/help/lroff/input.x
new file mode 100644
index 00000000..908bd81d
--- /dev/null
+++ b/pkg/system/help/lroff/input.x
@@ -0,0 +1,123 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	<chars.h>
+include	"lroff.h"
+
+.help input
+.nf ___________________________________________________________________________
+INPUT -- Read a line of text into the user supplied input buffer.  Convert
+any tabs therein to spaces to simplify further processing.  If generation
+of standout mode control chars is enabled, map "\f[BIR]" sequences into the
+appropriate control chars (defined in <chars.h>), otherwise delete any such
+sequences found (these control chars are later mapped by HELP, PAGE, LPRINT
+etc. into whatever sequence the actual output device requires).  Return the
+number of PRINTABLE chars in the input line.  Control characters are not
+counted, but are copied to the output line.  The trailing newline is stripped;
+Lroff deals mainly with words, not lines.  Only printable characters are made
+to "stand out", i.e., standout mode is always turned off between words and at
+the end of a line.
+.endhelp ______________________________________________________________________
+
+int procedure input (in, userbuf)
+
+extern	in()
+char	userbuf[ARB]
+
+bool	standout_mode_in_effect
+char	ch
+int	len_inputline, ocol
+pointer	sp, lbuf, ip, op
+int	stridx(), in()
+errchk	salloc, in
+include	"lroff.com"
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_IBUF, TY_CHAR)
+
+	# Get input line and deal with any tab characters therein.
+	if (in (in_magic_arg, Memc[lbuf]) == EOF) {
+	    call sfree (sp)
+	    return (EOF)
+	}
+
+	standout_mode_in_effect = false
+	len_inputline = 0
+	ip = lbuf
+	op = 1
+	ocol = 0
+
+	# Process the input buffer, converting any "\f?" font escape sequences
+	# found.  Terminate when newline is reached.  Delete the newline.
+	# Expand all tabs.
+
+	for (ch=Memc[ip];  ch != '\n' && ch != EOS;  ch=Memc[ip]) {
+	    if (ch == '\\')
+		if (Memc[ip+1] == 'f' && stridx (Memc[ip+2], "BIR") > 0) {
+		    # Turn standout mode on or off.  Can only be turned on
+		    # if "soflag" is YES.
+		    switch (Memc[ip+2]) {
+		    case 'B', 'I':			# bold, italic
+			if (soflag == YES)
+			    standout_mode_enabled = true
+		    case 'R':				# roman
+			if (standout_mode_in_effect) {
+			    userbuf[op] = SO_OFF
+			    op = op + 1
+			    standout_mode_in_effect = false
+			}
+			standout_mode_enabled = false
+		    }
+		    ip = ip + 3				# \f? = 3
+		    next
+		}
+
+	    # Only make alphanumeric chars "stand out".
+
+	    if (IS_ALNUM(ch)) {
+		len_inputline = len_inputline + 1
+		ocol = ocol + 1
+		if (standout_mode_enabled && !standout_mode_in_effect) {
+		    userbuf[op] = SO_ON
+		    op = op + 1
+		    standout_mode_in_effect = true
+		}
+
+	    } else if (ch == '\t') {
+		repeat {
+		    userbuf[op] = ' '
+		    op = op + 1
+		    ocol = ocol + 1
+		    len_inputline = len_inputline + 1
+		} until (ocol > 1 && mod (ocol, TABSIZE) == 0)
+		ip = ip + 1
+		next
+
+	    } else {
+		if (IS_PRINT(ch)) {
+		    len_inputline = len_inputline + 1
+		    ocol = ocol + 1
+		}
+		if (standout_mode_in_effect) {
+		    userbuf[op] = SO_OFF
+		    op = op + 1
+		    standout_mode_in_effect = false
+		}
+	    }
+
+	    userbuf[op] = ch
+	    op = op + 1
+	    ip = ip + 1
+	}
+
+	if (standout_mode_in_effect) {
+	    userbuf[op] = SO_OFF
+	    op = op + 1
+	    standout_mode_in_effect = false
+	}
+	userbuf[op] = EOS
+
+	call sfree (sp)
+	return (len_inputline)
+end
diff --git a/pkg/system/help/lroff/justify.o b/pkg/system/help/lroff/justify.o
new file mode 100644
index 00000000..db814dfb
--- /dev/null
+++ b/pkg/system/help/lroff/justify.o
diff --git a/pkg/system/help/lroff/justify.x b/pkg/system/help/lroff/justify.x
new file mode 100644
index 00000000..8085ca6b
--- /dev/null
+++ b/pkg/system/help/lroff/justify.x
@@ -0,0 +1,63 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	"lroff.h"
+
+# RIGHT_JUSTIFY -- Right justify the text string argument in LINEBUF on the
+# next input line, and write to the output.
+
+procedure right_justify (in, out, linebuf, ip)
+
+extern	in(), out()
+char	linebuf[ARB]
+int	ip
+pointer	sp, rjbuf
+int	in(), input()
+errchk	salloc, breakline, input, rjline
+include	"lroff.com"
+
+begin
+	call smark (sp)
+	call salloc (rjbuf, SZ_IBUF, TY_CHAR)
+
+	call breakline (out, NJ)
+	if (input (in, Memc[rjbuf]) != EOF)
+	    call rjline (out, Memc[rjbuf], linebuf[ip])
+
+	call sfree (sp)
+end
+
+
+
+# RJLINE -- Right justify a text string on an unfilled input line and
+# send it out.
+
+procedure rjline (out, input_line, rjtext)
+
+extern	out()
+char	input_line[ARB]		# unfilled input line
+char	rjtext[ARB]		# string to be right justified on same line
+
+int	i, nblanks, len_rjtext, next_output_column
+int	textlen()
+errchk	outstr, outc, outline
+include	"lroff.com"
+
+begin
+	# Breakline should already have been called by the time we get here.
+	# Output the input line at the left margin without filling.
+
+	call outstr (out, input_line)
+
+	# Determine the (printable) length of the rjtext string, and space
+	# over so that it comes out right justified.  Always output at least
+	# one space.  Exceed the right margin if necessary.
+
+	call getoutcol (next_output_column)
+	len_rjtext = textlen (rjtext)
+	nblanks = max (1, right_margin - (next_output_column + len_rjtext) + 1)
+	do i = 1, nblanks
+	    call outc (out, BLANK)
+
+	call outline (out, rjtext)
+end
diff --git a/pkg/system/help/lroff/lroff.com b/pkg/system/help/lroff/lroff.com
new file mode 100644
index 00000000..1e64173c
--- /dev/null
+++ b/pkg/system/help/lroff/lroff.com
@@ -0,0 +1,24 @@
+# Common for the Lroff text formatter.
+
+int	right_margin			# working margins
+int	left_margin
+int	perm_right_margin		# permanent margins
+int	perm_left_margin
+int	in_magic_arg			# magic args for in/out procedures
+int	out_magic_arg
+int	soflag				# if YES, output standout mode chars
+int	foflag				# if YES, output forms mode chars
+int	justify				# right justify text
+int	nls				# .LS nesting level
+int	ls_indent			# .LS, def number of spaces to indent
+int	sh_nskip			# .SH, def nlines to skip
+int	ih_nskip			# .IH, def nlines to skip
+int	ih_indent			# .IH, def nspaces to indent
+int	nh_nskip			# .NH, def nlines to skip
+int	nh_level[MAX_NHLEVEL]		# .NH, section level numbers
+bool	standout_mode_enabled		# see input()
+
+common	/lrfcom/ right_margin, left_margin, perm_right_margin,
+	perm_left_margin, in_magic_arg, out_magic_arg, soflag,
+	foflag, justify, nls, ls_indent, sh_nskip, ih_nskip, ih_indent,
+	nh_nskip, nh_level, standout_mode_enabled
diff --git a/pkg/system/help/lroff/lroff.h b/pkg/system/help/lroff/lroff.h
new file mode 100644
index 00000000..5111fcd0
--- /dev/null
+++ b/pkg/system/help/lroff/lroff.h
@@ -0,0 +1,41 @@
+# Input buffer must allow space for tab expansion and standout mode control
+# characters.  Word and output buffer dimensions depend on margins.
+
+define	SZ_IBUF		(2*SZ_LINE)
+define	MAX_NLS		20		# nesting level for LS
+define	MAX_NHLEVEL	10		# max level for numbered sections
+
+# Default formatter parameters.
+define	DEF_IHINDENT	4		# .ih indent level
+define	DEF_LSINDENT	4		# .ls indent level
+define	DEF_IHNSKIP	2		# .ih number of lines to skip
+define	DEF_NHNSKIP	2		# .nh number of lines to skip
+define	DEF_SHNSKIP	2		# .sh number of lines to skip
+define	DEF_TPNLINES	2		# .tp nlines left on page
+
+define	TABSIZE		8
+define	INVISIBLE	($1 < BLANK)
+
+# Lroff Directive Opcodes.
+define	FI		1		# enter fill mode
+define	NF		2		# leave fill mode (nofill)
+define	JU		3		# enter line justification mode
+define	NJ		4		# leave line justification mode
+define	RJ		5		# right justify text on nf,nj line
+define	SH		6		# section heading
+define	IH		7		# indented section heading
+define	NH		8		# numbered section heading
+define	BR		9		# break line
+define	CE		10		# center next line
+define	SP		11		# break, space N spaces on output
+define	IN		12		# indent +/- N spaces
+define	LS		13		# begin labelled section
+define	LE		14		# end labelled section
+define	BP		15		# break page
+define	TP		16		# test space left on page
+define	KS		17		# start floating keep
+define	KE		18		# end floating keep
+define	HR		19		# HTML href tag
+define	HN		20		# HTML name tag
+define	ENDHELP		21		# end of help block
+define	HELP		22		# start of help block
diff --git a/pkg/system/help/lroff/lroff.hlp b/pkg/system/help/lroff/lroff.hlp
new file mode 100644
index 00000000..6607977b
--- /dev/null
+++ b/pkg/system/help/lroff/lroff.hlp
@@ -0,0 +1,258 @@
+.help lroff Nov83 "Online Help Utilities"
+.ih
+NAME
+\fBlroff\fR -- line oriented text formatter
+.ih
+PURPOSE
+\fBLroff\fR is a simple text formatter used by the IRAF on-line Help command,
+and other utilities (MANPAGE, LIST), to format text.  
+\fBLroff\fR style documentation text may be embedded in program source files.
+\fBlroff\fR is line oriented, rather than page oriented,
+and is implemented as a library procedure rather than as a task.
+.ih
+USAGE
+status = lroff (input, output, left_margin, right_margin)
+.ih
+PARAMETERS
+.ls input
+An integer procedure, called by \fBlroff\fR to get lines of input,
+which takes the \fBlroff\fR input buffer as an argument,
+and which returns EOF upon End of File (like GETLINE).
+Each line of input must be terminated by a newline and an EOS
+(End Of String marker).
+.le
+.ls output
+A procedure, called by \fBlroff\fR to output formatted lines of text,
+which takes the \fBlroff\fR output buffer as an argument ("output (buffer)").
+.le
+.ls left_margin
+The first column to be filled (>= 1).
+.le
+.ls right_margin
+The last column to be filled.
+.le
+.ls status
+ERR is returned if meaningless margins are specified, or if an unrecoverable
+error occurs during processing.
+.le
+.ih
+DESCRIPTION
+\fBLroff\fR input may be bracketed by ".help" and ".endhelp" directives in
+the actual source file of the program being documented (if intended as input
+to the \fBhelp\fR utility), or may be in a separate file.
+The input text consists
+of a mixture of lines of text and \fBlroff\fR directives.
+\fBLroff\fR recognizes only a few directives,
+summarized in the "Request Summary" below.  Whenever a directive
+performs the same function as a UNIX TROFF directive, the name is the same.
+Unrecognized directives are ignored, and are not passed on to the output.
+Directives must be left justified and preceeded by a period.
+
+Help text need not be formatted unless desired.  Filling and justification
+are NOT ENABLED unless a legal directive (other than ".nf") is given on the
+line immediately following the ".help" directive.
+
+While filling, embedded whitespace in text IS significant to \fBlroff\fR,
+except at the end of a line.
+\fBlroff\fR recognizes no special characters.
+Blank lines cause a break, and are passed on to the output (a blank line
+is equivalent to ".sp"). 
+Case is not significant in command directives.
+Control characters embedded in text will be passed on to the output.
+
+Since both whitespace and blank lines are significant, \fBlroff\fR will properly
+format ordinary paragraphs of text, and single line section headers,
+without need for embedded directives.
+
+Since the i/o routines used by \fBlroff\fR are parameterized, pagination can be
+achieved by having the user supplied OUTPUT procedure count output lines.
+Similarly, pagination control directives can be added to the list of
+\fBlroff\fR directives, to be intercepted by the user supplied INPUT procedure.
+See the Manpage command for an example.
+
+
+DIRECTIVES
+
+Most of the \fBlroff\fR directives function the same as in the UNIX text
+formatters.  For the benefit of readers without experience with UNIX,
+"filling" means collecting words of text until an output line has been
+filled, and "justification" refers to adding extra spaces between words
+to cause the output line to be both left and right justified (as in this
+paragraph).  Filling is disabled with NF, and resumes following a FI.
+While filling is disabled, only the control directives FI and RJ will be
+recognized.  Justification is enabled with JU, and disabled with NJ.
+The filling of an output line may be stopped, and the line output, with BR.
+SP (or a blank line) does the same thing, outputting one or more blank
+lines as well.  CE causes the current line to be broken, and outputs the
+next line of input, centered.
+
+The directive ".rj text" breaks the current line, and outputs the next
+line of input, unfilled, with "text" right justified on the same line.
+RJ is especially useful for numbering equations.  The RJ directive is
+recognized whether or not filling is in effect.
+
+SH and IH may be used for section headers.  Both cause a break, followed
+by a couple blank lines, followed by the next line of input,
+left justified on the output line.  The left margin is reset to its
+initial value.  If IH is used, the text following the section header will
+be indented one level in from the left margin.
+The number of lines of blank lines before the heading,
+and the amount of indentation, are optional arguments.
+The default values are shown in the request summary below.  If values
+other than the defaults are desired, they need only be supplied as arguments
+once.  Succeeding calls will continue to use the new values.
+
+The IH and LS directives are especially useful in help text (manual pages).
+LS with a label string is useful for parameter lists,
+as shown in the example below.
+LS without a label string is used for relative indenting.
+A following LE restores the previous level of indentation.
+
+The LS directive has the form ".ls [n] [stuff]", where "n" (optional)
+is the amount by which the following text is to be indented,
+and "stuff" is the (optional) label for the indented text block.
+LS causes a break, followed by one blank line, then the label string (if given),
+left justified.
+If the length of "stuff" is less than N-1 characters, the text
+block will start filling on the same line, otherwise on the next line.
+The indented text block may contain anything, including additional LS
+directives if nesting is desired.  A matching LE eventually terminates the
+block, restoring the previous level of indentation.
+
+The LS directive takes the most recent argument as the new default
+indentation, allowing the argument to be omitted in subsequent calls.
+To keep the current default value from being changed, use a negative
+argument.
+
+.ih
+EXAMPLE
+.sp
+Many examples of the use of the \fBlroff\fR command directives in help text
+can be found by browsing about in source listings.
+A brief example is included here for convenient reference.
+.sp
+The ".help" directive, used to mark the beginning
+of a block of help text, is used by HELP and MANPAGE rather than \fBlroff\fR.
+The (optional) arguments to ".help" are the keyword name of the help
+text block, and two strings.
+The keyword argument may be a list of the form ".help keyw1,
+keyw2, ..., keywn", if more than one keyword is appropriate.
+The first keyword in the list is placed in the header of a manual page,
+followed by the first string, in parenthesis.  The second string,
+if given, is centered in the header line.  The strings need not be
+delimited unless they contain whitespace.
+.sp
+The \fBlroff\fR-format help text fragment
+.sp
+.ls
+.nf
+.help stcopy	2	"string utilities"
+.ih
+NAME
+stcopy -- copy a string.
+.ih
+PURPOSE
+Stcopy is used to copy an EOS delimited character
+string.  The EOS delimiter MUST be present.
+.ih
+USAGE
+stcopy (from, to, maxchar)
+.ih
+PARAMETERS
+.ls from
+The input string.
+.le
+.ls to
+The output string, of length no less than "maxchar"
+characters (excluding the EOS).
+.le
+.ls maxchar
+The maximum number of characters to be copied.
+Note that "maxchar" does not include the EOS.
+Thus, the destination string must contain storage
+for at least (maxchar + 1) characters.
+.le
+.ih
+DESCRIPTION
+.fi
+.sp 2
+.le
+would be converted by \fBlroff\fR (as called from Help) into something like
+the following.  Remember that the margins are runtime arguments to \fBlroff\fR.
+Help does not print a header line, or break pages.
+.sp 2
+.in 8
+NAME
+.in 5
+stcopy -- copy a string.
+.in -5
+.sp 2
+PURPOSE
+.in 5
+Stcopy  is  used  to  copy  an  EOS delimited character
+string.  The EOS delimiter MUST be present.
+.in -5
+.sp 2
+USAGE
+.in 5
+stcopy (from, to, maxchar)
+.in -5
+.sp 2
+PARAMETERS
+.in 5
+.ls from
+The input string.
+.le
+.ls to
+The output string, of length no less than "maxchar"
+characters (excluding the EOS).
+.le
+.ls maxchar
+The maximum number of characters to be copied.
+Note that "maxchar" does not include the EOS.
+Thus, the destination string must contain storage
+for at least (maxchar + 1) characters.
+.le
+.in -5
+.sp 2
+DESCRIPTION
+.sp
+.ih
+SEE ALSO
+help
+
+The reader should note that MANPAGE, which is page oriented,
+recognizes the following directives in addition to those recognized
+by \fBlroff\fR: BP (break page), and KS, KE (start and end keep).  MANPAGE also
+omits blank lines at the top of a page.  These directives may safely
+be included in \fBlroff\fR text, as they will be ignored by \fBlroff\fR if not
+intercepted by the procedure calling \fBlroff\fR.
+
+.ih
+REQUEST SUMMARY
+.sp
+.nf
+Request Initial Default  Break		Meaning
+
+  .fi	  yes		  yes	Begin filling output lines.
+  .nf	  no		  yes	Stop filling output lines.
+  .ju	  yes		  no	Right justify output lines.
+  .nj	  no		  no	Don't right justify.
+  .rj text		  yes	Rt justify text on next line.
+  .sh n		  n=2	  yes	Skip n lines, start section.
+  .ih m n	m=2,n=5	  yes	Like SH, but indent n spaces.
+  .br			  yes	Stop filling current line.
+  .ce			  yes	Center following line.
+  .sp n		  n=1	  yes	Space "n" lines.
+  .in n	  n=0	  n=0	  yes	Set left margin to "current+n".
+  .ls n	label	  n=8	  yes	Begin labeled text block.
+  .le			  yes	End labeled text block.
+
+additional directives provided by MANPAGE:
+
+  .bp			  yes	Start a new page of output.
+  .tp n   n=4		  yes	Break page if < n lines left.
+  .ks			  yes	Begin saving output.
+  .ke			  yes	Output saved text all on one page.
+.fi
+.endhelp
diff --git a/pkg/system/help/lroff/lroff.o b/pkg/system/help/lroff/lroff.o
new file mode 100644
index 00000000..4151c327
--- /dev/null
+++ b/pkg/system/help/lroff/lroff.o
diff --git a/pkg/system/help/lroff/lroff.x b/pkg/system/help/lroff/lroff.x
new file mode 100644
index 00000000..e9624827
--- /dev/null
+++ b/pkg/system/help/lroff/lroff.x
@@ -0,0 +1,220 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	<ctype.h>
+include	"lroff.h"
+
+.help lroff
+.nf ___________________________________________________________________________
+Source for the LROFF line-oriented text formatter.  LROFF is a simple text
+formatter patterned after the NROFF formatter of UNIX.  LROFF is unusual in
+that it is a "numerical" library procedure which does not directly do any
+i/o (except for buffer allocation and use of ERROR).
+
+LROFF -- The main entry point.  Lroff reads lines with the input procedure,
+performs the format conversion, and writes lines with the output procedure,
+both of which are passed as arguments.  Conversion proceeds until the .endhelp
+directive or EOF is reached.  The calling sequences for the input and output
+procedures are as follows:
+
+	stat =	 inproc (inarg, linebuf)
+		outproc (outarg, linebuf)
+
+where "inarg" and "outarg" are magic integer arguments which Lroff merely
+passes on to the input and output procedures when they are called.
+
+Other arguments to Lroff include PLM and PRM, the permanent left and right
+margins for the output text, and SOFLAG, set to YES if "standout mode" control
+chars are permitted in the output text.
+
+The forms control directives BP, TP, KS, and KE are ignored unless forms mode
+is enabled (foflag=YES).  If forms control is enabled these directives cause
+a breakline followed by output of a special control character forms directive,
+used to control the layout of text on a page.  When forms mode is in effect
+the section header directives also cause output of a TP (test page) directive
+before the section.  Processing of forms control characters is left to the
+program that reads lroff output.
+.endhelp ______________________________________________________________________
+
+procedure lroff (in, in_arg, out, out_arg, plm, prm, soflag_val, foflag_val)
+
+extern	in()			# called to get lines of input text
+int	in_arg			# magic argument for in()
+extern	out()			# called to output formatted lines of text
+int	out_arg			# magic argument for out()
+int	plm, prm		# permanent left and right margins
+int	soflag_val		# output standout mode control chars?
+int	foflag_val		# output form control chars?
+
+char	ctrlstr[2]
+pointer	sp, ibuf
+int	ip, command, last_command
+int	in(), nextcmd(), lgetarg(), input(), nofill()
+
+errchk	input, textout, nofill, rawcopy, skiplines, breakline, right_justify
+errchk	salloc, new_section, new_indented_section, center_text, init_ls
+errchk	init_nh, indent_left_margin, do_LS, textout, set_wordbuf, set_outbuf
+include	"lroff.com"
+
+define	text_	98
+
+begin
+	call smark (sp)
+	call salloc (ibuf, SZ_IBUF, TY_CHAR)
+
+	if (plm > prm || plm < 1)
+	    call error (1, "Lroff called with invalid margins")
+
+	# General initialization.  Set up the Lroff common.  Call the various
+	# initialization procedures to initialize the directives and to
+	# set up the word buffer and output buffer, the size of which depends
+	# on the margins.
+
+	justify = YES
+	perm_left_margin = plm
+	perm_right_margin = prm
+	left_margin = plm
+	right_margin = prm
+	last_command = NULL
+	in_magic_arg = in_arg
+	out_magic_arg = out_arg
+	soflag = soflag_val
+	foflag = foflag_val
+	standout_mode_enabled = false
+
+	ls_indent = DEF_LSINDENT
+	ih_indent = DEF_IHINDENT
+	sh_nskip  = DEF_SHNSKIP
+	nh_nskip  = DEF_NHNSKIP
+	ih_nskip  = DEF_IHNSKIP
+
+	call init_ls()
+	call init_nh()
+	call set_wordbuf (prm - plm + 2)
+	call set_outbuf (max (SZ_LINE, 2 * (prm - plm + 1)))
+
+	# If the first line of text is not an Lroff directive, we copy the
+	# input to the output without modification, except for moving the text
+	# to the desired left margin, stopping only at EOF or .endhelp.
+	# If any directive is given, the default mode is justify+fill.
+
+	if (input (in, Memc[ibuf]) == EOF) {
+	    call sfree (sp)
+	    return
+	} else if (nextcmd (Memc[ibuf], ip) < 0) {
+	    call rawcopy (in, out, Memc[ibuf])
+	    call sfree (sp)
+	    return
+	}
+
+
+	# The main Lroff interpreter loop.  Get input line: if directive,
+	# execute directive; else call textout() to process a line of text.
+	# The basic idea is to break the input stream up into words, saving
+	# these until we have one more than needed to fill the output line.
+	# The words are then copied into the output line, starting at the left
+	# margin, adding spaces as needed to right justify the line.  Many
+	# commands cause the "current output line" to be broken, forcing
+	# whatever has been accumulated out without right justification.
+
+	repeat {
+	    command = nextcmd (Memc[ibuf], ip)
+	    switch (command) {
+	    case FI:
+		call breakline (out, NJ)
+	    case NF:
+		call breakline (out, NJ)
+		if (nofill (in, out, Memc[ibuf]) == ENDHELP)
+		    break
+	    case JU:
+		justify = YES
+	    case NJ:
+		justify = NO
+	    case RJ:
+		call right_justify (in, out, Memc[ibuf], ip)
+	    case SH:
+		call new_section (in, out, Memc[ibuf], ip)
+	    case IH:
+		call new_indented_section (in, out, Memc[ibuf], ip)
+	    case NH:
+		call new_numbered_section (in, out, Memc[ibuf], ip)
+	    case BR:
+		call breakline (out, NJ)
+	    case CE:
+		call center_text (in, out, Memc[ibuf], ip)
+	    case SP:
+		call skiplines (out, lgetarg(Memc[ibuf],ip,1))
+	    case IN:
+		call indent_left_margin (in, out, lgetarg(Memc[ibuf],ip,0))
+	    case LS, LE:
+		call do_LS (out, Memc[ibuf+ip-1], command, last_command)
+
+	    case HR:
+		# HTML href target of the form ".hr <href> <text>", we skip
+		# ahead to the <text> and process as a normal line.
+		while (IS_WHITE(Memc[ibuf+ip]))		# skip space
+		    ip = ip + 1
+		while (!IS_WHITE(Memc[ibuf+ip]))	# skip <href>
+		    ip = ip + 1
+		call amovc (Memc[ibuf+ip+1], Memc[ibuf], SZ_IBUF)
+		ip = 0
+		goto text_
+
+	    case HN:
+		# HTML name target of the form ".hn <name>", ignore.
+		next
+
+	    # The following cases are for forms control.
+	    case BP:
+		call breakline (out, NJ)
+		if (foflag == YES)
+		    call outcc (out, FC_BREAKPAGE)
+	    case TP:
+		call breakline (out, NJ)
+		if (foflag == YES) {
+		    ctrlstr[1] = FC_TESTPAGE
+		    ctrlstr[2] = lgetarg (Memc[ibuf], ip, DEF_TPNLINES)
+		    ctrlstr[3] = EOS
+		    call out (out_magic_arg, ctrlstr)
+		}
+	    case KS:
+		call breakline (out, NJ)
+		if (foflag == YES)
+		    call outcc (out, FC_STARTKEEP)
+	    case KE:
+		call breakline (out, NJ)
+		if (foflag == YES)
+		    call outcc (out, FC_ENDKEEP)
+
+	    case ENDHELP:			# normal exit point
+		break
+
+	    default:				# ordinary line of text
+text_		if (Memc[ibuf] == '.') {
+		    # Ignore unrecognized directives.
+		    next
+		} else {
+		    # Determine if line is blank; skip a line if so, otherwise
+		    # process a normal line of text.
+		    for (ip=0;  Memc[ibuf+ip] == BLANK;  ip=ip+1)
+			;
+		    if (Memc[ibuf+ip] == EOS)
+			call skiplines (out, 1)
+		    else if (Memc[ibuf] == '\\')
+			call textout (out, Memc[ibuf+1])
+		    else
+			call textout (out, Memc[ibuf])
+		}
+	    }
+
+	    last_command = command
+	
+	} until (input (in, Memc[ibuf]) == EOF)
+
+
+99	call breakline (out, NJ)
+	call set_wordbuf (0)
+	call set_outbuf (0)
+
+	call sfree (sp)
+end
diff --git a/pkg/system/help/lroff/lroff2html.c b/pkg/system/help/lroff/lroff2html.c
new file mode 100644
index 00000000..26cf8227
--- /dev/null
+++ b/pkg/system/help/lroff/lroff2html.c
@@ -0,0 +1,1381 @@
+/* lroff2html.x -- translated by f2c (version 20100827).
+   You must link the resulting object file with libf2c:
+	on Microsoft Windows system, link with libf2c.lib;
+	on Linux or Unix systems, link with .../path/to/libf2c.a -lm
+	or, if you install libf2c.a in a standard place, with -lf2c -lm
+	-- in that order, at the end of the command line, as in
+		cc *.o -lf2c -lm
+	Source for libf2c is in /netlib/f2c/libf2c.zip, e.g.,
+
+		http://www.netlib.org/f2c/libf2c.zip
+*/
+
+#include "f2c.h"
+
+/* Common Block Declarations */
+
+struct {
+    doublereal memd[1];
+} mem_;
+
+#define mem_1 mem_
+
+struct {
+    integer rightn, leftmn, permrn, permln, inmagg, outmag, soflag, foflag, 
+	    justiy, nls, lsindt, shnskp, ihnskp, ihindt, nhnskp, nhlevl[10];
+    logical standd;
+} lrfcom_;
+
+#define lrfcom_1 lrfcom_
+
+struct {
+    logical xerflg, xerpad[84];
+} xercom_;
+
+#define xercom_1 xercom_
+
+/* Table of constant values */
+
+static integer c__2046 = 2046;
+static integer c__2 = 2;
+static integer c__1023 = 1023;
+static integer c__255 = 255;
+static integer c__256 = 256;
+static integer c__9 = 9;
+static integer c__0 = 0;
+static integer c__10 = 10;
+static integer c__1 = 1;
+static logical c_true = TRUE_;
+static integer c__3 = 3;
+
+/* Subroutine */ int lroffl_(in, out, module, parstr, center, lsblok, sectin)
+integer *in, *out;
+shortint *module, *parstr, *center, *lsblok, *sectin;
+{
+    /* Initialized data */
+
+    static shortint st0003[5] = { 60,80,62,10,0 };
+    static shortint st0004[7] = { 94,46,104,101,108,112,0 };
+    static shortint st0005[8] = { 60,47,80,82,69,62,10,0 };
+    static shortint st0006[7] = { 60,80,82,69,62,10,0 };
+    static shortint st0007[13] = { 60,47,68,68,62,10,60,47,68,76,62,10,0 };
+    static shortint st0008[13] = { 60,47,68,68,62,10,60,47,68,76,62,10,0 };
+    static shortint st0009[2] = { 124,0 };
+    static shortint st0010[13] = { 60,47,68,68,62,10,60,47,68,76,62,10,0 };
+    static shortint st0011[25] = { 60,67,69,78,84,69,82,62,37,115,60,47,67,69,
+	    78,84,69,82,62,60,66,82,62,10,0 };
+    static shortint st0012[6] = { 60,66,82,62,10,0 };
+    static shortint st0013[6] = { 60,66,82,62,10,0 };
+    static shortint st0014[6] = { 60,66,82,62,10,0 };
+    static shortint st0015[13] = { 60,68,76,62,10,60,68,84,62,60,66,62,0 };
+    static shortint st0016[22] = { 60,65,32,78,65,77,69,61,34,108,95,37,115,
+	    34,62,37,115,60,47,65,62,0 };
+    static shortint st0017[11] = { 60,47,66,62,60,47,68,84,62,10,0 };
+    static shortint st0018[46] = { 60,33,32,83,101,99,61,37,115,32,76,101,118,
+	    101,108,61,37,100,32,76,97,98,101,108,61,39,37,115,39,32,76,105,
+	    110,101,61,39,37,115,39,62,10,60,68,68,62,0 };
+    static shortint st0019[5] = { 78,111,110,101,0 };
+    static shortint st0020[2] = { 32,0 };
+    static shortint st0021[13] = { 60,47,68,68,62,10,60,47,68,76,62,10,0 };
+    static shortint st0022[21] = { 60,65,32,72,82,69,70,61,34,37,115,34,62,37,
+	    115,60,47,65,62,10,0 };
+    static shortint st0023[19] = { 60,65,32,78,65,77,69,61,34,37,115,34,62,60,
+	    47,65,62,10,0 };
+    static shortint st0024[7] = { 60,80,82,69,62,10,0 };
+    static shortint st0025[8] = { 60,47,80,82,69,62,10,0 };
+    static shortint st0026[5] = { 39,37,115,39,0 };
+    static shortint st0027[7] = { 60,47,85,76,62,10,0 };
+    static shortint st0028[22] = { 60,33,32,69,110,100,83,101,99,116,105,111,
+	    110,58,32,32,32,37,115,62,10,0 };
+    static shortint st0029[32] = { 60,72,50,62,60,65,32,78,65,77,69,61,34,115,
+	    95,37,115,34,62,37,115,60,47,65,62,60,47,72,50,62,10,0 };
+    static shortint st0030[35] = { 60,72,50,62,60,65,32,78,65,77,69,61,34,115,
+	    95,37,115,34,62,37,115,32,37,115,60,47,65,62,60,47,72,50,62,10,0 }
+	    ;
+    static shortint st0031[24] = { 60,33,32,66,101,103,105,110,83,101,99,116,
+	    105,111,110,58,32,39,37,115,39,62,10,0 };
+    static shortint st0032[6] = { 60,85,76,62,10,0 };
+    static shortint st0033[3] = { 37,115,0 };
+    static shortint st0034[30] = { 60,47,85,76,62,10,60,33,32,69,110,100,83,
+	    101,99,116,105,111,110,58,32,32,32,32,37,115,62,10,10,0 };
+    static shortint st0035[14] = { 60,33,32,67,111,110,116,101,110,116,115,58,
+	    32,0 };
+    static shortint st0036[4] = { 37,115,32,0 };
+    static shortint st0037[5] = { 32,62,10,10,0 };
+    static shortint st0038[17] = { 60,47,66,79,68,89,62,10,60,47,72,84,77,76,
+	    62,10,0 };
+    static shortint st0001[24] = { 60,84,73,84,76,69,62,37,115,60,47,84,73,84,
+	    76,69,62,10,60,85,76,62,10,0 };
+    static shortint st0002[2] = { 32,0 };
+
+    /* System generated locals */
+    integer i__1;
+
+    /* Local variables */
+    static integer i__, ip, sp, cmd, arg;
+#define memb ((logical *)&mem_1)
+#define memc ((shortint *)&mem_1)
+    static integer name__, ibuf;
+#define memi ((integer *)&mem_1)
+    static integer nsec;
+#define meml ((integer *)&mem_1)
+    static integer sw0001, sw0002;
+#define memr ((real *)&mem_1)
+#define mems ((shortint *)&mem_1)
+#define memx ((complex *)&mem_1)
+    static integer font, sptr;
+    extern /* Subroutine */ int aclrc_(), pargi_(), sfree_();
+    static integer level, unesc;
+    extern /* Subroutine */ int smark_();
+    static integer indend;
+    extern integer lgetag_();
+    extern /* Subroutine */ int lhesce_();
+    static logical formad;
+    extern integer getlie_(), lhfink_(), lhfinn_();
+    extern /* Subroutine */ int salloc_(), lhmkne_();
+    static integer lastle;
+    static logical format, quitae, quitah;
+    extern integer nextcd_(), strmah_();
+    static integer lslevl;
+    extern integer strids_();
+    extern /* Subroutine */ int amovki_(), ungete_(), lhprog_(), fprinf_(), 
+	    pargsr_(), lhsetl_(), sprinf_(), xffluh_(), zzepro_();
+    extern integer xstrln_();
+    extern /* Subroutine */ int xstrcy_();
+
+    /* Parameter adjustments */
+    --sectin;
+    --lsblok;
+    --center;
+    --parstr;
+    --module;
+
+    /* Function Body */
+    smark_(&sp);
+    salloc_(&ibuf, &c__2046, &c__2);
+    salloc_(&unesc, &c__2046, &c__2);
+    salloc_(&name__, &c__1023, &c__2);
+    salloc_(&level, &c__255, &c__2);
+    salloc_(&sptr, &c__256, &c__9);
+    aclrc_(&memc[ibuf - 1], &c__2046);
+    aclrc_(&memc[name__ - 1], &c__1023);
+    aclrc_(&memc[unesc - 1], &c__2046);
+    aclrc_(&memc[level - 1], &c__255);
+    lastle = 3;
+    font = 1;
+    indend = 1;
+    nsec = 0;
+    lslevl = 0;
+    format = TRUE_;
+    quitae = FALSE_;
+    quitah = FALSE_;
+    formad = FALSE_;
+    amovki_(&c__0, lrfcom_1.nhlevl, &c__10);
+L110:
+    if (! (getlie_(in, &memc[ibuf - 1]) == -2)) {
+	goto L120;
+    }
+    goto L98;
+L120:
+    ip = 1;
+L130:
+    if (! (memc[ibuf + ip - 2] == 32 || memc[ibuf + ip - 2] == 9)) {
+	goto L132;
+    }
+/* L131: */
+    ++ip;
+    goto L130;
+L132:
+/* L111: */
+    if (! (memc[ibuf + ip - 2] != 10)) {
+	goto L110;
+    }
+/* L112: */
+    ungete_(in, &memc[ibuf - 1]);
+    if (! (memc[ibuf - 1] == 46)) {
+	goto L140;
+    }
+    formad = TRUE_;
+L140:
+    if (! (sectin[1] != 0)) {
+	goto L150;
+    }
+    if (! (lhfinn_(in, &formad, &sectin[1]) == -2)) {
+	goto L160;
+    }
+    goto L98;
+L160:
+    goto L151;
+L150:
+    if (! (lsblok[1] != 0)) {
+	goto L170;
+    }
+    if (! (lhfink_(in, &formad, &lsblok[1]) == -2)) {
+	goto L180;
+    }
+    goto L98;
+L180:
+    quitae = TRUE_;
+L170:
+L151:
+    lhprog_(out, &module[1], &parstr[1], &center[1]);
+    fprinf_(out, st0001);
+    if (! (lsblok[1] != 0)) {
+	goto L190;
+    }
+    pargsr_(&lsblok[1]);
+    goto L191;
+L190:
+    if (! (sectin[1] != 0)) {
+	goto L200;
+    }
+    pargsr_(&sectin[1]);
+    goto L201;
+L200:
+    if (! (module[1] != 0)) {
+	goto L210;
+    }
+    pargsr_(&module[1]);
+    goto L211;
+L210:
+    pargsr_(st0002);
+L211:
+L201:
+L191:
+L220:
+    if (! (getlie_(in, &memc[ibuf - 1]) != -2)) {
+	goto L221;
+    }
+    memc[ibuf + xstrln_(&memc[ibuf - 1]) - 2] = 0;
+    xstrcy_(&memc[ibuf - 1], &memc[unesc - 1], &c__1023);
+    lhesce_(&memc[ibuf - 1], &font, &format, &c__0, &c__1023);
+    sw0001 = memc[ibuf - 1];
+    goto L230;
+L240:
+    fprinf_(out, st0003);
+    goto L231;
+L250:
+    if (! (strmah_(&memc[ibuf - 1], st0004) > 0)) {
+	goto L260;
+    }
+    goto L220;
+L260:
+    ip = 1;
+    lastle = 3;
+    cmd = nextcd_(&memc[ibuf - 1], &ip);
+L270:
+    if (! (memc[ibuf + ip - 1] == 32 || memc[ibuf + ip - 1] == 9)) {
+	goto L271;
+    }
+    ++ip;
+    goto L270;
+L271:
+    sw0002 = cmd;
+    goto L280;
+L290:
+    fprinf_(out, st0005);
+    format = TRUE_;
+    goto L281;
+L300:
+    fprinf_(out, st0006);
+    format = FALSE_;
+    goto L281;
+L310:
+    goto L220;
+L320:
+    goto L220;
+L330:
+    goto L220;
+L340:
+    if (! (lslevl > 0)) {
+	goto L350;
+    }
+    fprinf_(out, st0007);
+    lslevl = 0;
+L350:
+    lastle = 1;
+    memc[level - 1] = 0;
+    goto L220;
+L360:
+    if (! (lslevl > 0)) {
+	goto L370;
+    }
+    fprinf_(out, st0008);
+    lslevl = 0;
+L370:
+    lastle = 1;
+    memc[level - 1] = 0;
+    if (! quitah) {
+	goto L380;
+    }
+    if (! (strids_(st0009, &sectin[1]) > 0)) {
+	goto L390;
+    }
+    quitah = FALSE_;
+    ungete_(in, &memc[ibuf - 1]);
+    if (! (lhfinn_(in, &formad, &sectin[1]) == -2)) {
+	goto L400;
+    }
+    goto L221;
+L400:
+    goto L391;
+L390:
+    goto L221;
+L391:
+L380:
+    goto L220;
+L410:
+    if (! (lslevl > 0)) {
+	goto L420;
+    }
+    fprinf_(out, st0010);
+    lslevl = 0;
+L420:
+    i__1 = lgetag_(&memc[ibuf - 1], &ip, &c__1);
+    lhsetl_(&i__1, &memc[level - 1]);
+    lastle = 1;
+    goto L220;
+L430:
+    if (! (getlie_(in, &memc[ibuf - 1]) == -2)) {
+	goto L440;
+    }
+    goto L221;
+L440:
+    lhesce_(&memc[ibuf - 1], &font, &c_true, &c__0, &c__1023);
+    fprinf_(out, st0011);
+    pargsr_(&memc[ibuf - 1]);
+/* L441: */
+    goto L281;
+L450:
+    fprinf_(out, st0012);
+    goto L281;
+L460:
+    arg = lgetag_(&memc[ibuf - 1], &ip, &c__1);
+    fprinf_(out, st0013);
+    i__ = 1;
+L470:
+    if (! (i__ < arg)) {
+	goto L472;
+    }
+    fprinf_(out, st0014);
+/* L471: */
+    ++i__;
+    goto L470;
+L472:
+    goto L281;
+L480:
+    goto L220;
+L490:
+    arg = lgetag_(&memc[ibuf - 1], &ip, &c__0);
+    if (! (arg == 0)) {
+	goto L500;
+    }
+    ip = 5;
+L500:
+    xstrcy_(&memc[ibuf + ip - 2], &memc[name__ - 1], &c__1023);
+    i__ = 0;
+L510:
+    if (! (memc[name__ + i__ - 1] >= 65 && memc[name__ + i__ - 1] <= 90 || 
+	    memc[name__ + i__ - 1] >= 97 && memc[name__ + i__ - 1] <= 122 || 
+	    memc[name__ + i__ - 1] >= 48 && memc[name__ + i__ - 1] <= 57 || 
+	    memc[name__ + i__ - 1] == 95)) {
+	goto L512;
+    }
+/* L511: */
+    ++i__;
+    goto L510;
+L512:
+    memc[name__ + i__ - 1] = 0;
+    memc[ibuf + ip + xstrln_(&memc[ibuf + ip - 1]) - 2] = 0;
+    fprinf_(out, st0015);
+    fprinf_(out, st0016);
+    pargsr_(&memc[name__ - 1]);
+    pargsr_(&memc[ibuf + ip - 2]);
+    fprinf_(out, st0017);
+    lhesce_(&memc[unesc + ip - 2], &font, &c_true, &c__1, &c__1023);
+    memc[unesc + xstrln_(&memc[unesc - 1]) - 2] = 0;
+    fprinf_(out, st0018);
+    if (! (nsec > 0)) {
+	goto L520;
+    }
+    pargsr_(&memc[memi[sptr + nsec - 2] - 1]);
+    goto L521;
+L520:
+    pargsr_(st0019);
+L521:
+    pargi_(&lslevl);
+    pargsr_(&memc[name__ - 1]);
+    if (! (memc[unesc + ip - 2] == 10)) {
+	goto L530;
+    }
+    pargsr_(st0020);
+    goto L531;
+L530:
+    pargsr_(&memc[unesc + ip - 2]);
+L531:
+    ++lslevl;
+    goto L281;
+L540:
+    fprinf_(out, st0021);
+    --lslevl;
+    if (! quitae) {
+	goto L550;
+    }
+    goto L221;
+L550:
+    goto L281;
+L560:
+    memc[ibuf + ip + xstrln_(&memc[ibuf + ip - 1]) - 2] = 0;
+    i__ = 0;
+L570:
+    if (memc[ibuf + ip - 1] == 32 || memc[ibuf + ip - 1] == 9) {
+	goto L572;
+    }
+    memc[name__ + i__ - 1] = memc[ibuf + ip - 1];
+    ++i__;
+/* L571: */
+    ++ip;
+    goto L570;
+L572:
+    memc[name__ + i__ - 1] = 0;
+    fprinf_(out, st0022);
+    pargsr_(&memc[name__ - 1]);
+    pargsr_(&memc[ibuf + ip]);
+    goto L281;
+L580:
+    memc[ibuf + ip + xstrln_(&memc[ibuf + ip - 1]) - 2] = 0;
+    fprinf_(out, st0023);
+    pargsr_(&memc[ibuf + ip - 1]);
+    goto L281;
+L590:
+    goto L220;
+L600:
+    goto L220;
+L610:
+    fprinf_(out, st0024);
+    format = FALSE_;
+    goto L281;
+L620:
+    fprinf_(out, st0025);
+    format = TRUE_;
+    goto L281;
+L630:
+    goto L221;
+L280:
+    if (sw0002 < 1 || sw0002 > 21) {
+	goto L281;
+    }
+    switch ((int)sw0002) {
+	case 1:  goto L290;
+	case 2:  goto L300;
+	case 3:  goto L310;
+	case 4:  goto L320;
+	case 5:  goto L330;
+	case 6:  goto L340;
+	case 7:  goto L360;
+	case 8:  goto L410;
+	case 9:  goto L450;
+	case 10:  goto L430;
+	case 11:  goto L460;
+	case 12:  goto L480;
+	case 13:  goto L490;
+	case 14:  goto L540;
+	case 15:  goto L590;
+	case 16:  goto L600;
+	case 17:  goto L610;
+	case 18:  goto L620;
+	case 19:  goto L560;
+	case 20:  goto L580;
+	case 21:  goto L630;
+    }
+L281:
+    goto L231;
+L640:
+    if (! (lastle == 1)) {
+	goto L650;
+    }
+    salloc_(&memi[sptr + nsec - 1], &c__1023, &c__2);
+    aclrc_(&memc[memi[sptr + nsec - 1] - 1], &c__1023);
+    memc[ibuf + xstrln_(&memc[ibuf - 1]) - 2] = 0;
+    sprinf_(&memc[memi[sptr + nsec - 1] - 1], &c__1023, st0026);
+    pargsr_(&memc[ibuf - 1]);
+    if (! (indend == 1)) {
+	goto L660;
+    }
+    fprinf_(out, st0027);
+L660:
+    if (! (nsec > 0)) {
+	goto L670;
+    }
+    fprinf_(out, st0028);
+    pargsr_(&memc[memi[sptr + nsec - 2] - 1]);
+L670:
+    lhmkne_(&memc[ibuf - 1], &memc[name__ - 1]);
+    if (! (memc[level - 1] == 0)) {
+	goto L680;
+    }
+    fprinf_(out, st0029);
+    pargsr_(&memc[name__ - 1]);
+    pargsr_(&memc[ibuf - 1]);
+    goto L681;
+L680:
+    fprinf_(out, st0030);
+    pargsr_(&memc[name__ - 1]);
+    pargsr_(&memc[level - 1]);
+    pargsr_(&memc[ibuf - 1]);
+    memc[level - 1] = 0;
+L681:
+    fprinf_(out, st0031);
+    pargsr_(&memc[ibuf - 1]);
+    if (! (indend == 1)) {
+	goto L690;
+    }
+    fprinf_(out, st0032);
+L690:
+    lastle = 2;
+    ++nsec;
+    if (! (sectin[1] != 0)) {
+	goto L700;
+    }
+    quitah = TRUE_;
+L700:
+    goto L651;
+L650:
+/* L99: */
+    fprinf_(out, st0033);
+    pargsr_(&memc[ibuf - 1]);
+    lastle = 3;
+L651:
+    goto L231;
+L230:
+    if (sw0001 == 10) {
+	goto L240;
+    }
+    if (sw0001 == 46) {
+	goto L250;
+    }
+    goto L640;
+L231:
+    aclrc_(&memc[ibuf - 1], &c__2046);
+    aclrc_(&memc[unesc - 1], &c__2046);
+    aclrc_(&memc[name__ - 1], &c__1023);
+    goto L220;
+L221:
+    if (! (nsec > 0)) {
+	goto L710;
+    }
+    fprinf_(out, st0034);
+    pargsr_(&memc[memi[sptr + nsec - 2] - 1]);
+L710:
+    fprinf_(out, st0035);
+    i__ = 0;
+L720:
+    if (! (i__ < nsec)) {
+	goto L722;
+    }
+    fprinf_(out, st0036);
+    pargsr_(&memc[memi[sptr + i__ - 1] - 1]);
+/* L721: */
+    ++i__;
+    goto L720;
+L722:
+    fprinf_(out, st0037);
+    fprinf_(out, st0038);
+    xffluh_(out);
+L98:
+    sfree_(&sp);
+/* L100: */
+    zzepro_();
+    return 0;
+} /* lroffl_ */
+
+#undef memx
+#undef mems
+#undef memr
+#undef meml
+#undef memi
+#undef memc
+#undef memb
+
+
+/* Subroutine */ int lhprog_(fd, mod, date, title)
+integer *fd;
+shortint *mod, *date, *title;
+{
+    /* Initialized data */
+
+    static shortint st0001[15] = { 60,72,84,77,76,62,10,60,66,79,68,89,62,10,
+	    0 };
+    static shortint st0002[36] = { 60,84,65,66,76,69,32,87,73,68,84,72,61,34,
+	    49,48,48,37,37,34,32,66,79,82,68,69,82,61,48,62,60,84,82,62,10,0 }
+	    ;
+    static shortint st0003[30] = { 60,84,68,32,65,76,73,71,78,61,76,69,70,84,
+	    62,60,70,79,78,84,32,83,73,90,69,61,52,62,10,0 };
+    static shortint st0004[10] = { 60,66,62,37,115,60,47,66,62,0 };
+    static shortint st0005[15] = { 60,66,62,37,115,32,40,37,115,41,60,47,66,
+	    62,0 };
+    static shortint st0006[14] = { 60,47,70,79,78,84,62,60,47,84,68,62,10,0 };
+    static shortint st0007[32] = { 60,84,68,32,65,76,73,71,78,61,67,69,78,84,
+	    69,82,62,60,70,79,78,84,32,83,73,90,69,61,52,62,10,0 };
+    static shortint st0008[11] = { 60,66,62,37,115,60,47,66,62,10,0 };
+    static shortint st0009[14] = { 60,47,70,79,78,84,62,60,47,84,68,62,10,0 };
+    static shortint st0010[31] = { 60,84,68,32,65,76,73,71,78,61,82,73,71,72,
+	    84,62,60,70,79,78,84,32,83,73,90,69,61,52,62,10,0 };
+    static shortint st0011[10] = { 60,66,62,37,115,60,47,66,62,0 };
+    static shortint st0012[15] = { 60,66,62,37,115,32,40,37,115,41,60,47,66,
+	    62,0 };
+    static shortint st0013[14] = { 60,47,70,79,78,84,62,60,47,84,68,62,10,0 };
+    static shortint st0014[18] = { 60,47,84,82,62,60,47,84,65,66,76,69,62,60,
+	    80,62,10,0 };
+
+    extern /* Subroutine */ int fprinf_(), pargsr_(), zzepro_();
+
+    /* Parameter adjustments */
+    --title;
+    --date;
+    --mod;
+
+    /* Function Body */
+    fprinf_(fd, st0001);
+    if (! (date[1] == 0 && title[1] == 0)) {
+	goto L110;
+    }
+    goto L100;
+L110:
+    fprinf_(fd, st0002);
+    fprinf_(fd, st0003);
+    if (! (date[1] == 0)) {
+	goto L120;
+    }
+    fprinf_(fd, st0004);
+    pargsr_(&mod[1]);
+    goto L121;
+L120:
+    fprinf_(fd, st0005);
+    pargsr_(&mod[1]);
+    pargsr_(&date[1]);
+L121:
+    fprinf_(fd, st0006);
+    if (! (title[1] != 0)) {
+	goto L130;
+    }
+    fprinf_(fd, st0007);
+    fprinf_(fd, st0008);
+    pargsr_(&title[1]);
+    fprinf_(fd, st0009);
+L130:
+    fprinf_(fd, st0010);
+    if (! (date[1] == 0)) {
+	goto L140;
+    }
+    fprinf_(fd, st0011);
+    pargsr_(&mod[1]);
+    goto L141;
+L140:
+    fprinf_(fd, st0012);
+    pargsr_(&mod[1]);
+    pargsr_(&date[1]);
+L141:
+    fprinf_(fd, st0013);
+    fprinf_(fd, st0014);
+L100:
+    zzepro_();
+    return 0;
+} /* lhprog_ */
+
+/* Subroutine */ int lhesce_(str, font, format, speciy, maxch)
+shortint *str;
+integer *font;
+logical *format;
+integer *speciy, *maxch;
+{
+    /* Initialized data */
+
+    static shortint st0013[5] = { 60,47,66,62,0 };
+    static shortint st0014[4] = { 60,73,62,0 };
+    static shortint st0015[5] = { 60,47,66,62,0 };
+    static shortint st0016[5] = { 60,47,73,62,0 };
+    static shortint st0017[5] = { 60,47,66,62,0 };
+    static shortint st0018[5] = { 60,47,73,62,0 };
+    static shortint st0019[5] = { 60,66,82,62,0 };
+    static shortint st0020[3] = { 10,0,0 };
+    static shortint st0001[4] = { 60,62,38,0 };
+    static shortint st0002[5] = { 38,108,116,59,0 };
+    static shortint st0003[5] = { 38,103,116,59,0 };
+    static shortint st0004[6] = { 38,97,109,112,59,0 };
+    static shortint st0005[5] = { 60,84,84,62,0 };
+    static shortint st0006[6] = { 60,47,84,84,62,0 };
+    static shortint st0007[5] = { 60,84,84,62,0 };
+    static shortint st0008[6] = { 60,47,84,84,62,0 };
+    static shortint st0009[7] = { 60,47,84,84,62,34,0 };
+    static shortint st0010[6] = { 34,60,84,84,62,0 };
+    static shortint st0011[5] = { 60,47,73,62,0 };
+    static shortint st0012[4] = { 60,66,62,0 };
+
+    /* Local variables */
+    static integer i__, ip, sp, buf;
+#define memb ((logical *)&mem_1)
+#define memc ((shortint *)&mem_1)
+#define memi ((integer *)&mem_1)
+#define meml ((integer *)&mem_1)
+    static integer sw0001;
+#define memr ((real *)&mem_1)
+#define mems ((shortint *)&mem_1)
+#define memx ((complex *)&mem_1)
+    extern /* Subroutine */ int aclrc_(), sfree_(), amovc_(), smark_(), 
+	    salloc_();
+    static integer keywod;
+    extern integer gstrcy_(), stridx_();
+    extern /* Subroutine */ int zzepro_();
+
+    /* Parameter adjustments */
+    --str;
+
+    /* Function Body */
+    smark_(&sp);
+    salloc_(&buf, maxch, &c__2);
+    salloc_(&keywod, maxch, &c__2);
+    aclrc_(&memc[buf - 1], maxch);
+    aclrc_(&memc[keywod - 1], maxch);
+    ip = buf;
+    i__ = 1;
+L110:
+    if (! (str[i__] != 0 && i__ <= *maxch)) {
+	goto L112;
+    }
+    if (! (*speciy == 1 && stridx_(&str[i__], st0001) == 0)) {
+	goto L120;
+    }
+    goto L90;
+L120:
+    sw0001 = str[i__];
+    goto L130;
+L140:
+    ip += gstrcy_(st0002, &memc[ip - 1], &c__1023);
+    goto L131;
+L150:
+    ip += gstrcy_(st0003, &memc[ip - 1], &c__1023);
+    goto L131;
+L160:
+    ip += gstrcy_(st0004, &memc[ip - 1], &c__1023);
+    goto L131;
+L170:
+    if (! (str[i__ + 2] == 39)) {
+	goto L180;
+    }
+    ip += gstrcy_(st0005, &memc[ip - 1], &c__1023);
+    ip += gstrcy_(&str[i__], &memc[ip - 1], &c__3);
+    ip += gstrcy_(st0006, &memc[ip - 1], &c__1023);
+    i__ += 2;
+    goto L181;
+L180:
+    goto L90;
+L181:
+    goto L131;
+L190:
+    if (! (str[i__ + 2] == 96 || str[i__ + 2] == 39)) {
+	goto L200;
+    }
+    ip += gstrcy_(st0007, &memc[ip - 1], &c__1023);
+    ip += gstrcy_(&str[i__], &memc[ip - 1], &c__3);
+    ip += gstrcy_(st0008, &memc[ip - 1], &c__1023);
+    i__ += 2;
+    goto L201;
+L200:
+    goto L90;
+L201:
+    goto L131;
+L210:
+    if (! (*format && str[i__ + 1] != 47 && str[i__ + 2] != 47)) {
+	goto L220;
+    }
+    if (! (*font == 5)) {
+	goto L230;
+    }
+    ip += gstrcy_(st0009, &memc[ip - 1], &c__1023);
+    *font = 1;
+    goto L231;
+L230:
+    if (! (*font == 1)) {
+	goto L240;
+    }
+    ip += gstrcy_(st0010, &memc[ip - 1], &c__1023);
+    *font = 5;
+    goto L241;
+L240:
+    goto L90;
+L241:
+L231:
+    goto L221;
+L220:
+    goto L90;
+L221:
+    goto L131;
+L250:
+    if (! (str[i__ + 1] == 102)) {
+	goto L260;
+    }
+    if (! (str[i__ + 2] == 66)) {
+	goto L270;
+    }
+    if (! (*font == 3)) {
+	goto L280;
+    }
+    goto L111;
+L280:
+    if (! (*font == 2)) {
+	goto L290;
+    }
+    ip += gstrcy_(st0011, &memc[ip - 1], &c__1023);
+L290:
+    ip += gstrcy_(st0012, &memc[ip - 1], &c__1023);
+    *font = 3;
+    goto L271;
+L270:
+    if (! (str[i__ + 2] == 73)) {
+	goto L300;
+    }
+    if (! (*font == 2)) {
+	goto L310;
+    }
+    goto L111;
+L310:
+    if (! (*font == 3)) {
+	goto L320;
+    }
+    ip += gstrcy_(st0013, &memc[ip - 1], &c__1023);
+L320:
+    ip += gstrcy_(st0014, &memc[ip - 1], &c__1023);
+    *font = 2;
+    goto L301;
+L300:
+    if (! (str[i__ + 2] == 82)) {
+	goto L330;
+    }
+    if (! (*font == 3)) {
+	goto L340;
+    }
+    ip += gstrcy_(st0015, &memc[ip - 1], &c__1023);
+    goto L341;
+L340:
+    if (! (*font == 2)) {
+	goto L350;
+    }
+    ip += gstrcy_(st0016, &memc[ip - 1], &c__1023);
+L350:
+L341:
+    *font = 1;
+    goto L331;
+L330:
+    if (! (str[i__ + 2] == 80)) {
+	goto L360;
+    }
+    if (! (*font == 3)) {
+	goto L370;
+    }
+    ip += gstrcy_(st0017, &memc[ip - 1], &c__1023);
+    goto L371;
+L370:
+    if (! (*font == 2)) {
+	goto L380;
+    }
+    ip += gstrcy_(st0018, &memc[ip - 1], &c__1023);
+L380:
+L371:
+    *font = 1;
+L360:
+L331:
+L301:
+L271:
+    i__ += 2;
+    goto L261;
+L260:
+    if (! (str[i__ + 1] == 10 || str[i__ + 1] == 0)) {
+	goto L390;
+    }
+    memc[ip - 1] = str[i__];
+    ++ip;
+    ++i__;
+    ip += gstrcy_(st0019, &memc[ip - 1], &c__1023);
+    goto L391;
+L390:
+    goto L90;
+L391:
+L261:
+    goto L131;
+L400:
+L90:
+    memc[ip - 1] = str[i__];
+    ++ip;
+    goto L131;
+L130:
+    if (sw0001 == 34) {
+	goto L210;
+    }
+    if (sw0001 == 38) {
+	goto L160;
+    }
+    if (sw0001 == 39) {
+	goto L170;
+    }
+    if (sw0001 == 60) {
+	goto L140;
+    }
+    if (sw0001 == 62) {
+	goto L150;
+    }
+    if (sw0001 == 92) {
+	goto L250;
+    }
+    if (sw0001 == 96) {
+	goto L190;
+    }
+    goto L400;
+L131:
+L111:
+    ++i__;
+    goto L110;
+L112:
+    ip += gstrcy_(st0020, &memc[ip - 1], &c__1023);
+    amovc_(&memc[buf - 1], &str[1], maxch);
+    sfree_(&sp);
+/* L100: */
+    zzepro_();
+    return 0;
+} /* lhesce_ */
+
+#undef memx
+#undef mems
+#undef memr
+#undef meml
+#undef memi
+#undef memc
+#undef memb
+
+
+/* Subroutine */ int lhsetl_(n, level)
+integer *n;
+shortint *level;
+{
+    /* Initialized data */
+
+    static shortint st0001[4] = { 37,100,46,0 };
+
+    /* System generated locals */
+    integer i__1;
+
+    /* Local variables */
+    static integer i__;
+    extern /* Subroutine */ int pargi_(), amovki_(), sprinf_(), zzepro_();
+    extern integer xstrln_();
+
+    /* Parameter adjustments */
+    --level;
+
+    /* Function Body */
+    ++lrfcom_1.nhlevl[(60 + (0 + (*n - 1 << 2)) - 60) / 4];
+    i__1 = 10 - *n;
+    amovki_(&c__0, &lrfcom_1.nhlevl[*n], &i__1);
+    level[1] = 0;
+    i__1 = *n;
+    for (i__ = 1; i__ <= i__1; ++i__) {
+	sprinf_(&level[xstrln_(&level[1]) + 1], &c__2046, st0001);
+	pargi_(&lrfcom_1.nhlevl[i__ - 1]);
+/* L110: */
+    }
+/* L111: */
+    if (! (*n > 1 && level[xstrln_(&level[1])] == 46)) {
+	goto L120;
+    }
+    level[xstrln_(&level[1])] = 0;
+L120:
+/* L100: */
+    zzepro_();
+    return 0;
+} /* lhsetl_ */
+
+integer lhfink_(fd, formad, param)
+integer *fd;
+logical *formad;
+shortint *param;
+{
+    /* Initialized data */
+
+    static shortint st0001[5] = { 123,37,115,125,0 };
+    static shortint st0002[7] = { 94,46,123,108,115,125,0 };
+    static shortint st0003[7] = { 94,35,123,37,115,125,0 };
+
+    /* System generated locals */
+    integer ret_val;
+
+    /* Local variables */
+    static integer sp, len;
+#define memb ((logical *)&mem_1)
+#define memc ((shortint *)&mem_1)
+#define memi ((integer *)&mem_1)
+    static integer lbuf;
+#define meml ((integer *)&mem_1)
+#define memr ((real *)&mem_1)
+#define mems ((shortint *)&mem_1)
+#define memx ((complex *)&mem_1)
+    extern /* Subroutine */ int sfree_(), smark_();
+    static logical matchd;
+    extern integer getlie_();
+    extern /* Subroutine */ int salloc_(), ungete_();
+    static integer patten;
+    extern integer strmah_();
+    extern /* Subroutine */ int pargsr_(), sprinf_(), zzepro_();
+    extern integer xstrln_();
+
+    /* Parameter adjustments */
+    --param;
+
+    /* Function Body */
+    smark_(&sp);
+    salloc_(&patten, &c__255, &c__2);
+    salloc_(&lbuf, &c__1023, &c__2);
+    matchd = FALSE_;
+    if (! (getlie_(fd, &memc[lbuf - 1]) == -2) && ! xercom_1.xerflg) {
+	goto L110;
+    }
+    if (xercom_1.xerflg) {
+	goto L100;
+    }
+    goto L90;
+L110:
+    if (! (*formad)) {
+	goto L120;
+    }
+    sprinf_(&memc[patten - 1], &c__255, st0001);
+    pargsr_(&param[1]);
+L130:
+    if (! (strmah_(&memc[lbuf - 1], st0002) > 0)) {
+	goto L140;
+    }
+    if (! (strmah_(&memc[lbuf - 1], &memc[patten - 1]) > 0)) {
+	goto L150;
+    }
+    matchd = TRUE_;
+    goto L132;
+L150:
+L140:
+/* L131: */
+    if (! (getlie_(fd, &memc[lbuf - 1]) == -2) && ! xercom_1.xerflg) {
+	goto L130;
+    }
+    if (xercom_1.xerflg) {
+	goto L100;
+    }
+L132:
+    goto L121;
+L120:
+    sprinf_(&memc[patten - 1], &c__255, st0003);
+    pargsr_(&param[1]);
+L160:
+    if (! (strmah_(&memc[lbuf - 1], &memc[patten - 1]) > 0)) {
+	goto L170;
+    }
+    matchd = TRUE_;
+    goto L162;
+L170:
+/* L161: */
+    if (! (getlie_(fd, &memc[lbuf - 1]) == -2) && ! xercom_1.xerflg) {
+	goto L160;
+    }
+    if (xercom_1.xerflg) {
+	goto L100;
+    }
+L162:
+L121:
+    ungete_(fd, &memc[lbuf - 1]);
+L90:
+    len = xstrln_(&memc[lbuf - 1]);
+    sfree_(&sp);
+    if (! matchd) {
+	goto L180;
+    }
+    ret_val = len;
+    goto L100;
+L180:
+    ret_val = -2;
+    goto L100;
+/* L181: */
+L100:
+    zzepro_();
+    return ret_val;
+} /* lhfink_ */
+
+#undef memx
+#undef mems
+#undef memr
+#undef meml
+#undef memi
+#undef memc
+#undef memb
+
+
+integer lhfinn_(fd, formad, sectis)
+integer *fd;
+logical *formad;
+shortint *sectis;
+{
+    /* Initialized data */
+
+    static shortint st0001[7] = { 94,46,123,105,104,125,0 };
+    static shortint st0002[5] = { 46,105,104,10,0 };
+
+    /* System generated locals */
+    integer ret_val, i__1, i__2;
+
+    /* Local variables */
+    static integer ip, op, sp;
+#define memb ((logical *)&mem_1)
+#define memc ((shortint *)&mem_1)
+#define memi ((integer *)&mem_1)
+    static shortint lbuf[1024];
+#define meml ((integer *)&mem_1)
+    static integer sw0001;
+#define memr ((real *)&mem_1)
+#define mems ((shortint *)&mem_1)
+    static integer npat;
+#define memx ((complex *)&mem_1)
+    extern /* Subroutine */ int sfree_(), smark_();
+    static logical matchd;
+    extern integer getlie_();
+    extern logical lhmath_();
+    static integer patoff[10];
+    extern /* Subroutine */ int salloc_();
+    static integer patbuf;
+    extern /* Subroutine */ int ungete_();
+    extern integer strmah_();
+    extern /* Subroutine */ int zzepro_();
+
+    /* Parameter adjustments */
+    --sectis;
+
+    /* Function Body */
+    smark_(&sp);
+    salloc_(&patbuf, &c__1023, &c__2);
+    npat = 1;
+    op = patbuf;
+    patoff[0] = op;
+    if (! (getlie_(fd, lbuf) == -2) && ! xercom_1.xerflg) {
+	goto L110;
+    }
+    if (xercom_1.xerflg) {
+	goto L100;
+    }
+    goto L91;
+L110:
+    ip = 1;
+L120:
+    if (! (sectis[ip] != 0)) {
+	goto L122;
+    }
+    sw0001 = sectis[ip];
+    goto L130;
+L140:
+    memc[op - 1] = 0;
+    ++op;
+/* Computing MIN */
+    i__1 = 10, i__2 = npat + 1;
+    npat = min(i__1,i__2);
+    patoff[npat - 1] = op;
+    goto L131;
+L150:
+    memc[op - 1] = sectis[ip];
+    ++op;
+    goto L131;
+L130:
+    if (sw0001 == 124) {
+	goto L140;
+    }
+    goto L150;
+L131:
+/* L121: */
+    ++ip;
+    goto L120;
+L122:
+    memc[op - 1] = 0;
+    matchd = FALSE_;
+    if (! (*formad)) {
+	goto L160;
+    }
+L170:
+    if (! (strmah_(lbuf, st0001) > 0)) {
+	goto L180;
+    }
+    if (! (getlie_(fd, lbuf) != -2) && ! xercom_1.xerflg) {
+	goto L190;
+    }
+    if (xercom_1.xerflg) {
+	goto L100;
+    }
+    matchd = lhmath_(lbuf, patoff, &npat);
+    if (! matchd) {
+	goto L200;
+    }
+    goto L172;
+L200:
+L190:
+L180:
+/* L171: */
+    if (! (getlie_(fd, lbuf) == -2) && ! xercom_1.xerflg) {
+	goto L170;
+    }
+    if (xercom_1.xerflg) {
+	goto L100;
+    }
+L172:
+    ungete_(fd, lbuf);
+    ungete_(fd, st0002);
+    goto L161;
+L160:
+L210:
+    matchd = lhmath_(lbuf, patoff, &npat);
+    if (! matchd) {
+	goto L220;
+    }
+    goto L212;
+L220:
+/* L211: */
+    if (! (getlie_(fd, lbuf) == -2) && ! xercom_1.xerflg) {
+	goto L210;
+    }
+    if (xercom_1.xerflg) {
+	goto L100;
+    }
+L212:
+    ungete_(fd, lbuf);
+L161:
+L91:
+    sfree_(&sp);
+    if (! matchd) {
+	goto L230;
+    }
+    ret_val = 0;
+    goto L100;
+L230:
+    ret_val = -2;
+    goto L100;
+/* L231: */
+L100:
+    zzepro_();
+    return ret_val;
+} /* lhfinn_ */
+
+#undef memx
+#undef mems
+#undef memr
+#undef meml
+#undef memi
+#undef memc
+#undef memb
+
+
+logical lhmath_(lbuf, patoff, npat)
+shortint *lbuf;
+integer *patoff, *npat;
+{
+    /* Initialized data */
+
+    static shortint st0001[6] = { 94,123,37,115,125,0 };
+
+    /* System generated locals */
+    logical ret_val;
+
+    /* Local variables */
+    static integer sp, pat;
+#define memb ((logical *)&mem_1)
+#define memc ((shortint *)&mem_1)
+#define memi ((integer *)&mem_1)
+#define meml ((integer *)&mem_1)
+#define memr ((real *)&mem_1)
+#define mems ((shortint *)&mem_1)
+#define memx ((complex *)&mem_1)
+    extern /* Subroutine */ int sfree_(), smark_(), salloc_();
+    static integer patten;
+    extern integer strmah_();
+    extern /* Subroutine */ int pargsr_(), sprinf_(), zzepro_();
+
+    /* Parameter adjustments */
+    --lbuf;
+    --patoff;
+
+    /* Function Body */
+    smark_(&sp);
+    salloc_(&patten, &c__255, &c__2);
+    pat = 1;
+L110:
+    if (! (pat <= *npat)) {
+	goto L112;
+    }
+    sprinf_(&memc[patten - 1], &c__255, st0001);
+    pargsr_(&memc[patoff[pat] - 1]);
+    if (! (strmah_(&lbuf[1], &memc[patten - 1]) > 0)) {
+	goto L120;
+    }
+    sfree_(&sp);
+    ret_val = TRUE_;
+    goto L100;
+L120:
+/* L111: */
+    ++pat;
+    goto L110;
+L112:
+    sfree_(&sp);
+    ret_val = FALSE_;
+    goto L100;
+L100:
+    zzepro_();
+    return ret_val;
+} /* lhmath_ */
+
+#undef memx
+#undef mems
+#undef memr
+#undef meml
+#undef memi
+#undef memc
+#undef memb
+
+
+/* Subroutine */ int lhmkne_(instr, outstr)
+shortint *instr, *outstr;
+{
+    static integer i__;
+    extern /* Subroutine */ int zzepro_(), xstrcy_(), strlwr_();
+
+    /* Parameter adjustments */
+    --outstr;
+    --instr;
+
+    /* Function Body */
+    xstrcy_(&instr[1], &outstr[1], &c__1023);
+    strlwr_(&outstr[1]);
+    i__ = 1;
+L110:
+    if (! (i__ < 1023)) {
+	goto L112;
+    }
+    if (! (outstr[i__] == 0 || outstr[i__] == 10)) {
+	goto L120;
+    }
+    goto L112;
+L120:
+    if (outstr[i__] >= 65 && outstr[i__] <= 90 || outstr[i__] >= 97 && outstr[
+	    i__] <= 122 || outstr[i__] >= 48 && outstr[i__] <= 57) {
+	goto L130;
+    }
+    outstr[i__] = 95;
+L130:
+/* L121: */
+/* L111: */
+    ++i__;
+    goto L110;
+L112:
+/* L100: */
+    zzepro_();
+    return 0;
+} /* lhmkne_ */
+
diff --git a/pkg/system/help/lroff/lroff2html.x b/pkg/system/help/lroff/lroff2html.x
new file mode 100644
index 00000000..1e3815ae
--- /dev/null
+++ b/pkg/system/help/lroff/lroff2html.x
@@ -0,0 +1,781 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	"lroff.h"
+
+define  DIRECTIVE       1			# processing codes
+define  NAME            2
+define  TEXT            3
+
+define	F_ROMAN		1			# font changes
+define	F_ITALIC	2
+define	F_BOLD		3
+define	F_PREVIOUS	4
+define	F_TELETYPE	5			# HTML-specific font
+
+define	SPTR		Memi[$1+$2]
+define	SECTION		Memc[SPTR($1,$2)]
+define	MAX_SECTIONS	256
+
+
+# LROFF2HTML -- Convert LROFF text to HTML.  By default we process the
+# entire file however we allow for the printing of only a particular section
+# or labelled text block to be compatible with the HELP task options.
+# If a section name is given that section will be printed and and .ls
+# block request will be ignored.
+
+procedure lroff2html (in, out, module, parstr, center, ls_block, section)
+
+int	in					#I input file descriptor
+int	out					#I output file descriptor
+char	module[ARB]				#I .help module name
+char	parstr[ARB]				#I .help optional keyword 2
+char	center[ARB]				#I .help optional keyword 3
+char	ls_block[ARB]				#I .ls block to search for
+char	section[ARB]				#I section to print
+
+pointer sp, ip, sptr
+pointer ibuf, unesc, name, level
+int	lastline, font, indented, ls_level
+int	i, arg, nsec, cmd
+bool	format, quit_at_le, quit_at_ih, formatted
+
+int	lh_findsection(), lh_findblock(), nextcmd()
+int	stridxs(), getline(), strlen(), strmatch(), lgetarg()
+
+define	text_ 	99
+define	err_ 	98
+
+include	"lroff.com"
+
+begin
+	call smark (sp)
+	call salloc (ibuf, SZ_IBUF, TY_CHAR)
+	call salloc (unesc, SZ_IBUF, TY_CHAR)
+	call salloc (name, SZ_LINE, TY_CHAR)
+	call salloc (level, SZ_FNAME, TY_CHAR)
+	call salloc (sptr, MAX_SECTIONS, TY_POINTER)
+
+	call aclrc (Memc[ibuf], SZ_IBUF)
+	call aclrc (Memc[name], SZ_LINE)
+	call aclrc (Memc[unesc], SZ_IBUF)
+	call aclrc (Memc[level], SZ_FNAME)
+
+	# Initialize.
+	lastline  = TEXT
+	font 	  = F_ROMAN
+	indented  = YES
+	nsec 	  = 0
+	ls_level  = 0
+        format    = true
+	quit_at_le = false
+	quit_at_ih = false
+        formatted  = false
+
+	# Initialize the section numbering.
+	call amovki (0, nh_level, MAX_NHLEVEL)
+
+        # Determine whether or not the text is formatted.
+        repeat {
+            if (getline (in, Memc[ibuf]) == EOF)
+                goto err_
+            for (ip=1;  IS_WHITE(Memc[ibuf+ip-1]);  ip=ip+1)
+               ;
+        } until (Memc[ibuf+ip-1] != '\n')
+        call ungetline (in, Memc[ibuf])
+        if (Memc[ibuf] == '.')
+           formatted = true
+
+	# Scan forward if searching for and item.
+	if (section[1] != EOS) {
+	    if (lh_findsection (in, formatted, section) == EOF)
+		goto err_
+	} else if (ls_block[1] != EOS) {
+	    if (lh_findblock (in, formatted, ls_block) == EOF)
+		goto err_
+	    quit_at_le = true
+	}
+
+	# Begin the output.
+	call lh_prolog (out, module, parstr, center)
+	call fprintf (out, "<TITLE>%s</TITLE>\n<UL>\n")
+	    if (ls_block[1] != EOS)
+		call pargstr (ls_block)
+	    else if (section[1] != EOS)
+		call pargstr (section)
+	    else if (module[1] != EOS)
+		call pargstr (module)
+	    else
+		call pargstr (" ")
+
+
+	# Process the file.
+	while (getline (in, Memc[ibuf]) != EOF) {
+
+	    # Make a copy of the raw line minus the newline char, we may need
+	    # this to extract comments later.
+	    Memc[ibuf+strlen(Memc[ibuf])-1] = EOS
+	    call strcpy (Memc[ibuf], Memc[unesc], SZ_LINE)
+
+
+	    # Escape problem chars for HTML and handle font changes. Changes
+	    # are done in-place.
+	    call lh_escape (Memc[ibuf], font, format, NO, SZ_LINE)
+
+	    switch (Memc[ibuf]) {
+	    case '\n':
+		call fprintf (out, "<P>\n")
+
+	    case '.':
+		# Swallow any help strings if present.
+		if (strmatch (Memc[ibuf], "^.help") > 0)
+		    next
+
+		ip = 1
+		lastline = TEXT
+
+		# Process the directive, position the ip at the beginning
+		# of any argument.
+		cmd = nextcmd (Memc[ibuf], ip)
+                while (IS_WHITE(Memc[ibuf+ip]))		# skip spaces
+                    ip = ip + 1
+
+		switch (cmd) {
+		case FI:		# enter fill mode
+		    call fprintf (out, "</PRE>\n")
+		    format = true
+		case NF:		# leave fill mode (nofill)
+		    call fprintf (out, "<PRE>\n")
+		    format = false
+		case JU:		# enter line justification mode
+		    # no-op
+		    next
+		case NJ:		# leave line justification mode
+		    # no-op
+		    next
+		case RJ:		# right justify text on nf,nj line
+		    # no-op
+		    next
+
+		case SH:		# section heading
+		    if (ls_level > 0) {	# for missing .le statements
+		        call fprintf (out, "</DD>\n</DL>\n")
+		        ls_level = 0
+		    }
+		    lastline = DIRECTIVE
+       	    	    Memc[level] = EOS
+		    next
+		case IH:		# indented section heading
+		    if (ls_level > 0) {	# for missing .le statements
+		        call fprintf (out, "</DD>\n</DL>\n")
+		        ls_level = 0
+		    }
+		    lastline = DIRECTIVE
+       	    	    Memc[level] = EOS
+		    if (quit_at_ih)
+			if (stridxs ("|", section) > 0) {
+			    quit_at_ih = false
+			    call ungetline (in, Memc[ibuf])
+			    if (lh_findsection (in, formatted, section) == EOF)
+		    	        break
+			} else
+		    	    break
+		    next
+		case NH:		# numbered section heading
+		    if (ls_level > 0) {	# for missing .le statements
+		        call fprintf (out, "</DD>\n</DL>\n")
+		        ls_level = 0
+		    }
+		    call lh_set_level (lgetarg(Memc[ibuf],ip,1), Memc[level])
+		    lastline = DIRECTIVE
+		    next
+
+		case CE:		# center next line
+		    if (getline (in, Memc[ibuf]) == EOF)
+			break
+		    else {
+	    	        call lh_escape (Memc[ibuf], font, true, NO, SZ_LINE)
+		        call fprintf (out, "<CENTER>%s</CENTER><BR>\n")
+		            call pargstr (Memc[ibuf])
+		    }
+
+		case BR:		# break line
+		    call fprintf (out, "<BR>\n")
+		case SP:		# break, space N spaces on output
+		    arg = lgetarg (Memc[ibuf], ip, 1)
+		    call fprintf (out, "<BR>\n")
+		    for (i=1; i < arg; i = i + 1)
+		        call fprintf (out, "<BR>\n")
+		case IN:		# indent +/- N spaces
+		    # no-op
+		    next
+
+		case LS:		# begin labelled section
+		    arg = lgetarg (Memc[ibuf], ip, 0)
+		    if (arg == 0) 
+			ip = 5
+
+		    # Generate a HREF of the label, we use only the first word.
+		    call strcpy (Memc[ibuf+ip-1], Memc[name], SZ_LINE)
+		    for (i=0; IS_ALNUM(Memc[name+i]) || 
+				Memc[name+i] == '_'; i=i+1)
+			    	    ;
+		    Memc[name+i] = EOS
+	    	    Memc[ibuf+ip+strlen(Memc[ibuf+ip])-1] = EOS
+
+		    call fprintf (out, "<DL>\n<DT><B>")
+		    call fprintf (out,"<A NAME=\"l_%s\">%s</A>")
+		        call pargstr (Memc[name])
+		        call pargstr (Memc[ibuf+ip-1])
+		    call fprintf (out, "</B></DT>\n")
+
+		    # Write out a comment line for the GUI to use.
+	    	    call lh_escape(Memc[unesc+ip-1], font, true, YES, SZ_LINE)
+	    	    Memc[unesc+strlen(Memc[unesc])-1] = EOS
+		    call fprintf (out, 
+			"<! Sec=%s Level=%d Label=\'%s\' Line=\'%s\'>\n<DD>")
+			    if (nsec > 0)
+		                call pargstr (SECTION(sptr, nsec-1))
+			    else
+		                call pargstr ("None")
+		            call pargi (ls_level)
+		            call pargstr (Memc[name])
+			    if (Memc[unesc+ip-1] == '\n')
+		                call pargstr (" ")
+			    else
+		                call pargstr (Memc[unesc+ip-1])
+		    ls_level = ls_level + 1
+
+		case LE:		# end labelled section
+		    call fprintf (out, "</DD>\n</DL>\n")
+		    ls_level = ls_level - 1
+		    if (quit_at_le)
+			break
+
+                case HR:		# HREF anchor
+                    # HTML href anchor of the form ".hr <href> <anch_text>",
+		    # we skip ahead to the <text> and process as a normal line.
+	    	    Memc[ibuf+ip+strlen(Memc[ibuf+ip])-1] = EOS
+                    for (i=0; !IS_WHITE(Memc[ibuf+ip]); ip=ip+1) {
+			Memc[name+i] = Memc[ibuf+ip]
+                        i = i + 1
+		    }
+		    Memc[name+i] = EOS
+
+		    call fprintf (out, "<A HREF=\"%s\">%s</A>\n")
+		        call pargstr (Memc[name])
+		        call pargstr (Memc[ibuf+ip+1])
+
+                case HN:		# NAME target
+                    # HTML name target of the form ".hn <name>", strip the
+		    # newline added in the escape routine.
+	    	    Memc[ibuf+ip+strlen(Memc[ibuf+ip])-1] = EOS
+		    call fprintf (out, "<A NAME=\"%s\"></A>\n")
+			call pargstr (Memc[ibuf+ip])
+
+		case BP:		# break page
+		    # no-op
+		    next
+		case TP:		# test space left on page
+		    # no-op
+		    next
+		case KS:		# start floating keep
+		    call fprintf (out, "<PRE>\n")
+		    format = false
+		case KE:		# end floating keep
+		    call fprintf (out, "</PRE>\n")
+		    format = true
+		case ENDHELP:		# end of help block
+		    break
+		}
+
+	    default:
+		if (lastline == DIRECTIVE) {
+
+		    # Section directive name.  For certain standard sections
+		    # we'll force an indention to make the output look better,
+		    # everything else gets written normally.
+
+		    # Save the section name.
+		    call salloc (SPTR(sptr,nsec), SZ_LINE, TY_CHAR)
+	    	    call aclrc (SECTION(sptr,nsec), SZ_LINE)
+	    	    Memc[ibuf+strlen(Memc[ibuf])-1] = EOS
+		    call sprintf (SECTION(sptr,nsec), SZ_LINE, "\'%s\'")
+			call pargstr (Memc[ibuf])
+
+		    if (indented == YES)
+		        call fprintf (out, "</UL>\n")
+		    if (nsec > 0) {
+		        call fprintf (out, "<! EndSection:   %s>\n")
+		            call pargstr (SECTION(sptr,nsec-1))
+		    }
+
+		    # Make the section name a URL target.
+		    call lh_mkname (Memc[ibuf], Memc[name])
+		    if (Memc[level] == EOS) {
+		        call fprintf (out, "<H2><A NAME=\"s_%s\">%s</A></H2>\n")
+			    call pargstr (Memc[name])
+			    call pargstr (Memc[ibuf])
+		    } else {
+		        call fprintf (out,
+			    "<H2><A NAME=\"s_%s\">%s %s</A></H2>\n")
+			    	call pargstr (Memc[name])
+			    	call pargstr (Memc[level])
+			    	call pargstr (Memc[ibuf])
+       	    	    	Memc[level] = EOS
+		    }
+
+		    call fprintf (out, "<! BeginSection: \'%s\'>\n")
+		        call pargstr (Memc[ibuf])
+		    if (indented == YES)
+		        call fprintf (out, "<UL>\n")
+
+		    lastline = NAME
+		    nsec = nsec + 1
+		    if (section[1] != EOS)
+	    	 	quit_at_ih = true
+
+		} else {
+		    # Ordinary text line.
+text_	    	    call fprintf (out, "%s")
+	   	        call pargstr (Memc[ibuf])
+		    lastline = TEXT
+		}
+	    }
+
+	    call aclrc (Memc[ibuf], SZ_IBUF)
+	    call aclrc (Memc[unesc], SZ_IBUF)
+	    call aclrc (Memc[name], SZ_LINE)
+	}
+
+	# Close the last section.
+	if (nsec > 0) {
+	    call fprintf (out, "</UL>\n<! EndSection:    %s>\n\n")
+                call pargstr (SECTION(sptr,nsec-1))
+	}
+
+	# Write out an HTML comment giving the document section names.
+	call fprintf (out, "<! Contents: ")
+	for (i=0; i < nsec; i=i+1) {
+	    call fprintf (out, "%s ")
+		call pargstr (SECTION(sptr,i))
+	}
+	call fprintf (out, " >\n\n")
+	call fprintf (out, "</BODY>\n</HTML>\n")
+
+	call flush (out)
+err_	call sfree (sp)
+end
+
+
+# LH_PROLOG --  Begin the HTML output, print the header table for a help
+# page if we have the information.
+
+procedure lh_prolog (fd, mod, date, title)
+
+int	fd					#I output file descriptor
+char	mod[ARB]				#I .help module name
+char	date[ARB]				#I .help keyword 2
+char	title[ARB]				#I .help keyword 3
+
+begin
+        call fprintf (fd, "<HTML>\n<BODY>\n")
+
+	# If we only have the module name don't bother with header.
+	if (date[1] == EOS && title[1] == EOS)
+	    return
+
+	# Begin the HTML output prolog.
+        call fprintf (fd, "<TABLE WIDTH=\"100%%\" BORDER=0><TR>\n")
+
+	# Left side page header.
+        call fprintf (fd, "<TD ALIGN=LEFT><FONT SIZE=4>\n")
+	if (date[1] == EOS) {
+            call fprintf (fd, "<B>%s</B>")
+                call pargstr (mod)
+	} else {
+            call fprintf (fd, "<B>%s (%s)</B>")
+                call pargstr (mod)
+                call pargstr (date)
+	}
+        call fprintf (fd, "</FONT></TD>\n")
+
+	# Center page header.
+	if (title[1] != EOS) {
+            call fprintf (fd, "<TD ALIGN=CENTER><FONT SIZE=4>\n")
+            call fprintf (fd, "<B>%s</B>\n")
+                call pargstr (title)
+            call fprintf (fd, "</FONT></TD>\n")
+	}
+
+	# Right side page header.
+        call fprintf (fd, "<TD ALIGN=RIGHT><FONT SIZE=4>\n")
+	if (date[1] == EOS) {
+            call fprintf (fd, "<B>%s</B>")
+                call pargstr (mod)
+	} else {
+            call fprintf (fd, "<B>%s (%s)</B>")
+                call pargstr (mod)
+                call pargstr (date)
+	}
+        call fprintf (fd, "</FONT></TD>\n")
+
+        call fprintf (fd, "</TR></TABLE><P>\n")
+end
+
+
+# LH_ESCAPE -- Escape any HTML problem characters in the line ('<','>','&')
+# as well as the font changes.
+
+procedure lh_escape (str, font, format, special_only, maxch)
+
+char	str[ARB]				#I string to edit
+int	font					#U current font
+bool	format					#I formatting flag
+int	special_only				#I escape only special chars?
+int	maxch					#I max length of string
+
+pointer	sp, ip, buf, keyword
+int	i, gstrcpy(), stridx()
+
+define	copy_	90
+
+begin
+	call smark (sp)
+	call salloc (buf, maxch, TY_CHAR)
+	call salloc (keyword, maxch, TY_CHAR)
+	call aclrc (Memc[buf], maxch)
+	call aclrc (Memc[keyword], maxch)
+
+	ip = buf
+	for (i=1; str[i] != EOS && i <= maxch; i = i + 1) {
+
+	    if (special_only == YES && stridx (str[i], "<>&") == 0)
+		goto copy_
+
+	    switch (str[i]) {
+
+	    # Handle special chars.
+	    case '<':
+		ip = ip + gstrcpy ("&lt;", Memc[ip], SZ_LINE)
+	    case '>':
+		ip = ip + gstrcpy ("&gt;", Memc[ip], SZ_LINE)
+	    case '&':
+		ip = ip + gstrcpy ("&amp;", Memc[ip], SZ_LINE)
+
+	    # Quoted single chars and strings get a special font.
+	    case '\'':
+		if (str[i+2] == '\'') {
+		    ip = ip + gstrcpy ("<TT>",  Memc[ip], SZ_LINE)
+		    ip = ip + gstrcpy (str[i],  Memc[ip], 3)
+		    ip = ip + gstrcpy ("</TT>", Memc[ip], SZ_LINE)
+		    i = i + 2
+		} else
+		    goto copy_
+	    case '`':
+		if (str[i+2] == '`' || str[i+2] == '\'') {
+		    ip = ip + gstrcpy ("<TT>",  Memc[ip], SZ_LINE)
+		    ip = ip + gstrcpy (str[i],  Memc[ip], 3)
+		    ip = ip + gstrcpy ("</TT>", Memc[ip], SZ_LINE)
+		    i = i + 2
+		} else
+		    goto copy_
+	    case '"':
+                if (format && str[i+1] != '/' && str[i+2] != '/') {
+		    if (font == F_TELETYPE) {
+		        # Do a closing quote.
+		        ip = ip + gstrcpy ("</TT>\"", Memc[ip], SZ_LINE)
+		        font = F_ROMAN
+		    } else if (font == F_ROMAN) {
+		        # Do an opening quote.
+		        ip = ip + gstrcpy ("\"<TT>", Memc[ip], SZ_LINE)
+		        font = F_TELETYPE
+		    } else
+		        goto copy_
+		} else 
+		    goto copy_
+
+	    # Process font changes.
+	    case '\\':
+		if (str[i+1] == 'f') {
+		    if (str[i+2] == 'B') {
+			if (font == F_BOLD)
+			    next
+			if (font == F_ITALIC)
+			    ip = ip + gstrcpy ("</I>", Memc[ip], SZ_LINE)
+			ip = ip + gstrcpy ("<B>", Memc[ip], SZ_LINE)
+			font = F_BOLD
+
+		    } else if (str[i+2] == 'I') {
+			if (font == F_ITALIC)
+			    next
+			if (font == F_BOLD)
+			    ip = ip + gstrcpy ("</B>", Memc[ip], SZ_LINE)
+			ip = ip + gstrcpy ("<I>", Memc[ip], SZ_LINE)
+			font = F_ITALIC
+
+		    } else if (str[i+2] == 'R') {
+			if (font == F_BOLD)
+			    ip = ip + gstrcpy ("</B>", Memc[ip], SZ_LINE)
+			else if (font == F_ITALIC)
+			    ip = ip + gstrcpy ("</I>", Memc[ip], SZ_LINE)
+			font = F_ROMAN
+
+		    } else if (str[i+2] == 'P') {
+			if (font == F_BOLD) {
+			    ip = ip + gstrcpy ("</B>", Memc[ip], SZ_LINE)
+			} else if (font == F_ITALIC) {
+			    ip = ip + gstrcpy ("</I>", Memc[ip], SZ_LINE)
+			}
+			font = F_ROMAN
+		    }
+		    i = i + 2
+		} else if (str[i+1] == '\n' || str[i+1] == EOS) {
+		    Memc[ip] = str[i]
+		    ip = ip + 1
+		    i = i + 1
+		    ip = ip + gstrcpy ("<BR>", Memc[ip], SZ_LINE)
+		} else
+		    goto copy_
+
+	    default:
+copy_		Memc[ip] = str[i]
+		ip = ip + 1
+	    }
+	}
+
+	# Add the trailing newline we stripped above.
+	ip = ip + gstrcpy ("\n\0", Memc[ip], SZ_LINE)
+
+	# Move the string back.
+	call amovc (Memc[buf], str, maxch)
+
+	call sfree (sp)
+end
+
+
+# LH_SET_LEVEL -- Increment the level number of a numbered header.
+
+procedure lh_set_level (n, level)
+
+int	n					#I level number
+char	level[ARB]				#U level string
+
+int	i, strlen()
+include	"lroff.com"
+
+begin
+        # Increment the desired section number; zero all higher
+	# numbered section counters.
+        nh_level[n] = nh_level[n] + 1
+        call amovki (0, nh_level[n+1], MAX_NHLEVEL - n)
+
+        # Output the section number followed by a blank and then
+	# the section label.
+	level[1] = EOS
+        do i = 1, n {
+	    call sprintf (level[strlen(level)+1], SZ_IBUF, "%d.")
+	        call pargi (nh_level[i])
+	}
+
+        # Cancel the final "." if subsection heading.  Add a blank.
+        if (n > 1 && level[strlen(level)] == '.')
+            level[strlen(level)] = EOS
+end
+
+
+# LH_FINDBLOCK -- If text contains format directives, eat input lines until
+# a ".ls" directive is found which contains the block name as a substring.
+# If the text is not formatted, search for a line beginning with the pattern.
+
+int procedure lh_findblock (fd, formatted, param)
+
+int	fd
+bool	formatted
+char	param[ARB]
+
+bool	match_found
+pointer	sp, lbuf, pattern
+int	len
+int	getline(), strmatch(), strlen()
+errchk	getline
+
+define	err_	90
+
+begin
+	call smark (sp)
+	call salloc (pattern, SZ_FNAME, TY_CHAR)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+
+	match_found = false
+
+	# Get the first line.
+	if (getline (fd, Memc[lbuf]) == EOF)
+	    goto err_
+
+	if (formatted) {
+	    call sprintf (Memc[pattern], SZ_FNAME, "{%s}")
+		call pargstr (param)
+	    repeat {
+		if (strmatch (Memc[lbuf], "^.{ls}") > 0)
+		    if (strmatch (Memc[lbuf], Memc[pattern]) > 0) {
+			match_found = true
+			break
+		    }
+	    } until (getline (fd, Memc[lbuf]) == EOF)
+
+	} else {
+	    call sprintf (Memc[pattern], SZ_FNAME, "^#{%s}")
+		call pargstr (param)
+	    repeat {
+		if (strmatch (Memc[lbuf], Memc[pattern]) > 0) {
+		    match_found = true
+		    break
+		}
+	    } until (getline (fd, Memc[lbuf]) == EOF)
+	}
+        call ungetline (fd, Memc[lbuf])
+
+err_	len = strlen (Memc[lbuf])
+	call sfree (sp)
+	if (match_found)
+	    return (len)
+	else
+	    return (EOF)
+end
+
+
+# LH_FINDSECTION -- If text contains format directives, eat input lines until
+# a ".ih" directive is found for the named section.  If the text is not
+# formatted, search for a line beginning with the section name.
+
+define	MAXPAT		10
+
+int procedure lh_findsection (fd, formatted, sections)
+
+int	fd			# input file
+bool	formatted		# is help block formatted
+char	sections[ARB]		# list of sections "a|b|c"
+
+bool	match_found
+int	npat, ip
+pointer	sp, patbuf, patoff[MAXPAT], op
+char	lbuf[SZ_LINE]
+
+bool	lh_match()
+int	getline(), strmatch()
+errchk	getline
+
+define	err_	91
+
+begin
+	call smark (sp)
+	call salloc (patbuf, SZ_LINE, TY_CHAR)
+
+	# Process the list of sections into patbuf and patoff, i.e., into a
+	# list of EOS delimited strings in the string buffer patbuf.  Each
+	# section name or abbreviation is delimited by '|' (or).
+
+	npat = 1
+	op = patbuf
+	patoff[1] = op
+
+	# Get the first line.
+	if (getline (fd, lbuf) == EOF)
+	    goto err_
+
+	for (ip=1;  sections[ip] != EOS;  ip=ip+1)
+	    switch (sections[ip]) {
+	    case '|':
+		Memc[op] = EOS
+		op = op + 1
+		npat = min (MAXPAT, npat + 1)
+		patoff[npat] = op
+	    default:
+		Memc[op] = sections[ip]
+		op = op + 1
+	    }
+	Memc[op] = EOS
+
+	match_found = false
+
+	if (formatted) {
+	    repeat {
+		if (strmatch (lbuf, "^.{ih}") > 0)
+		    if (getline (fd, lbuf) != EOF) {
+			match_found = lh_match (lbuf, patoff, npat)
+			if (match_found)
+			    break
+		    }
+	    } until (getline (fd, lbuf) == EOF)
+            call ungetline (fd, lbuf)
+	    call ungetline (fd, ".ih\n")
+
+	} else {
+	    repeat {
+		match_found = lh_match (lbuf, patoff, npat)
+		if (match_found)
+		    break
+	    } until (getline (fd, lbuf) == EOF)
+            call ungetline (fd, lbuf)
+	}
+
+err_	call sfree (sp)
+        if (match_found)
+            return (OK)
+        else
+            return (EOF)
+end
+
+
+# LH_MATCH -- Match a set of patterns against a line of test, matching only
+# at the beginning of line in either case.
+
+bool procedure lh_match (lbuf, patoff, npat)
+
+char    lbuf[ARB]               # line of text
+pointer patoff[npat]            # pointers to pattern strings
+int     npat                    # number of patterns
+
+int     pat
+pointer sp, pattern
+int     strmatch()
+
+begin
+        call smark (sp)
+        call salloc (pattern, SZ_FNAME, TY_CHAR)
+
+        for (pat=1;  pat <= npat;  pat=pat+1) {
+            call sprintf (Memc[pattern], SZ_FNAME, "^{%s}")
+                call pargstr (Memc[patoff[pat]])
+            if (strmatch (lbuf, Memc[pattern]) > 0) {
+        	call sfree (sp)
+                return (true)
+	    }
+        }
+
+        call sfree (sp)
+        return (false)
+end
+
+
+# LH_MKNAME -- Given a string make it suitable for use as an HREF name.
+
+procedure lh_mkname (instr, outstr)
+
+char	instr[ARB]
+char	outstr[ARB]
+
+int	i
+
+begin
+	# Make it a URL.  First convert the section name to a
+	# lower-case string and replace the blanks.
+	call strcpy (instr, outstr, SZ_LINE)
+	call strlwr (outstr)
+	for (i=1; i < SZ_LINE; i=i+1)
+	    if (outstr[i] == EOS || outstr[i] == '\n')
+		break
+	    else if (!IS_ALNUM(outstr[i]))
+	        outstr[i] = '_'
+end
diff --git a/pkg/system/help/lroff/lroff2ps.x b/pkg/system/help/lroff/lroff2ps.x
new file mode 100644
index 00000000..a782ca21
--- /dev/null
+++ b/pkg/system/help/lroff/lroff2ps.x
@@ -0,0 +1,460 @@
+include	<syserr.h>
+include	<ctype.h>
+include	<psset.h>
+include	"lroff.h"
+
+
+define  DIRECTIVE       1			# processing codes
+define  NAME            2
+define  TEXT            3
+define  NEWLINE         4
+
+define	INDENT		5			# size of indentitudedness
+
+
+# LROFF2PS -- Convert LROFF text to Postscript.  By default we process the
+# entire file however we allow for the printing of only a particular section
+# or labelled text block to be compatible with the HELP task options.
+# If a section name is given that section will be printed and and .ls
+# block request will be ignored.
+
+procedure lroff2ps (in, out, psptr, ls_block, section)
+
+int	in					#i input file descriptor
+int	out					#i output file descriptor
+pointer	psptr					#i PSIO pointer
+char	ls_block[ARB]				#i .ls block to search for
+char	section[ARB]				#i section to print
+
+pointer sp, ip, ps
+pointer ibuf, line, level
+int	lastline, font, indent, ls_level
+int	i, arg, nsec, cmd, last_cmd
+bool	format, quit_at_le, quit_at_ih, formatted
+
+pointer	ps_open()
+int	lh_findsection(), lh_findblock(), nextcmd()
+int	stridxs(), getline(), strlen(), strmatch(), lgetarg()
+
+define	text_ 	99
+define	err_ 	98
+
+include	"lroff.com"
+
+begin
+	call smark (sp)
+	call salloc (ibuf, SZ_IBUF, TY_CHAR)
+	call salloc (line, SZ_LINE, TY_CHAR)
+	call salloc (level, SZ_FNAME, TY_CHAR)
+
+	call aclrc (Memc[ibuf], SZ_LINE)
+	call aclrc (Memc[line], SZ_LINE)
+	call aclrc (Memc[level], SZ_LINE)
+
+	# Initialize.
+	lastline  = TEXT
+	font 	  = F_ROMAN
+	indent    = 0
+	nsec 	  = 0
+	ls_level  = 0
+        format    = true
+	quit_at_le = false
+	quit_at_ih = false
+        formatted  = false
+
+	# Initialize the section numbering.
+	call amovki (0, nh_level, MAX_NHLEVEL)
+
+        # Determine whether or not the text is formatted.
+        repeat {
+            if (getline (in, Memc[ibuf]) == EOF)
+                goto err_
+            for (ip=1;  IS_WHITE(Memc[ibuf+ip-1]);  ip=ip+1)
+               ;
+        } until (Memc[ibuf+ip-1] != '\n')
+        call ungetline (in, Memc[ibuf])
+        if (Memc[ibuf] == '.')
+           formatted = true
+
+	# Scan forward if searching for and item.
+	if (section[1] != EOS) {
+	    if (lh_findsection (in, formatted, section) == EOF)
+		goto err_
+	} else if (ls_block[1] != EOS) {
+	    if (lh_findblock (in, formatted, ls_block) == EOF)
+		goto err_
+	    quit_at_le = true
+	}
+
+	# Begin the output is we aren't passed a pointer that may have
+	# already set up headers, page parameters, etc.
+	if (psptr == NULL)
+	    ps = ps_open (out, YES)
+	else
+	    ps = psptr
+	call ps_write_prolog (ps)
+
+	# Process the file.
+	while (getline (in, Memc[ibuf]) != EOF) {
+
+	    # Escape the line, inserting special font changes for quotes
+	    # chars and strings.
+	    call lp_escape (Memc[ibuf], font, format, SZ_LINE)
+
+	    switch (Memc[ibuf]) {
+	    case '\n':
+		if (lastline != NEWLINE)
+		    call ps_linebreak (ps, NO)
+		call ps_newline (ps)
+		lastline = NEWLINE
+
+	    case '.':
+		# Swallow any help strings if present.
+		if (strmatch (Memc[ibuf], "^.{help}") > 0)
+		    next
+
+	        # Stomp the newline.
+	        Memc[ibuf+strlen(Memc[ibuf])-1] = EOS
+
+		# Process the directive, position the ip at the beginning
+		# of any argument.
+		ip = 1
+		cmd = nextcmd (Memc[ibuf], ip)
+                while (IS_WHITE(Memc[ibuf+ip]))		# skip spaces
+                    ip = ip + 1
+
+		switch (cmd) {
+		case FI, KE:		# enter fill mode
+		    call ps_setfont (ps, F_ROMAN)
+		    call ps_set_justify (ps, YES)
+		    lastline = NEWLINE
+		    format = true
+		case NF, KS:		# leave fill mode (nofill)
+		    if (lastline != NEWLINE && last_cmd != KS && last_cmd != NF)
+		        call ps_linebreak (ps, NO)
+		    call ps_setfont (ps, F_TELETYPE)
+		    call ps_set_justify (ps, NO)
+		    lastline = NEWLINE
+		    format = false
+		case JU:		# enter line justification mode
+		    call ps_set_justify (ps, YES)
+		    next
+		case NJ:		# leave line justification mode
+		    call ps_set_justify (ps, NO)
+		    next
+		case RJ:		# right justify text on next line
+		    call ps_linebreak (ps, NO)
+		    call strcpy (Memc[ibuf+4], Memc[line], SZ_FNAME)
+	    	    Memc[line+strlen(Memc[line])-1] = EOS
+		    if (getline (in, Memc[ibuf]) != EOF) {
+		        call ps_output (ps, Memc[ibuf], NO)
+		        call ps_rightjustify (ps, Memc[line])
+		    }
+		    next
+
+		case SH:		# section heading
+		    if (ls_level > 0) {	# for missing .le statements
+		        ls_level = 0
+		        indent = 0
+		    }
+		    if (nsec > 0) {
+		        call ps_linebreak (ps, NO)
+		        call ps_linebreak (ps, NO)
+		    }
+		    lastline = DIRECTIVE
+       	    	    Memc[level] = EOS
+		    next
+		case IH:		# indented section heading
+		    if (ls_level > 0) {	# for missing .le statements
+		        ls_level = 0
+		        indent = 0
+		    }
+			
+		    if (nsec > 0)
+		        call ps_linebreak (ps, NO)
+		    if (lastline == TEXT)
+		        call ps_linebreak (ps, NO)
+		    lastline = DIRECTIVE
+		    Memc[level] = EOS
+		    if (quit_at_ih)
+			if (stridxs ("|", section) > 0) {
+			    quit_at_ih = false
+			    call ungetline (in, Memc[ibuf])
+			    if (lh_findsection (in, formatted, section) == EOF)
+		    	        break
+			} else
+		    	    break
+		    next
+		case NH:		# numbered section heading
+		    if (ls_level > 0) {	# for missing .le statements
+		        ls_level = 0
+		        indent = 0
+		    }
+		    call lh_set_level (lgetarg(Memc[ibuf],ip,1), Memc[level])
+		    if (nsec > 0) {
+		        call ps_linebreak (ps, NO)
+		        call ps_linebreak (ps, NO)
+		    }
+		    lastline = DIRECTIVE
+		    next
+
+		case CE:		# center next line
+		    if (getline (in, Memc[ibuf]) != EOF) {
+	    	        call lp_escape (Memc[ibuf], font, true, SZ_LINE)
+			call ps_linebreak (ps, NO)
+			call ps_center (ps, Memc[ibuf])
+		    }
+
+		case BR:		# break line
+		    call ps_linebreak (ps, NO)
+		case SP:		# break, space N spaces on output
+		    arg = lgetarg (Memc[ibuf], ip, 1)
+		    call ps_linebreak (ps, NO)
+		    for (i=1; i < arg; i = i + 1)
+		        call ps_linebreak (ps, NO)
+		case IN:		# indent +/- N spaces
+		    arg = lgetarg (Memc[ibuf], ip, 0)
+		    call ps_indent (ps, arg)
+		    next
+
+		case LS:		# begin labelled section
+		    arg = lgetarg (Memc[ibuf], ip, INDENT)
+	  	    if (arg == 0)
+			ip = 5
+
+		    if (lastline == TEXT)
+		        call ps_linebreak (ps, NO)
+		    call ps_testpage (ps, 2)
+		    if (ls_level < 1 || (last_cmd == LS && lastline == TEXT))
+		        call ps_linebreak (ps, NO)
+		    call ps_spfont (ps, F_BOLD)
+		    call ps_output (ps, Memc[ibuf+ip-1], NO)
+		    if (strlen (Memc[ibuf+ip-1]) > (INDENT-1))
+		        call ps_linebreak (ps, NO)
+		    call ps_spfont (ps, NULL)
+		    call ps_setfont (ps, F_ROMAN)
+
+		    indent = max (0, indent + INDENT)
+		    call ps_indent (ps, indent)
+		    ls_level = ls_level + 1
+
+		case LE:		# end labelled section
+		    if (last_cmd != LE || (last_cmd == LE && lastline == TEXT))
+			call ps_linebreak (ps, NO)
+		    indent = max (0, indent - INDENT)
+		    call ps_indent (ps, indent)
+		    ls_level = ls_level - 1
+		    lastline = NEWLINE
+		    if (quit_at_le)
+			break
+
+                case HR:		# HREF anchor
+                    # HTML href anchor of the form ".hr <href> <anch_text>",
+		    # we skip ahead to the <text> and process as a normal line.
+                    for (i=0; !IS_WHITE(Memc[ibuf+ip]); ip=ip+1)
+                        i = i + 1
+		    call ps_deposit (ps, Memc[ibuf+ip+1])
+
+                case HN:		# NAME target
+                    # HTML name target of the form ".hn <name>".
+		    # no-op
+
+		case BP:		# break page
+		    call ps_linebreak (ps, NO)
+		    call ps_pagebreak (ps)
+		    next
+		case TP:		# test space left on page
+		    # no-op
+		    next
+		case ENDHELP:		# end of help block
+		    break
+		}
+		last_cmd = cmd
+
+	    default:
+	        # Stomp the newline.
+	        Memc[ibuf+strlen(Memc[ibuf])-1] = EOS
+
+		if (lastline == DIRECTIVE) {
+
+		    # Section directive name.  For certain standard sections
+		    # we'll force an indention to make the output look better,
+		    # everything else gets written normally.
+
+		    indent = max (0, indent - INDENT)
+		    call ps_indent (ps, indent)
+
+		    call ps_setfont (ps, F_BOLD)
+		    call ps_testpage (ps, 3)
+		    if (Memc[level] == EOS) {
+			call ps_output (ps, Memc[ibuf], NO)
+		    } else {
+			call sprintf (Memc[line], SZ_LINE, "%s %s")
+			    call pargstr (Memc[level])
+			    call pargstr (Memc[ibuf])
+
+			call ps_output (ps, Memc[line], NO)
+       	    	    	Memc[level] = EOS
+		    }
+		    call ps_setfont (ps, F_ROMAN)
+		    call ps_linebreak (ps, NO)
+
+		    indent = max (0, indent + INDENT)
+		    call ps_indent (ps, indent)
+
+		    lastline = NAME
+		    nsec = nsec + 1
+		    if (section[1] != EOS)
+	    	 	quit_at_ih = true
+
+		} else {
+		    # Ordinary text line.
+text_		    if (format) {
+	    	        call ps_deposit (ps, Memc[ibuf])
+		    } else {
+			call lp_strdetab (Memc[ibuf], Memc[line], SZ_LINE)
+	    	        call ps_output (ps, Memc[line], NO)
+		        call ps_linebreak (ps, NO)
+		    }
+		    lastline = TEXT
+		}
+	    }
+
+	    call aclrc (Memc[ibuf], SZ_LINE)
+	    call aclrc (Memc[line], SZ_LINE)
+	}
+
+	# Close the last section.
+	call ps_linebreak (ps, NO)
+	call ps_close (ps)
+	
+	call flush (out)
+err_	call sfree (sp)
+end
+
+
+# LP_ESCAPE -- Escape any HTML problem characters in the line ('<','>','&')
+# as well as the font changes.
+
+procedure lp_escape (str, font, format, maxch)
+
+char	str[ARB]				#i string to edit
+int	font					#u current font
+bool	format					#i formatting flag
+int	maxch					#i max length of string
+
+pointer	sp, ip, buf, keyword
+int	i, strmatch(), gstrcpy()
+bool	is_ls
+
+define	copy_	90
+
+begin
+	call smark (sp)
+	call salloc (buf, maxch, TY_CHAR)
+	call salloc (keyword, maxch, TY_CHAR)
+	call aclrc (Memc[buf], maxch)
+	call aclrc (Memc[keyword], maxch)
+
+	ip = buf
+        is_ls = FALSE
+        if (strmatch (str, "^.{ls}") > 0)
+            is_ls = TRUE
+
+	for (i=1; str[i] != EOS && i <= maxch; i = i + 1) {
+
+	    switch (str[i]) {
+
+	    # Quoted single chars and strings get a special font.
+	    case '\'', '`':
+		if (str[i+2] == '`' || str[i+2] == '\'') {
+		    if (format)
+		        ip = ip + gstrcpy ("\\fT",  Memc[ip], SZ_LINE)
+		    ip = ip + gstrcpy (str[i],  Memc[ip], 3)
+		    if (format) {
+			if (is_ls)
+		            ip = ip + gstrcpy ("\\fB", Memc[ip], SZ_LINE)
+			else {
+			    switch (font) {
+			    case F_ROMAN:
+		                ip = ip + gstrcpy ("\\fR", Memc[ip], SZ_LINE)
+			    case F_BOLD:
+		                ip = ip + gstrcpy ("\\fB", Memc[ip], SZ_LINE)
+			    case F_ITALIC:
+		                ip = ip + gstrcpy ("\\fI", Memc[ip], SZ_LINE)
+			    case F_TELETYPE:
+		                ip = ip + gstrcpy ("\\fT", Memc[ip], SZ_LINE)
+			    default:
+		                ip = ip + gstrcpy ("\\fR", Memc[ip], SZ_LINE)
+			    }
+			}
+		    }
+		    i = i + 2
+		} else
+		    goto copy_
+	    case '"':
+                if (format && str[i+1] != '/' && str[i+2] != '/') {
+		    if (font == F_TELETYPE) {
+		        # Do a closing quote.
+			if (format) {
+			    if (is_ls)
+		                ip = ip + gstrcpy ("\"\\fB", Memc[ip], SZ_LINE)
+			    else
+		                ip = ip + gstrcpy ("\"\\fR", Memc[ip], SZ_LINE)
+			}
+		        font = F_ROMAN
+		    } else if (font == F_ROMAN) {
+		        # Do an opening quote.
+			if (format)
+		            ip = ip + gstrcpy ("\\fT\"", Memc[ip], SZ_LINE)
+		        font = F_TELETYPE
+		    } else
+		        goto copy_
+		} else 
+		    goto copy_
+
+	    default:
+copy_		Memc[ip] = str[i]
+		ip = ip + 1
+	    }
+	}
+
+	# Add the trailing newline we stripped above.
+	ip = ip + gstrcpy ("\n\0", Memc[ip], SZ_LINE)
+
+	# Move the string back.
+	call amovc (Memc[buf], str, maxch)
+
+	call sfree (sp)
+end
+
+
+# LP_STRDETAB -- Procedure to remove tabs from a line of text.
+
+procedure lp_strdetab (line, outline, maxch)
+
+char   line[ARB], outline[ARB]
+int    maxch
+
+int    ip, op
+
+begin
+        ip = 1
+        op = 1
+
+        while (line[ip] != EOS && op <= maxch) {
+            if (line[ip] == '\t') {
+                repeat {
+                    outline[op] = ' '
+                    op = op + 1
+                } until (mod(op,8) == 0 || op > maxch)
+                ip = ip + 1
+            } else {
+                outline[op] = line [ip]
+                ip = ip + 1
+                op = op + 1
+            }
+        }
+
+        outline[op] = EOS
+end
diff --git a/pkg/system/help/lroff/mkpkg b/pkg/system/help/lroff/mkpkg
new file mode 100644
index 00000000..73a157d4
--- /dev/null
+++ b/pkg/system/help/lroff/mkpkg
@@ -0,0 +1,27 @@
+# Make the LROFF line-oriented text formatter.
+
+$checkout libpkg.a ../../
+$update   libpkg.a
+$checkin  libpkg.a ../../
+$exit
+
+libpkg.a:
+	breakline.x	lroff.com lroff.h words.com <chars.h>
+	center.x	lroff.com lroff.h <chars.h>
+	dols.x		lroff.com lroff.h <chars.h> <error.h>
+	getarg.x	lroff.h <chars.h>
+	indent.x	lroff.com lroff.h
+	input.x		lroff.com lroff.h <chars.h> <ctype.h>
+	justify.x	lroff.com lroff.h <chars.h>
+	lroff.x		lroff.com lroff.h <chars.h> <ctype.h>
+	lroff2html.x	lroff.com lroff.h <ctype.h>
+	lroff2ps.x	lroff.com lroff.h <psset.h> <ctype.h>
+	nextcmd.x	lroff.h <ctype.h>
+	nofill.x	lroff.com lroff.h
+	output.x	lroff.com lroff.h <chars.h> <mach.h>
+	rawcopy.x	lroff.com lroff.h
+	section.x	lroff.com lroff.h <chars.h> <ctype.h>
+	skiplines.x	lroff.h
+	textlen.x	lroff.h <chars.h>
+	textout.x	lroff.com lroff.h words.com <chars.h>
+	;
diff --git a/pkg/system/help/lroff/nextcmd.x b/pkg/system/help/lroff/nextcmd.x
new file mode 100644
index 00000000..d4b09f44
--- /dev/null
+++ b/pkg/system/help/lroff/nextcmd.x
@@ -0,0 +1,56 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	"lroff.h"
+
+.help nextcmd
+.nf ________________________________________________________________________
+NEXTCMD -- Examine the input line; if it is an Lroff directive, return
+the integer code of the directive, otherwise NULL.  Leave IP pointing
+to the argument field if a command, otherwise leave it pointing at the
+first char of the text line.  Note that the "directives" string must
+match the opcode definitions given in lroff.h
+.endhelp ___________________________________________________________________
+
+define	SZ_OPCODE	2
+
+int procedure nextcmd (linebuf, ip)
+
+char	linebuf[ARB]
+int	ip, op
+char	opcode[SZ_OPCODE]
+int	command, kwp, strmatch(), strncmp()
+string	directives "finfjunjrjshihnhbrcespinlslebptpkskehrhn"
+
+begin
+	if (linebuf[1] != '.')			# not a command line?
+	    return (NULL)
+	if (strmatch (linebuf, "^.endhelp") > 0)
+	    return (ENDHELP)
+	ip = 2					# skip the '.'
+
+	# Directives may be either upper or lower case.
+	for (op=1;  op <= SZ_OPCODE;  op=op+1) {
+	    opcode[op] = linebuf[ip]
+	    if (IS_UPPER (opcode[op]))
+		opcode[op] = TO_LOWER (opcode[op])
+	    ip = ip + 1
+	}
+
+	# Just in case a directive happens to be longer than 2 chars, make
+	# sure IP points past the directive name.
+	while (IS_ALPHA (linebuf[ip]))
+	    ip = ip + 1
+
+	# Lookup directive, return opcode number if found.
+	command = NULL
+	for (kwp=1;  directives[kwp] != EOS;  kwp=kwp+SZ_OPCODE)
+	    if (strncmp (opcode, directives[kwp], SZ_OPCODE) == 0) {
+		command = (kwp+1) / SZ_OPCODE
+		break
+	    }
+
+	if (command == NULL)			# unrecognized directive
+	    ip = 1		
+	return (command)
+end
diff --git a/pkg/system/help/lroff/nofill.x b/pkg/system/help/lroff/nofill.x
new file mode 100644
index 00000000..f81b0d88
--- /dev/null
+++ b/pkg/system/help/lroff/nofill.x
@@ -0,0 +1,45 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"lroff.h"
+
+.help nofill
+.nf __________________________________________________________________________
+NOFILL -- Copy a block of text in ".nf" (nofill) mode, leaving the text
+alone except for left justification.  The only directives recognized in
+a nofill block are FI (resume filling) and RJ (right justify).
+.endhelp _____________________________________________________________________
+
+int procedure nofill (in, out, linebuf)
+
+extern	in(), out()
+char	linebuf[ARB]
+int	ip, command
+pointer	sp, rjbuf
+int	in(), input(), nextcmd()
+errchk	salloc, breakline, input, rjline, outline
+include	"lroff.com"
+
+begin
+	call smark (sp)
+	call salloc (rjbuf, SZ_IBUF, TY_CHAR)
+
+	call breakline (out, NJ)
+
+	while (input (in, linebuf) != EOF) {
+	    command = nextcmd (linebuf, ip)
+	    switch (command) {
+	    case FI, ENDHELP:
+		call sfree (sp)
+		return (command)
+	    case RJ:				# right justify text
+		if (input (in, Memc[rjbuf]) == EOF)
+		    break
+		call rjline (out, Memc[rjbuf], linebuf[ip])
+	    default:
+		call outline (out, linebuf)
+	    }
+	}
+
+	call sfree (sp)
+	return (ENDHELP)
+end
diff --git a/pkg/system/help/lroff/output.x b/pkg/system/help/lroff/output.x
new file mode 100644
index 00000000..84ebea0d
--- /dev/null
+++ b/pkg/system/help/lroff/output.x
@@ -0,0 +1,190 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<mach.h>
+include	<chars.h>
+include	"lroff.h"
+
+.help outstr, outc, getoutcol, set_outbuf
+.nf __________________________________________________________________________
+OUTSTR, OUTC, OUTCOL -- Routines for buffering the output line.  An output
+line is built up with calls to OUTPUT to pass strings, and to OUTC to pass
+characters.  OUTCOL may be called to determine the index of the PRINTABLE
+column at which the next character will be deposited (this is unaffected by
+control chars in the output stream).
+
+The output buffer is flushed when OUTC is called to deposit a newline
+character, by calling the user supplied output procedure.  The output column
+pointer is reset to the CURRENT left margin when the first character of the
+new line is deposited.  Any changes to the left margin made after the first
+character is deposited do not take effect until the next line.  OUTCOL returns
+the value of the current left margin if called when the buffer is empty.
+
+NOTE: set_outbuf() must be called during Lroff startup and shutdown to
+allocate the output buffer, the size of which depends on the maximum output
+line length.
+.endhelp _____________________________________________________________________
+
+
+# SET_OUTBUF -- Allocate and initialize the output buffer.
+
+procedure set_outbuf (outbuf_size)
+
+int	outbuf_size		# new buffer size in chars
+
+include	"lroff.com"
+errchk	malloc, out
+bool	first_time
+data	first_time /true/
+
+pointer	obuf, op, otop
+int	col, old_left_margin, buffer_empty
+common	/lroout/ obuf, op, otop, col, old_left_margin, buffer_empty
+
+begin
+	if (first_time) {
+	    obuf = NULL
+	    buffer_empty = YES
+	    first_time = false
+	}
+
+	if (outbuf_size <= 0 && obuf != NULL)
+	    call mfree (obuf, TY_CHAR)
+	else {
+	    call malloc (obuf, outbuf_size, TY_CHAR)
+	    otop = obuf + outbuf_size - 1
+	    op = obuf
+	    buffer_empty = YES
+	    old_left_margin = 1
+	}
+end
+
+
+# OUTC -- Output a single character.  Note that the character value is
+# passed as an integer.
+
+procedure outc (out, ch)
+
+extern	out()			# user line output procedure
+int	ch			# character to be output
+
+char	text[1]
+data	text[1] /EOS/
+
+begin
+	text[1] = ch
+	call outstr (out, text)
+end
+
+
+# OUTSTR -- Output a text string.
+
+procedure outstr (out, text)
+
+extern	out()			# user line output procedure
+char	text[ARB]		# text string to be output
+
+int	ch, ip, i
+pointer	obuf, op, otop
+int	col, old_left_margin, buffer_empty
+common	/lroout/ obuf, op, otop, col, old_left_margin, buffer_empty
+include	"lroff.com"
+
+begin
+	for (ip=1;  text[ip] != EOS;  ) {
+	    if (buffer_empty == YES) {
+		if (obuf == NULL)
+		    call error (1, "No Lroff output buffer allocated")
+
+		# If left margin has been moved inward, blank out the unused
+		# columns.
+
+		if (left_margin != old_left_margin) {
+		    for (i=old_left_margin;  i < left_margin;  i=i+1)
+			Memc[obuf+i-1] = BLANK
+		    old_left_margin = left_margin
+		}
+
+		op = obuf + left_margin - 1
+		col = left_margin
+		buffer_empty = NO
+	    }
+
+	    # Move the text string into the buffer.  The string may contain
+	    # more than one line of text.
+
+	    for (;  text[ip] != EOS;  ip=ip+1) {
+		ch = text[ip]
+		Memc[op] = ch
+		op = op + 1
+
+		if (INVISIBLE(ch) || op > otop) {
+		    if (ch == '\r' || ch == '\n') {
+			# Flush the buffer.
+			Memc[op] = EOS
+			call out (out_magic_arg, Memc[obuf])
+			buffer_empty = YES
+
+			# If all text data has not been copied (buffer overflow
+			# or newline embedded in the text), we must reinit the
+			# buffer and copy the remaining data.  Otherwise we must
+			# return without calling the buffer_empty code to give
+			# the caller a chance to change the left margin.
+
+			if (text[ip+1] == EOS)
+			    return
+		    }
+		} else
+		    col = col + 1
+	    }
+	}
+end
+
+
+# GETOUTCOL -- Return the index of the next column of output.
+
+procedure getoutcol (next_column)
+
+int	next_column		# next col to be written (output)
+pointer	obuf, op, otop
+int	col, old_left_margin, buffer_empty
+common	/lroout/ obuf, op, otop, col, old_left_margin, buffer_empty
+include	"lroff.com"
+
+begin
+	if (buffer_empty == YES)
+	    next_column = left_margin
+	else
+	    next_column = col
+end
+
+
+# OUTCC -- Output a control sequence, i.e., a forms control sequence.
+# Called only after a line has already been output.  Does not interfere
+# with output buffer.  Sequence is not newline terminated.
+
+procedure outcc (out, ctrlchar)
+
+extern	out()			# user supplied line output procedure
+int	ctrlchar		# character to be output (INT)
+char	ctrlstr[1]
+include	"lroff.com"
+
+begin
+	ctrlstr[1] = ctrlchar
+	ctrlstr[2] = EOS
+	call out (out_magic_arg, ctrlstr)
+end
+
+
+# OUTLINE -- Output a string and append a newline to flush the output buffer.
+
+procedure outline (out, text)
+
+extern	out()
+char	text[ARB]
+errchk	outstr
+
+begin
+	call outstr (out, text)
+	call outc (out, '\n')
+end
diff --git a/pkg/system/help/lroff/rawcopy.x b/pkg/system/help/lroff/rawcopy.x
new file mode 100644
index 00000000..d303b96e
--- /dev/null
+++ b/pkg/system/help/lroff/rawcopy.x
@@ -0,0 +1,26 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"lroff.h"
+
+.help rawcopy
+.nf ________________________________________________________________________
+RAWCOPY -- Copy an unformatted help block without modification, except
+for moving to the desired left margin.  Stop only when the .endhelp
+directive is seen, or at EOF.  Ignore all other directives.
+.endhelp ___________________________________________________________________
+
+procedure rawcopy (in, out, linebuf)
+
+extern	in(), out()
+char	linebuf[ARB]
+int	ip, in(), input(), nextcmd()
+errchk	input, outline
+include	"lroff.com"
+
+begin
+	while (input (in, linebuf) != EOF)
+	    if (nextcmd (linebuf, ip) == ENDHELP)
+		break
+	    else
+		call outline (out, linebuf)
+end
diff --git a/pkg/system/help/lroff/section.x b/pkg/system/help/lroff/section.x
new file mode 100644
index 00000000..3df732c8
--- /dev/null
+++ b/pkg/system/help/lroff/section.x
@@ -0,0 +1,224 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	<ctype.h>
+include	"lroff.h"
+
+# NEW_SECTION -- Begin a section heading.  Argument is the number of lines
+# to skip.  Output section heading string given on next input line.  Reset left
+# margin and cancel out any LS indents.
+
+procedure new_section (in, out, linebuf, ip)
+
+extern	in(), out()
+char	linebuf[ARB]
+int	ip
+
+int	inbold(), in(), lgetarg()
+errchk	skiplines, inbold, outline
+include	"lroff.com"
+
+begin
+	sh_nskip = lgetarg (linebuf, ip, sh_nskip)
+	call skiplines (out, sh_nskip)
+	left_margin = perm_left_margin
+	call testpage (out, DEF_TPNLINES)
+
+	if (inbold (in, linebuf) != EOF)
+	    call outline (out, linebuf)
+
+	call init_ls()
+end
+
+
+# NEW_NUMBERED_SECTION -- Begin a numbered section heading.  Arguments are
+# the number of lines to skip and the section level to be incremented (default
+# is 1).  If only one arg is given, we assume it is the section level.
+# Output section number followed by section heading string given on next input
+# line.  Reset left margin and cancel out any LS indents.
+
+procedure new_numbered_section (in, out, linebuf, ip)
+
+extern	in(), out()
+char	linebuf[ARB]
+int	ip
+
+int	i, n
+int	inbold(), in(), lgetarg(), strlen()
+errchk	skiplines, sprintf, pargi, outstr, outc, inbold, outline
+include	"lroff.com"
+
+begin
+	# Get level, nskip arguments.
+	n = max (1, min (MAX_NHLEVEL, lgetarg (linebuf, ip, 1)))
+	nh_nskip = lgetarg (linebuf, ip, nh_nskip)
+
+	call skiplines (out, nh_nskip)
+	left_margin = perm_left_margin
+	call testpage (out, DEF_TPNLINES)
+
+	# Increment the desired section number; zero all higher numbered
+	# section counters.
+
+	nh_level[n] = nh_level[n] + 1
+	call amovki (0, nh_level[n+1], MAX_NHLEVEL - n)
+
+	# Output the section number followed by a blank and then the section
+	# label.
+
+	linebuf[1] = EOS
+	do i = 1, n {
+	    call sprintf (linebuf[strlen(linebuf)+1], SZ_IBUF, "%d.")
+		call pargi (nh_level[i])
+	}
+
+	# Cancel the final "." if subsection heading.  Add a blank.
+	if (n > 1 && linebuf[strlen(linebuf)] == '.')
+	    linebuf[strlen(linebuf)] = EOS
+	call outstr (out, linebuf)
+	call outc (out, BLANK)
+
+	# Get section label from next input line, write that out on the same
+	# line in standout mode, then terminate the line.
+
+	if (inbold (in, linebuf) != EOF)
+	    call outline (out, linebuf)
+
+	call init_ls()
+end
+
+
+# INIT_NH -- Initialize section numbering.
+
+procedure init_nh()
+
+include	"lroff.com"
+
+begin
+	call amovki (0, nh_level, MAX_NHLEVEL)
+end
+
+
+# NEW_INDENTED_SECTION -- Begin an indented section heading.  Optional
+# arguments are the number of spaces to indent subsequent text and the number
+# of lines to skip.  Output section heading string given on next input line.
+# Reset left margin and cancel out any LS indents.
+
+procedure new_indented_section (in, out, linebuf, ip)
+
+extern	in(), out()
+char	linebuf[ARB]
+int	ip
+
+int	inbold(), in(), lgetarg()
+errchk	skiplines, inbold, outline
+include	"lroff.com"
+
+begin
+	ih_indent = lgetarg (linebuf, ip, ih_indent)
+	ih_nskip = lgetarg (linebuf, ip, ih_nskip)
+
+	call skiplines (out, ih_nskip)
+	left_margin = perm_left_margin
+	call testpage (out, DEF_TPNLINES)
+
+	# Read in and output the section heading in boldface.
+	if (inbold (in, linebuf) != EOF)
+	    call outline (out, linebuf)
+
+	# Reset the left margin and cancel out any LS indents.
+	left_margin = max (perm_left_margin, min (right_margin,
+	    perm_left_margin + ih_indent))
+
+	call init_ls()
+end
+
+
+# INBOLD -- Input a line in standout mode.  If the line is already all in
+# upper case, do not use standout mode.  The input procedure processes
+# all font escape sequences.  We must get the raw input line by calling the
+# user input procedure, then pass it on to input() enclosed in \fB...\fR
+# font escape sequences, to enable standout mode.
+
+int procedure inbold (in, user_linebuf)
+
+extern	in()
+int	in()
+char	user_linebuf[ARB]
+
+pointer	sp, ip, lbuf, first
+int	save_in_magic_arg, status
+int	stropen(), input()
+extern	getline()
+errchk	salloc, stropen, input
+include	"lroff.com"
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE + 3 + 3, TY_CHAR)
+
+	# Deposit escape sequence to turn bold on.
+	call strcpy ("\\fB", Memc[lbuf], ARB)
+
+	# Read in the input line after the three char escape sequence.
+	if (in (in_magic_arg, Memc[lbuf+3]) == EOF) {
+	    call sfree (sp)
+	    return (EOF)
+	}
+
+	# Scan the line to see if there are any lower case characters.
+	# If all upper case, omit the mode control sequences (this procedure
+	# becomes equivalent to input()).
+
+	first = lbuf + 3
+	for (ip=lbuf+3;  Memc[ip] != EOS;  ip=ip+1)
+	    if (IS_LOWER (Memc[ip])) {
+		first = lbuf
+		break
+	    }
+
+	# Step on the newline if there is one, then add the \fR
+	# sequence to turn bold off.
+	if (first == lbuf) {
+	    for (ip=lbuf;  Memc[ip] != EOS;  ip=ip+1)
+		;
+	    if (Memc[ip-1] == '\n')
+		Memc[ip-1] = EOS
+	    call strcat ("\\fR\n", Memc[lbuf], ARB)
+	}
+
+	# Now open the string as a file and call input to process it
+	# into our caller's buffer.  We must save and restore the input
+	# magic argument, set to the fd of the string file when input is
+	# called.  This is a good example of the disadvantages of commons...
+
+	save_in_magic_arg = in_magic_arg
+	in_magic_arg = stropen (Memc[first], ARB, READ_ONLY)
+	status = input (getline, user_linebuf)
+	call close (in_magic_arg)
+	in_magic_arg = save_in_magic_arg
+
+	call sfree (sp)
+	return (status)
+end
+
+
+# TESTPAGE -- If forms mode is enabled, output the control code for a test
+# page followed by the number of lines to test for.  Test page tests if the
+# specified number of lines are left on a page, and breaks the page if not.
+
+procedure testpage (out, nlines)
+
+extern	out()
+int	nlines
+char	ctrlstr[2]
+include	"lroff.com"
+
+begin
+	if (foflag == YES) {
+	    ctrlstr[1] = FC_TESTPAGE
+	    ctrlstr[2] = nlines
+	    ctrlstr[3] = EOS
+	    call out (out_magic_arg, ctrlstr)
+	}
+end
diff --git a/pkg/system/help/lroff/skiplines.x b/pkg/system/help/lroff/skiplines.x
new file mode 100644
index 00000000..0449dc50
--- /dev/null
+++ b/pkg/system/help/lroff/skiplines.x
@@ -0,0 +1,19 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"lroff.h"
+
+# SKIPLINES -- Skip one or more lines on the output, i.e., break the current
+# output line and add a few empty lines.
+
+procedure skiplines (out, nlines)	
+
+extern	out()
+int	nlines, i
+errchk	breakline, outc
+
+begin
+	call breakline (out, NJ)
+
+	do i = 1, nlines
+	    call outc (out, '\n')
+end
diff --git a/pkg/system/help/lroff/textlen.x b/pkg/system/help/lroff/textlen.x
new file mode 100644
index 00000000..e192314f
--- /dev/null
+++ b/pkg/system/help/lroff/textlen.x
@@ -0,0 +1,20 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	"lroff.h"
+
+# TEXTLEN -- Return the number of printable characters in a text string.
+
+int procedure textlen (text_string)
+
+char	text_string[ARB]
+int	ip, nchars
+
+begin
+	nchars = 0
+	for (ip=1;  text_string[ip] != EOS;  ip=ip+1)
+	    if (!INVISIBLE (text_string[ip]))
+		nchars = nchars + 1
+
+	return (nchars)
+end
diff --git a/pkg/system/help/lroff/textout.x b/pkg/system/help/lroff/textout.x
new file mode 100644
index 00000000..663bccc8
--- /dev/null
+++ b/pkg/system/help/lroff/textout.x
@@ -0,0 +1,140 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	"lroff.h"
+
+# TEXTOUT -- Process a line of text.  Move words from the text buffer into
+# the word buffer WBUF, maintaining an array of pointers to the words, until
+# an output line has been filled.  Leading whitespace is part of the word,
+# if it is the first word on a line (thus we get paragraph indents).
+# Thereafter only trailing whitespace is included in the word.  The last word
+# on a line gets one trailing space, unless the last char is a period, in which
+# case it gets two.  Otherwise whitespace at the end of the input text line is
+# stripped.  BREAKLINE is subsequently called to reassemble the words to form
+# an output line.
+# 
+# WBUF is the word buffer (a set of strings separated by EOS markers).  WP is
+# a pointer to the next available char in WBUF.  NWORDS is the number of words
+# in the buffer.  WORDS is a pointer to the array of word pointers.  We do not
+# check for word buffer overflow because the word buffer is allocated large
+# enough to accommodate the worst case (the buffer is flushed when an output
+# line is filled, which always happens before the buffer overflows).  WCOLS is
+# the number of printable characters in the word buffer.  The word buffer
+# variables are all stored in the "words" common for use by TEXTOUT and
+# BREAKLINE.  Set_wordbuf() must be called upon startup and shutdown to
+# allocate/deallocate the word buffer.
+
+procedure set_wordbuf (max_words)
+
+int	max_words		#I output word buffer size
+
+int	word_buffer_size
+errchk	malloc
+
+include	"lroff.com"
+include	"words.com"
+
+begin
+	word_buffer_size = (max_words * 2) + SZ_LINE		# worst case
+	if (max_words <= 0 && words != NULL) {
+	    call mfree (wbuf, TY_CHAR)
+	    call mfree (words, TY_POINTER)
+	} else {
+	    call malloc (wbuf, word_buffer_size, TY_CHAR)
+	    call malloc (words, max_words, TY_POINTER)
+	    wp = wbuf
+	    nwords = 0
+	    wcols = 0
+	}
+end
+
+
+# TEXTOUT -- Output a newline delimited line of text.
+
+procedure textout (out, text)
+
+extern	out()
+char	text[1]
+
+char	ch
+int	ip_save, wcols_save, ip
+errchk	breakline
+include	"lroff.com"
+include	"words.com"
+
+begin
+	if (wbuf == NULL || words == NULL)
+	    call error (1, "No Lroff word buffer allocated")
+
+	for (ip=1;  text[ip] != EOS;  ) {
+	    # Set up descriptors of new word.  Save the input pointer in case
+	    # the output line fills and we have to "put the word back".
+
+	    Memi[words+nwords] = wp			# word pointer 
+	    ip_save = ip
+	    wcols_save = wcols
+
+	    # The following is a nop except at the beginning of a line.
+	    for (;  text[ip] == BLANK;  ip=ip+1) {
+		Memc[wp] = BLANK
+		wp = wp + 1
+		wcols = wcols + 1
+	    }
+
+	    # Copy the word itself.
+	    for (ch=text[ip];  ch != BLANK && ch != EOS;  ch=text[ip]) {
+		Memc[wp] = ch
+		wp = wp + 1
+		if (!INVISIBLE (ch))
+		    wcols = wcols + 1
+		ip = ip + 1
+	    }
+
+	    # And then any trailing whitespace.
+	    for (;  text[ip] == BLANK;  ip=ip+1) {
+		Memc[wp] = BLANK
+		wp = wp + 1
+		wcols = wcols + 1
+	    }
+
+	    # End of word string.
+	    Memc[wp] = EOS
+	    wp = wp + 1
+
+	    # If line has been filled, call breakline to format output line
+	    # and send it out.  Put word which caused break back for next line.
+	    # Do not put word back if it is the first word, or we will have
+	    # an infinite loop.
+	    if (wcols > (right_margin - left_margin + 1) && nwords > 0) {
+		ip = ip_save
+		wcols = wcols_save
+		wp = Memi[words+nwords]
+		call breakline (out, JU)
+	    } else
+		nwords = nwords + 1
+	}
+
+	# Strip trailing whitespace at the end of an input line.  If input
+	# line ends with a period, assume it is a sentence and add a blank.
+	# Otherwise add a blank to separate words when output line is filled.
+	# If a sentence ends within a line, the user is responsible for placing
+	# two spaces after the period.
+
+	if (nwords > 0) {
+	    for (wp=wp-2;  Memc[wp] == BLANK && wp > wbuf;  wp=wp-1)
+		wcols = wcols - 1
+	    if (Memc[wp] == '.') {
+		wp = wp + 1
+		Memc[wp] = BLANK
+		wcols = wcols + 1
+	    }
+	    wp = wp + 1					# point to next avail
+
+	    Memc[wp] = BLANK				# need at least one
+	    wp = wp + 1
+	    wcols = wcols + 1
+
+	    Memc[wp] = EOS
+	    wp = wp + 1
+	}
+end
diff --git a/pkg/system/help/lroff/words.com b/pkg/system/help/lroff/words.com
new file mode 100644
index 00000000..d0c220da
--- /dev/null
+++ b/pkg/system/help/lroff/words.com
@@ -0,0 +1,9 @@
+# Common for Textout()/Breakline().
+
+pointer	wbuf			# the word buffer, a string buffer
+pointer	wp			# pointer to next available char in word buffer
+pointer	words			# array of word pointers
+int	nwords			# number of words in word buffer
+int	wcols			# number of printable columns in word buffer
+
+common	/wrdcom/ wbuf, wp, words, nwords, wcols
diff --git a/pkg/system/help/manout.x b/pkg/system/help/manout.x
new file mode 100644
index 00000000..b81abc8b
--- /dev/null
+++ b/pkg/system/help/manout.x
@@ -0,0 +1,330 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<chars.h>
+include	<ctype.h>
+include	"help.h"
+
+.help man_output, man_close
+.nf ___________________________________________________________________________
+Process a stream of output lines containing forms control codes.  This mode is
+used when manual page format is desired.  Input is a sequence of manual pages.
+A manual page consists of a formfeed, a two line page header, and then an
+arbitrary number of lines of text.  Pages are broken every H_NLPP lines,
+the page header is printed at the top of each page, and the page number is
+printed at the bottom of each page.  The following special directives may
+be inserted in the text as control character forms codes; the forms code must
+be the first character in a line but may be followed by text.
+
+	FC_BREAKPAGE		Start a new page, unless already at the
+				top of a page.
+
+	FC_TESTPAGE n		Test if there are at least N (binary code)
+				lines left on the page.  If no, break the
+				page before more text is output.
+
+	FC_STARTKEEP		Start spooling lines into an internal buffer.
+				Try to output the keep block all on one page.
+
+	FC_ENDKEEP		Stop spooling and output the keep block.
+
+MAN_CLOSE should be called at program termination to return buffer space and
+reinit variables.
+.endhelp ______________________________________________________________________
+
+define	LEN_HEADER		2
+define	LEN_FOOTER		3
+define	LEN_KEEP		1024
+define	SZ_PAGELABEL		6
+define	SZ_HEADER_GAP		2
+
+define	IN_HEADER		1
+define	IN_TEXT			2
+define	IN_KEEP			3
+define	FLUSH_KEEP		4
+
+
+# MAN_OUTPUT -- Output a line, optionally preceded by one or more forms control
+# characters.  Start a new manual page, save lines in header, save lines in keep
+# buffer, or write line to output.
+
+procedure man_output (fd, lbuf, nlines_per_page, lmargin, rmargin)
+
+int	fd			# output file
+char	lbuf[ARB]		# next output line
+int	nlines_per_page
+int	lmargin, rmargin
+
+char	end_keep_chars[3]
+int	ip, mode, max_lineno, i
+
+int	nlines_in_keep, nlines_in_header, init
+int	lineno, pageno, center_col, nlpp
+pointer	header_lines[LEN_HEADER], keep_lines[LEN_KEEP]
+common	/hlpman/ lineno, pageno, nlpp, center_col, init,
+	nlines_in_header, nlines_in_keep, header_lines, keep_lines
+
+int	stridx()
+pointer	man_putstr()
+data	end_keep_chars /FC_STARTKEEP, FC_BREAKPAGE, '\f', EOS/
+
+begin
+	if (init == YES) {
+	    # Initialize internal variables.
+	    lineno = 0
+	    pageno = 0
+	    mode = IN_TEXT
+	    nlines_in_header = 0
+	    nlines_in_keep = 0
+	    nlpp = nlines_per_page
+	    max_lineno = max(1, min(LEN_KEEP, nlpp - LEN_FOOTER))
+	    center_col = (lmargin + rmargin) / 2
+	    init = NO
+	}
+
+	for (ip=1;  lbuf[ip] != EOS;  ) {
+	    switch (mode) {
+
+	    case IN_HEADER:
+		# Just started a new manual page; still accumulating header
+		# lines.  Forms control chars not recognized in header area.
+		# Output header to start new page as soon as we have all the
+		# lines.
+
+		nlines_in_header = nlines_in_header + 1
+		header_lines[nlines_in_header] = man_putstr (lbuf[ip])
+		if (nlines_in_header == LEN_HEADER) {
+		    mode = IN_TEXT
+		    call man_breakpage (fd, header_lines, nlines_in_header,
+			lineno, pageno, nlpp, center_col)
+		    pageno = 1
+		}
+		break
+
+	    case IN_KEEP:
+		# Any forms control character in a keep ends the keep, in case
+		# the KE is missing.  Don't bump input pointer unless sentinel
+		# is ENDKEEP, else forms control character will be skipped.
+
+		if (lbuf[ip] == FC_ENDKEEP) {
+		    mode = FLUSH_KEEP
+		    ip = ip + 1
+		} else if (stridx (lbuf[ip], end_keep_chars) > 0) {
+		    mode = FLUSH_KEEP
+		} else if (lineno + nlines_in_keep > max_lineno) {
+		    mode = FLUSH_KEEP
+		} else {
+		    # Put line in keep buffer.
+		    nlines_in_keep = nlines_in_keep + 1
+		    keep_lines[nlines_in_keep] = man_putstr (lbuf[ip])
+		    break
+		}
+
+	    case FLUSH_KEEP:
+		# Break page and output header if insufficient room left on
+		# page.
+		if (lineno + nlines_in_keep > max_lineno)
+		    call man_breakpage (fd, header_lines, nlines_in_header,
+			lineno, pageno, nlpp, center_col)
+
+		# Output contents of keep buffer and return buffer space.
+		for (i=1;  i <= nlines_in_keep;  i=i+1) {
+		    call putline (fd, Memc[keep_lines[i]])
+		    call mfree (keep_lines[i], TY_CHAR)
+		    lineno = lineno + 1
+		}
+		nlines_in_keep = 0
+		mode = IN_TEXT
+
+	    case IN_TEXT:
+		# Not in any special mode.  Check for forms chars and loop if
+		# one is found.  Otherwise output line and bump line pointer.
+		# Break page when it fills.
+
+		switch (lbuf[ip]) {
+		case FC_BREAKPAGE:
+		    call man_breakpage (fd, header_lines, nlines_in_header,
+			lineno, pageno, nlpp, center_col)
+		    ip = ip + 1
+
+		case FC_TESTPAGE:
+		    # The first char following the testpage forms control char
+		    # is the number of lines to test for, in binary.
+
+		    if (lineno + lbuf[ip+1] > max_lineno)
+			call man_breakpage (fd, header_lines, nlines_in_header,
+			    lineno, pageno, nlpp, center_col)
+		    ip = ip + 2
+
+		case FC_STARTKEEP:
+		    mode = IN_KEEP
+		    ip = ip + 1
+		case FC_ENDKEEP:
+		    ip = ip + 1
+
+		case '\f':
+		    # Start accumulating new header.  Page cannot be broken
+		    # until we have the header.
+
+		    for (i=1;  i <= nlines_in_header;  i=i+1)
+			call mfree (header_lines[i], TY_CHAR)
+		    nlines_in_header = 0
+		    mode = IN_HEADER
+		    ip = ip + 1
+
+		default:
+		    # This is the case that gets called most often; output an
+		    # ordinary line, not in any special mode.  Eat blank lines
+		    # at top of page.
+
+		    if (lineno <= nlines_in_header + SZ_HEADER_GAP + 1) {
+			for (i=ip;  IS_WHITE (lbuf[i]);  i=i+1)
+			    ;
+			if (lbuf[i] == '\n' || lbuf[i] == EOS)
+			    break
+		    }
+		    call putline (fd, lbuf[ip])
+		    lineno = lineno + 1
+		    if (lineno > max_lineno)
+			call man_breakpage (fd, header_lines, nlines_in_header,
+			    lineno, pageno, nlpp, center_col)
+		    break
+		}
+	    }
+	}	   
+end
+
+
+# MAN_INIT -- Called to initialize the manpage internal data structures.
+
+procedure man_init()
+
+int	i
+bool	first_time
+data	first_time /true/
+
+int	nlines_in_keep, nlines_in_header, init
+int	lineno, pageno, center_col, nlpp
+pointer	header_lines[LEN_HEADER], keep_lines[LEN_KEEP]
+common	/hlpman/ lineno, pageno, nlpp, center_col, init,
+	nlines_in_header, nlines_in_keep, header_lines, keep_lines
+
+begin
+	# Clean up in the event of an interrupt.
+	if (!first_time) {
+	    for (i=1;  i <= nlines_in_header;  i=i+1)
+		call mfree (header_lines[i], TY_CHAR)
+	    for (i=1;  i <= nlines_in_keep;  i=i+1)
+		call mfree (keep_lines[i], TY_CHAR)
+	}
+
+	init = YES
+	first_time = false
+end
+
+
+# MAN_CLOSE -- Called at program termination to write the page number at the
+# bottom of the last manual page.
+
+procedure man_close (fd)
+
+int	fd
+int	nlines_in_keep, nlines_in_header, i
+int	lineno, pageno, center_col, nlpp, init
+pointer	header_lines[LEN_HEADER], keep_lines[LEN_KEEP]
+common	/hlpman/ lineno, pageno, nlpp, center_col, init,
+	nlines_in_header, nlines_in_keep, header_lines, keep_lines
+
+begin
+	if (init == YES)
+	    return
+
+	# Trash any data in keep buffer if missing the KE by the time we are
+	# called (at program termination).
+
+	call man_breakpage (fd, header_lines, 0, lineno, pageno, nlpp,
+	    center_col)
+
+	for (i=1;  i <= nlines_in_header;  i=i+1)
+	    call mfree (header_lines[i], TY_CHAR)
+	for (i=1;  i <= nlines_in_keep;  i=i+1)
+	    call mfree (keep_lines[i], TY_CHAR)
+
+	nlines_in_header = 0
+	nlines_in_keep = 0
+	init = YES
+end
+
+
+# MAN_BREAKPAGE -- Advance to bottom of page, print page number, bump page
+# number, init line counter, formfeed, output page header of new page.
+
+procedure man_breakpage (fd, header_lines, nlines_in_header, lineno, pageno,
+			 nlines_per_page, center_col)
+int	fd
+pointer	header_lines[ARB]
+int	nlines_in_header
+int	lineno, pageno
+int	nlines_per_page
+int	center_col
+
+char	pagelabel[SZ_PAGELABEL]
+int	i, destcol
+int	strlen()
+
+begin
+	# Output page footer if there is anything on the page and if there
+	# was a page header.
+
+	if (lineno > 0 && nlines_in_header > 0) {
+	    while (lineno <= nlines_per_page - 1) {
+		call putline (fd, "\n")
+		lineno = lineno + 1
+	    }
+
+	    call sprintf (pagelabel, SZ_PAGELABEL, "-%d-")
+		call pargi (pageno)
+
+	    destcol = max (1, center_col - strlen (pagelabel) / 2)
+	    for (i=1;  i < destcol;  i=i+1)
+		call putci (fd, ' ')
+	    call putline (fd, pagelabel)
+	    call putci (fd, '\n')
+	}
+
+	# Break page and output page header if there are any header lines.
+	# Leave two blank lines between header and first line of text.
+
+	if (nlines_in_header > 0) {
+	    if (pageno > 0)
+		call putci (fd, '\f')
+	    lineno = 1
+	    pageno = pageno + 1
+
+	    for (i=1;  i <= nlines_in_header;  i=i+1) {
+		call putline (fd, Memc[header_lines[i]])
+		lineno = lineno + 1
+	    }
+	    for (i=1;  i <= SZ_HEADER_GAP;  i=i+1) {
+		call putline (fd, "\n")
+		lineno = lineno + 1
+	   }
+	}
+end
+
+
+# MAN_PUTSTR -- Save a string on the heap and return a pointer to the string.
+
+pointer procedure man_putstr (str)
+
+char	str[ARB]
+int	nchars
+pointer	bp
+int	strlen()
+
+begin
+	nchars = strlen (str)
+	call malloc (bp, nchars, TY_CHAR)
+	call strcpy (str, Memc[bp], nchars)
+
+	return (bp)
+end
diff --git a/pkg/system/help/mkhelpdb.hlp b/pkg/system/help/mkhelpdb.hlp
new file mode 100644
index 00000000..d3534cf6
--- /dev/null
+++ b/pkg/system/help/mkhelpdb.hlp
@@ -0,0 +1,75 @@
+.help mkhelpdb Feb89 softools
+.ih
+NAME
+mkhelpdb -- update the help database
+.ih
+USAGE
+mkhelpdb helpdir helpdb
+.ih
+PARAMETERS
+.ls helpdir = "lib$root.hd"
+The filename of the root help directory file (".hd" file) defining the
+help tree to be updated.  By convention this is \fIroot.hd\fR in some
+directory.
+.le
+.ls helpdb = "lib$helpdb.mip"
+The filename of the help database file to be written.  By convention this
+is \fIhelpdb.mip\fR in some directory (the ".mip" signifies that the file
+format is machine independent).
+.le
+.ls verbose = no
+If this switch is enabled, \fImkhelpdb\fR will print a detailed description
+of the help database as it is being compiled.  A more concise summary listing
+only the packages and the number of help modules in each package is printed
+by default.
+.le
+.ih
+DESCRIPTION
+The \fImkhelpdb\fR task descends a tree of help directory (".hd") files and
+compiles a binary help database from the information therein.  The help
+database is used to speed global searches when help is requested for a
+module, the ".hlp" file for which might be anywhere in the system.
+The help database defines the packages and modules in the help database,
+and stores the filenames of the associated help files.  No actual help text
+is stored in the help database, only sufficient index information to find
+the help files when the \fIhelp\fR task is run.  The help directory files
+are text files which define the packages and modules in the help database.
+The format of these files is self explanatory hence is documented by example
+only.
+
+By default, \fImkhelpdb\fR recompiles the standard IRAF help database,
+although any other similar database may be recompiled by changing the values
+of the parameters \fIhelpdir\fR and \fIhelpdb\fR.  The standard
+IRAF help database is rooted in the file \fBlib$root.hd\fR.
+
+The help database must be updated whenever a new help module (e.g., manual
+page) is added, deleted, or renamed.  It is also necessary for sites receiving
+a source only version of IRAF to run \fImkhelpdb\fR to rebuild the help
+database once the system is up, since the database is a binary file and
+is not included in a source only distribution.  It is not necessary to rerun
+\fImkhelpdb\fR when an existing manual page is edited, since only index
+information is stored in the database.
+
+The \fIhelp\fR utilities make use of the following types of files.  Examples
+of these files will be found throughout the IRAF directories.
+
+.nf
+	.hd		help directory file (tree structured)
+	.hlp		manual page
+	.men		package menu (module listing)
+.fi
+.ih
+EXAMPLES
+1. Update the standard IRAF help database.
+
+.nf
+	cl> softools
+	so> mkhelpdb helpdir=lib$root.hd helpdb=lib$helpdb.mip
+.fi
+
+2. Update the NOAO package help database.
+
+	so> mkhelpdb helpdir=noao$lib/root.hd helpdb=noao$lib/helpdb.mip
+.ih
+SEE ALSO
+hdbexamine, help
diff --git a/pkg/system/help/mkpkg b/pkg/system/help/mkpkg
new file mode 100644
index 00000000..3cb437ec
--- /dev/null
+++ b/pkg/system/help/mkpkg
@@ -0,0 +1,36 @@
+# Make the HELP and LROFF library modules.
+
+$checkout libpkg.a ../
+$update   libpkg.a
+$checkin  libpkg.a ../
+$exit
+
+libpkg.a:
+	@lroff
+	@xhelp
+
+	filetemp.x	help.h <ctype.h>
+	getoption.x	help.h <ctype.h>
+	hbgetblk.x	help.h <ctype.h>
+	helpdb.x	help.h helpdir.h <error.h> <finfo.h> <fset.h> <mach.h>\
+			<time.h>
+	helpdir.x	help.h helpdir.h <ctype.h> <error.h>
+	hinput.x	help.h <ctype.h>
+	houtput.x	help.h <ttyset.h> <chars.h> <fset.h>
+	manout.x	help.h <chars.h> <ctype.h>
+	modlist.x	help.h helpdir.h <ctype.h> <error.h>
+	modtemp.x	help.h <ctype.h>
+	prblkhdr.x	help.h
+	prdir.x		help.h <error.h> <ttyset.h>
+	prfile.x	help.h <finfo.h> <time.h> <ctype.h>
+	prfnames.x	help.h helpdir.h
+	prhelp.x	help.h helpdir.h <error.h>
+	prhlpblk.x	help.h <error.h>
+	prmodname.x	help.h
+	prsummary.x	help.h
+	t_hdbexamine.x	help.h <error.h>
+	t_help.x	help.h <error.h> <fset.h> <finfo.h>
+	t_lroff.x	
+	t_mkhelpdb.x	<error.h>
+	tlist.x		help.h <ctype.h>
+	;
diff --git a/pkg/system/help/modlist.x b/pkg/system/help/modlist.x
new file mode 100644
index 00000000..dce63f59
--- /dev/null
+++ b/pkg/system/help/modlist.x
@@ -0,0 +1,200 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	<error.h>
+include	"help.h"
+include	"helpdir.h"
+
+.help modlist
+.nf ___________________________________________________________________________
+MODLIST -- Routines for expanding a template into a list of module names,
+and accessing successive elements of the list, given a template and a help
+directory defining the pattern matching domain as input.
+
+The permissible complexity of a help template is somewhat less than that of
+a file template; we do not have to deal with logical directories, virtual
+file names, and list files.  The class of characters permitted in a CL
+module name is fairly limited.  By the time we are called, the template
+list will already have been expanded into a list of templates, wherein
+each list element has a full package name (the package field has already
+been expanded).  Our job is to process a single template.
+
+	*			all modules
+	im*			all modules whose names start with 'im'
+	[a-z]?*			all modules beg. with lowercase letter 
+	alpha			all modules alpha*
+
+Note that a simple alphanumeric name like 'alpha' is treated as an
+abbreviation, i.e., as 'alpha?*'.  All occurrences of the closure metacharacter
+* are converted to ?*, except after a character class.
+
+Entry points in this package include:
+
+	ml_open		set up to match and extract module names
+	ml_read		get next module name which matches pattern
+	ml_close	called when done to return buffers
+.endhelp ______________________________________________________________________
+
+define	LEN_MLSTRUCT	165		# module list structure
+define	SZ_PATBUF	159
+
+define	ML_HP		Memi[$1]	# help directory
+define	ML_INDEX	Memi[$1+1]	# index of next module in helpdir
+define	ML_SIMPAT	Memi[$1+2]	# pattern is simple (no metacharacters)
+define	ML_PATBUF	Memc[P2C($1+5)] # encoded pattern buffer
+
+
+# ML_OPEN -- Set up to read module names from a directory which match
+# a certain pattern.  Allocate descriptor, save pointer to directory,
+# encode pattern, rewind module index.
+
+pointer procedure ml_open (hp, template)
+
+pointer	hp				# help directory
+char	template[ARB]			# "alpha", "*", etc.
+
+bool	metachars
+int	ip
+pointer	ml, sp, pat, op
+int	patmake()
+
+begin
+	call smark (sp)
+	call salloc (pat, SZ_LINE, TY_CHAR)
+
+	iferr (call calloc (ml, LEN_MLSTRUCT, TY_STRUCT))
+	    call erract (EA_FATAL)
+	ML_INDEX(ml) = 1
+	ML_HP(ml) = hp
+
+	# Match only at beginning of module name.  Remember to scrap this
+	# metacharacter if matching a SIMPAT simple pattern.
+
+	Memc[pat] = '^'		
+	op = pat + 1
+
+	# If no metacharacters present in pattern set SIMPAT flag so that we
+	# can use a more efficient pattern matching algorithm.  If closure
+	# metacharacter is present and it does not follow ], replace it with ?*.
+
+	metachars = false
+	for (ip=1;  template[ip] != EOS && op-pat < SZ_LINE;  ip=ip+1) {
+	    switch (template[ip]) {
+	    case '?', '[', ']':
+		metachars = true
+	    case '*':
+		metachars = true
+		if (ip == 1 || (ip > 1 && template[ip-1] != ']')) {
+		    Memc[op] = '?'
+		    op = op + 1
+		}
+	    }
+	    Memc[op] = template[ip]
+	    op = op + 1
+	}
+	Memc[op] = EOS
+
+	if (!metachars) {
+	    ML_SIMPAT(ml) = YES
+	    # Copy pattern string, stripping off the leading '^'.
+	    call strcpy (Memc[pat+1], ML_PATBUF(ml), SZ_PATBUF)
+	} else {
+	    ML_SIMPAT(ml) = NO
+	    if (patmake (Memc[pat], ML_PATBUF(ml), SZ_PATBUF) == ERR)
+		call error (16, "Cannot encode help pattern")
+	}
+
+	call sfree (sp)
+	return (ml)
+end
+
+
+# ML_READ -- Get next module name from help directory matching the
+# encoded pattern.  Return EOF when directory is exhausted.  We directly
+# access the HD structure for efficiency.
+
+int procedure ml_read (ml, modnum, module, maxch)
+
+pointer	ml			# pointer to module list descriptor
+int	modnum			# module number (output)
+char	module[ARB]		# module name (output)
+int	maxch
+
+bool	match_found
+char	first_char
+int	m
+pointer	hp, sbuf, mp, modname
+int	patmatch(), ml_patmatch(), gstrcpy()
+errchk	patmatch
+
+begin
+	if (ml == NULL)
+	    call error (17, "help.ml_read called with null descriptor")
+	hp = ML_HP(ml)
+	sbuf = HD_SBUF(hp)
+	first_char = ML_PATBUF(ml)
+	m = ML_INDEX(ml)
+	match_found = false
+
+	# Scan down the list of modules in the package, returning the name
+	# of each matching the pattern in successive calls.  Return EOF when
+	# the list is exhausted.
+
+	while (m <= HD_NMODULES(hp)) {
+	    mp = HD_MODULE(hp, m)
+	    modname = sbuf + M_NAME(mp)
+	    m = m + 1
+	    if (ML_SIMPAT(ml) == YES) {
+		if (Memc[modname] == first_char)
+		    if (ml_patmatch (Memc[modname], ML_PATBUF(ml)) > 0) {
+			match_found = true
+			break
+		    }
+	    } else if (patmatch (Memc[modname], ML_PATBUF(ml)) > 0) {
+		match_found = true
+		break
+	    }
+	}
+
+	ML_INDEX(ml) = m
+	modnum = m - 1
+
+	# If no modules were found matching the template then we can only be
+	# at the end of the list.
+
+	if (match_found)
+	    return (gstrcpy (Memc[modname], module, maxch))
+	else {
+	    module[1] = EOS
+	    return (EOF)
+	}
+end
+
+
+# ML_PATMATCH -- Determine if a SIMPAT simple pattern is an abbreviation of
+# a module name.  The pattern is simple, i.e., contains no metacharacters.
+
+int procedure ml_patmatch (module, pattern)
+
+char	module[ARB]		# module name to be compared to pattern
+char	pattern[ARB]		# pattern of the form "^abbrev"
+int	ip
+
+begin
+	for (ip=1;  pattern[ip] != EOS;  ip=ip+1)
+	    if (pattern[ip] != module[ip])
+		return (0)
+
+	return (1)
+end
+
+
+# ML_CLOSE -- Return buffers allocated by open_template.
+
+procedure ml_close (ml)
+
+pointer	ml
+
+begin
+	call mfree (ml, TY_STRUCT)
+end
diff --git a/pkg/system/help/modtemp.x b/pkg/system/help/modtemp.x
new file mode 100644
index 00000000..d539a2f2
--- /dev/null
+++ b/pkg/system/help/modtemp.x
@@ -0,0 +1,190 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	<syserr.h>
+include	"help.h"
+
+# DO_MODULE_TEMPLATE -- Called with a template defining the packages and
+# modules for which help is desired, and the control parameters defining
+# the type of help desired.  Expand the template into a list of packages
+# and modules and process the help for each module.
+# 
+# Most of the work here is done by HD_LOAD, which opens and interprets a
+# help directory, and TL_OPEN/TL_READ, which expand the template list.
+# Each element of the template list is a module matching pattern for a
+# single package.  When this is expanded we get the names of the individual
+# modules and call PRINT_HELP to print the help text for a single module
+# of a particular package.  In the worst case (template "*.*"), we might
+# process all modules in the help database.
+
+procedure do_module_template (db, tlist, ctrl)
+
+pointer	db			# help database to be used
+char	tlist[ARB]		# template list ("a.*,b.*,...")
+pointer	ctrl			# Help control structure
+
+int	m, pk
+bool	no_matches, matchall
+pointer	sp, fname, template, module, pakname, modtemp, curmod
+pointer	hp_sys, hp_pak, tl, ip, op, modlist
+
+bool	strne()
+int	tl_read(), ml_read(), print_help()
+int	stridxs(), ml_patmatch(), hd_findmod(), hd_getname()
+pointer	hdb_load(), tl_open(), ml_open()
+errchk	tl_read, hd_findmod, hd_getname, ml_read, hdb_load
+errchk	tl_open, ml_open, print_help
+
+begin
+	call smark (sp)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+	call salloc (template, SZ_LINE, TY_CHAR)
+	call salloc (module, SZ_FNAME, TY_CHAR)
+	call salloc (pakname, SZ_FNAME, TY_CHAR)
+	call salloc (modtemp, SZ_FNAME, TY_CHAR)
+	call salloc (curmod, SZ_FNAME, TY_CHAR)
+
+	# The help database contains the package names and references to
+	# all help modules in the package.  Load the help database index
+	# ("_index") which contains a list of all packages in the database.
+
+	hp_sys = hdb_load (db, "_index")
+
+	# Expand the template list into a list of simple "pak.mod"
+	# templates, where the package name is spelled out and only
+	# the module template remains to be expanded.  The null
+	# template list gets turned into "curpack.".
+
+	tl = tl_open (db, hp_sys, tlist, ctrl)
+	Memc[curmod] = EOS
+	no_matches = true
+	matchall = (H_ALLMODULES(ctrl) == YES)
+
+	# Process each "package_name.module_template" element of the TL
+	# list.  The package_name field is always valid.  If the module
+	# field is null, package help is desired, so print help on the
+	# package as if it were a module.  Otherwise, expand the module
+	# template for the package and process help on each module.
+
+	while (tl_read (tl, Memc[template], SZ_FNAME) != EOF) {
+	    # Extract package name.
+	    op = pakname
+	    for (ip=template;  IS_WHITE(Memc[ip]);  ip=ip+1)
+		;
+	    for (;  Memc[ip] != '.' && Memc[ip] != EOS;  ip=ip+1) {
+		Memc[op] = Memc[ip]
+		op = op + 1
+	    }
+	    Memc[op] = EOS
+
+	    # Extract module template.
+	    op = modtemp
+	    if (Memc[ip] == '.')
+		ip = ip + 1
+	    for (;  Memc[ip] != EOS;  ip=ip+1) {
+		Memc[op] = Memc[ip]
+		op = op + 1
+	    }
+	    Memc[op] = EOS
+
+	    # The following is used to count the number of matches for a module
+	    # (used to generate the "No help available" message).  The
+	    # template list will typically contain sequences like "curpak.mod",
+	    # "paki.", "pakj.", "pak1.mod", "pak2.mod", and so on (where the
+	    # pak[ij] are the full names of packages matched by an ambiguous
+	    # template).  Reset no_matches whenever the module name changes.
+	    # If the module name disappears, i.e., if the template is a package
+	    # name, then we have a match if the old module name is an
+	    # abbreviation for the package.
+
+	    if (Memc[curmod] == EOS || strne(Memc[modtemp],Memc[curmod])) {
+		if (Memc[modtemp] == EOS) {
+		    if (ml_patmatch (Memc[pakname], Memc[modtemp]) > 0)
+			no_matches = false
+		} else {
+		    if (no_matches && Memc[curmod] != EOS) {
+			call eprintf ("No help available for `%s'\n")
+			    call pargstr (Memc[curmod])
+		    }
+		    no_matches = true
+		    call strcpy (Memc[modtemp], Memc[curmod], SZ_FNAME)
+		}
+	    }
+
+	    # Search system help directory for the named package.  If it
+	    # cannot be found it is not an error if the original template
+	    # was ambiguous (did not have a ".").  A package reference is
+	    # always generated for an ambiguous template because we do not
+	    # know what the user wants help for.
+
+	    pk = hd_findmod (hp_sys, Memc[pakname])
+	    if (pk == 0) {
+		if (Memc[modtemp] == EOS)
+		    if (stridxs (".", tlist) != 0) {
+			call eprintf ("Package `%s' not in Help database\n")
+			    call pargstr (Memc[pakname])
+		    }
+		next
+	    }
+
+	    # If module template is null, user wants help on the package
+	    # itself; print help on the package as a module of hp_sys.
+	    # Otherwise expand the module template and print help on each
+	    # module.
+
+	    if (Memc[modtemp] == EOS) {
+		if (print_help (db, hp_sys, Memc[pakname], pk, ctrl) == ERR) {
+		    call eprintf ("No help available for package `%s'\n")
+			call pargstr (Memc[pakname])
+		}
+	    } else {
+		# Get name of package helpdir for package `pk'.
+		if (hd_getname (hp_sys,pk,TY_PKG,Memc[fname],SZ_FNAME) == 0) {
+		    call eprintf ("No help available for package `%s'\n")
+			call pargstr (Memc[pakname])
+		    next
+		}
+
+		# Open the help directory for the package, then open the
+		# template.
+
+		iferr (hp_pak = hdb_load (db, Memc[fname]))
+		    next
+		modlist = ml_open (hp_pak, Memc[modtemp])
+	
+		# Process each module matching the module template.  The QUIT
+		# flag is set if the user responds no to the "more" query
+		# between help blocks or files.
+
+		while (ml_read (modlist, m, Memc[module], SZ_FNAME) != EOF) {
+		    if (print_help (db, hp_pak, Memc[module], m, ctrl) == OK) {
+			no_matches = false
+			if (!matchall)
+			    break
+		    }
+		    if (H_QUIT(ctrl) == YES)
+			break
+		}
+
+		call ml_close (modlist)
+		call hdb_free (db, hp_pak)
+	    }
+
+	    if (H_QUIT(ctrl) == YES)
+		break
+	    else if (!matchall && !no_matches)
+		break
+	}
+
+	# A final check for no matches is necessary for the last module in
+	# the template list.
+
+	if (no_matches && Memc[curmod] != EOS) {
+	    call eprintf ("No help available for `%s'\n")
+		call pargstr (Memc[curmod])
+	}
+
+	call tl_close (tl)
+	call hdb_free (db, hp_sys)
+	call sfree (sp)
+end
diff --git a/pkg/system/help/prblkhdr.x b/pkg/system/help/prblkhdr.x
new file mode 100644
index 00000000..168965ab
--- /dev/null
+++ b/pkg/system/help/prblkhdr.x
@@ -0,0 +1,80 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"help.h"
+
+# PR_BLOCK_HEADER -- Print the help block header.  Clear screen first if
+# enabled.  Print header in the form
+#
+#		 MODNAME (section)      title      MODNAME (section) 
+#
+# followed by a blank line.
+
+procedure pr_block_header (hb, modname, ctrl)
+
+pointer	hb		# encode help block header
+char	modname[ARB]	# module name
+pointer	ctrl		# help control block
+
+char	blank
+int	n, center, offset, lmargin, rmargin
+pointer	sp, lbuf, edge, op, hbuf
+int	strlen(), gstrcpy()
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+	call salloc (edge, SZ_LINE, TY_CHAR)
+	call salloc (hbuf, SZ_LINE, TY_CHAR)
+
+	# Clear screen.
+	if (H_FORMAT(ctrl) == HF_TEXT) {
+	    if (H_SOFLAG(ctrl) == YES)
+	        call houtput (ctrl, "\f")
+	} else {
+	    call sfree (sp)
+	    return
+	}
+
+	lmargin = H_LMARGIN(ctrl)
+	rmargin = min (H_RMARGIN(ctrl), SZ_LINE-1)
+
+	# Initialize the output line to blanks.
+	blank = ' '
+	call amovkc (blank, Memc[lbuf], SZ_LINE)
+
+	n = strlen (HB_TITLE(hb))
+	center = (lmargin + rmargin) / 2
+	offset = max (lmargin, center - n/2)
+
+	# Center help block title in output line.
+	call amovc (HB_TITLE(hb), Memc[lbuf+offset], min(n,rmargin-offset))
+
+	# Format the MODNAME (section) into the "edge" buffer.  MODNAME is the
+	# module name (one of the keys), transformed to upper case.
+
+	if (HB_NKEYS(hb) >= 1) {
+	    op = edge + gstrcpy (modname, Memc[edge], SZ_LINE)
+	    call strupr (Memc[edge])
+	} else {
+	    Memc[edge] = EOS
+	    op = edge
+	}
+
+	if (HB_SECTION(hb) != EOS) {
+	    call sprintf (Memc[op], SZ_LINE-SZ_KEY, " (%s)")
+		call pargstr (HB_SECTION(hb))
+	}
+
+	n = strlen (Memc[edge])
+	call amovc (Memc[edge], Memc[lbuf+lmargin-1], n)
+	call amovc (Memc[edge], Memc[lbuf + rmargin - n], n)
+
+	Memc[lbuf+rmargin] = '\n'
+	Memc[lbuf+rmargin+1] = EOS
+
+	# Write out the header line, followed by a blank line.
+	call houtput (ctrl, Memc[lbuf])
+	call houtput (ctrl, "\n")
+
+	call sfree (sp)
+end
diff --git a/pkg/system/help/prdir.x b/pkg/system/help/prdir.x
new file mode 100644
index 00000000..08322522
--- /dev/null
+++ b/pkg/system/help/prdir.x
@@ -0,0 +1,108 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ttyset.h>
+include	<error.h>
+include	"help.h"
+
+# PR_DIRECTORY -- Print a directory of the help blocks for a package.
+# Extract the names of all the modules therein into a string buffer
+# and print it as a table.
+
+define	SZ_OBUF		2048		# buffer for directory
+define	MAX_MODULES	500		# max modules that can be sorted
+define	MAX_NAMELEN	20		# max chars in a module name
+define	FIRST_COL	6
+define	exit_		90
+
+
+procedure pr_directory (db, hp, pakname, paknum, ctrl)
+
+pointer	db			# help database descriptor
+pointer	hp			# root help directory
+char	pakname[ARB]		# name of package
+int	paknum			# package number in root directory
+pointer	ctrl			# help control params
+
+bool	multiple_directories
+int	m, fd
+pointer	sp, indices, hdr, fname, hp_pak, lbuf, obuf
+
+int	hd_getname(), ttystati(), stropen(), getline()
+pointer	hdb_load()
+errchk	hd_getname, ttystati
+
+begin
+	call smark (sp)
+	call salloc (indices, MAX_MODULES, TY_POINTER)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+	call salloc (hdr, SZ_FNAME, TY_CHAR)
+	call salloc (obuf, SZ_OBUF, TY_CHAR)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+
+	# Get the filename of the help directory for the package.
+	if (hd_getname (hp,paknum,TY_PKG,Memc[fname],SZ_FNAME) == 0) {
+	    call eprintf ("No help available for package `%s'\n")
+		call pargstr (pakname)
+	    goto exit_
+	}
+
+	# Open the help directory for the package.
+	iferr (hp_pak = hdb_load (db, Memc[fname])) {
+	    call sfree (sp)
+	    call erract (EA_WARN)
+	    goto exit_
+	}
+
+	# Extract the names of the modules in the package.  Save the pointers
+	# in an array for the table print routine.
+
+	for (m=0;  m < MAX_MODULES;  m=m+1) {
+	    call salloc (Memi[indices+m], MAX_NAMELEN, TY_CHAR)
+	    if (hd_getname (hp_pak, m+1, TY_MODNAME, Memc[Memi[indices+m]],
+		MAX_NAMELEN) == 0)
+		    break
+	}
+
+	# If the template supplied by the user specified more than a single
+	# package, print the name of the package.
+
+	multiple_directories = (H_LENTL(ctrl) > 1)
+
+	# We may have been called with any legal abbreviation of the package
+	# name; fetch the full name from the system package directory.
+
+	if (multiple_directories) {
+	    call houtput (ctrl, "\n")
+	    if (hd_getname (hp, paknum, TY_MODNAME, Memc[hdr], SZ_FNAME) != 0) {
+		call strcat (":\n", Memc[hdr], SZ_FNAME)
+		call houtput (ctrl, Memc[hdr])
+	    }
+	}
+
+	# Now print the table.  It is not necessary to sort the table,
+	# because the "helpdir" code (which reads the help directory) has
+	# already done so.  The directory is written into a string buffer
+	# and then output line by line with houtput, since there is no
+	# easy way to make STRTBL call houtput.
+	
+	# Format the table into the buffer.  A blank line marks the end of
+	# the table.
+
+	fd = stropen (Memc[obuf], SZ_OBUF, NEW_FILE)
+	call strtbl (fd, Memc, Memi[indices], m, FIRST_COL,
+	    ttystati (H_TTY(ctrl), TTY_NCOLS), MAX_NAMELEN, 0)
+	call putline (fd, "\n")
+	call close (fd)
+
+	# Copy lines from the buffer to the HELP output.  Do not read off
+	# end of buffer or FIO will file fault and abort.
+
+	fd = stropen (Memc[obuf], SZ_OBUF, READ_ONLY)
+	while (getline (fd, Memc[lbuf]) > 1)
+	    call houtput (ctrl, Memc[lbuf])
+	call close (fd)
+
+	call hdb_free (db, hp_pak)
+exit_
+	call sfree (sp)
+end
diff --git a/pkg/system/help/prfile.x b/pkg/system/help/prfile.x
new file mode 100644
index 00000000..8ecf1af1
--- /dev/null
+++ b/pkg/system/help/prfile.x
@@ -0,0 +1,84 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<finfo.h>
+include	<time.h>
+include <ctype.h>
+include	"help.h"
+
+# PR_FILE -- Print a file.  Called to print menu files and source files.
+# Do not print block header, but do page output if enabled by user.
+
+procedure pr_file (fname, ctrl, pakname)
+
+char	fname[ARB]
+pointer	ctrl
+char	pakname[ARB]
+
+int	center_col, ip
+long	fi[LEN_FINFO]
+pointer	sp, lbuf, time
+int	open(), hinput(), strlen(), finfo()
+errchk	houtput, open, hinput
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+	call salloc (time, SZ_TIME, TY_CHAR)
+
+	H_EOF(ctrl) = NO
+	H_RAWIN(ctrl) = YES
+	center_col = (H_LMARGIN(ctrl) + H_RMARGIN(ctrl)) / 2
+
+	# Clear screen and print filename if printing source file.
+	# Filename is printed centered between left and right margins.
+	# Note header must be two lines for man_output().
+
+	if (H_OPTION(ctrl) == O_SOURCE) {
+	    if (finfo (fname, fi) == ERR) {
+		call sprintf (Memc[lbuf], SZ_LINE, "%s `%s'")
+		    call pargstr ("Cannot get file info for file")
+		    call pargstr (fname)
+		call error (20, Memc[lbuf])
+	    }
+	    call cnvtime (FI_MTIME(fi), Memc[time], SZ_TIME)
+
+	    # Header format: "FILE  date  FILE".
+	    call sprintf (Memc[lbuf], SZ_LINE, "%*t%s %*t%s %*t%s\n")
+		call pargi   (H_LMARGIN(ctrl))
+		call pargstr (fname)
+		call pargi   (center_col - strlen(Memc[time]) / 2)
+		call pargstr (Memc[time])
+		call pargi   (H_RMARGIN(ctrl) - strlen (fname) + 1)
+		call pargstr (fname)
+
+	    call houtput (ctrl, "\f")
+	    call houtput (ctrl, Memc[lbuf])
+	    call houtput (ctrl, "\n")
+	}
+
+	# Use hinput/houtput to read and print file so that output is
+	# paginated.
+	H_IN(ctrl) = open (fname, READ_ONLY, TEXT_FILE)
+
+	while (hinput (ctrl, Memc[lbuf]) != EOF) {
+	    if (H_OPTION(ctrl) == O_REFERENCES) {
+		# Replace the newline character.
+		ip = strlen (Memc[lbuf]) - 1
+		Memc[lbuf+ip] = ' '
+
+		# Append the package name.
+		call strcat (" [", Memc[lbuf], SZ_LINE)
+		call strcat (pakname, Memc[lbuf], SZ_LINE)
+		call strcat ("]\n", Memc[lbuf], SZ_LINE)
+
+		# Strip leading whitespace.
+		for (ip=0; IS_WHITE(Memc[lbuf+ip]); ip=ip+1)
+		    ;
+	        call houtput (ctrl, Memc[lbuf+ip])
+	    } else
+	        call houtput (ctrl, Memc[lbuf])
+	}
+
+	call close (H_IN(ctrl))
+	call sfree (sp)
+end
diff --git a/pkg/system/help/prfnames.x b/pkg/system/help/prfnames.x
new file mode 100644
index 00000000..f4e42205
--- /dev/null
+++ b/pkg/system/help/prfnames.x
@@ -0,0 +1,69 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"help.h"
+include	"helpdir.h"
+
+# PR_FILENAMES -- Print file names associated with module.
+
+define	NFILES	5
+
+procedure pr_filenames (hp, module, modnum, ctrl)
+
+pointer	hp
+char	module[ARB]
+int	modnum
+pointer	ctrl
+
+int	i, nfiles, ftype[NFILES]
+pointer	sp, lbuf, pakname, modname, fname[NFILES]
+int	hd_getname()
+errchk	hd_getname, houtput
+data	ftype /TY_HLP, TY_SYS, TY_SRC, TY_PKG, TY_MEN/
+string	fcode "hlpsyssrcpkgmen"
+
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+	call salloc (pakname, SZ_FNAME, TY_CHAR)
+	call salloc (modname, SZ_FNAME, TY_CHAR)
+
+	do i = 1, NFILES
+	    call salloc (fname[i], SZ_PATHNAME, TY_CHAR)
+
+	nfiles = 0
+	do i = 1, NFILES
+	    if (hd_getname (hp, modnum, ftype[i], Memc[fname[i]],
+	    SZ_PATHNAME) > 0)
+		nfiles = nfiles + 1
+
+	if (nfiles > 0) {
+	    # Get the package and module names.  If there is no package name
+	    # in the help directory, it is root help directory, and the module
+	    # name is actually the name of a package.
+
+	    if (HD_PAKNAME(hp) == 0) {
+		call strcpy (module, Memc[pakname], SZ_FNAME)
+		Memc[modname] = EOS
+	    } else {
+		call strcpy (Memc[HD_SBUF(hp) + HD_PAKNAME(hp)],
+		    Memc[pakname], SZ_FNAME)
+		call strcpy (module, Memc[modname], SZ_FNAME)
+	    }
+
+	    # Print "pakname.modname".
+	    if (H_LENTL(ctrl) > 1)
+		call pr_modname (ctrl, Memc[pakname], Memc[modname])
+
+	    # List the names of the help files.
+	    do i = 1, NFILES
+		if (Memc[fname[i]] != EOS) {
+		    call sprintf (Memc[lbuf], SZ_LINE, "\t%0.3s = %s\n")
+			call pargstr (fcode[(i-1)*3+1])
+			call pargstr (Memc[fname[i]])
+		    call houtput (ctrl, Memc[lbuf])
+		}
+	}
+
+	call sfree (sp)
+end
diff --git a/pkg/system/help/prhelp.x b/pkg/system/help/prhelp.x
new file mode 100644
index 00000000..7fba611d
--- /dev/null
+++ b/pkg/system/help/prhelp.x
@@ -0,0 +1,144 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+include	"help.h"
+include	"helpdir.h"
+
+.help print_help
+.nf __________________________________________________________________________
+PRINT_HELP -- Print help documentation for the named module or parameter.
+We are called with the name of a single module; all fiddling with packages
+and templates has been performed by the time we are called.  The help
+directory is open and contains the names of the files containing the help
+source for the module.  Our main task is to determine what kind of help
+is desired and call the appropriate routine.
+
+Recall that the principal options are
+
+        option                    meaning                  file
+
+	help		print help block for module	   hlp
+	param		print help for single param	   hlp
+	section		print a single section		   hlp
+	files	        print file names		   ...
+	source		print source file		   src
+	sysdoc		print system documentation	   sys
+	alldoc		print all documentation		 hlp,sys
+	summary		print help block titles		 hlp,sys
+.endhelp ______________________________________________________________________
+
+define	exit_		90
+
+
+# PRINT_HELP -- Print the indicated type of help text for the named module.
+# If we are called the module exists in the package, but the type of help
+# information indicated may not exist.  Return OK if help is printed, ERR
+# otherwise.
+
+int procedure print_help (db, hp, module, modnum, ctrl)
+
+pointer	db			# help database descriptor
+pointer	hp			# help directory for package
+char	module[ARB]		# name of module for which help is desired
+int	modnum			# module number within package directory
+pointer	ctrl			# help control parameters
+
+bool	not_found
+int	option
+pointer	sp, fname, pakname, modname
+int	hd_getname()
+
+begin
+	call smark (sp)
+	call salloc (fname, SZ_PATHNAME, TY_CHAR)
+	call salloc (pakname, SZ_FNAME, TY_CHAR)
+	call salloc (modname, SZ_FNAME, TY_CHAR)
+
+	# Handle options which do not access a help file.
+	option = H_OPTION(ctrl)
+	H_STATE(ctrl) = BOF
+	not_found = true
+
+	if (option == O_MENU) {
+	    if (hd_getname (hp, modnum, TY_MEN, Memc[fname], SZ_PATHNAME) == 0)
+		goto exit_
+	    iferr (call pr_file (Memc[fname], ctrl, "test"))
+		call erract (EA_WARN)
+	    else
+		not_found = false
+
+	} else if (option == O_FILES) {
+	    # Print names of all files associated with module.
+	    ifnoerr (call pr_filenames (hp, module, modnum, ctrl))
+		not_found = false
+
+	} else if (option == O_DIR) {
+	    # Print a directory of all help blocks for the named package.
+	    ifnoerr (call pr_directory (db, hp, module, modnum, ctrl))
+		not_found = false
+
+	} else if (option == O_SOURCE) {
+	    # Print source file, if given.
+	    if (hd_getname (hp, modnum, TY_SRC, Memc[fname], SZ_PATHNAME) == 0)
+		goto exit_
+	    iferr (call pr_file (Memc[fname], ctrl, "test"))
+		call erract (EA_WARN)
+	    else
+		not_found = false
+
+	} else if (option == O_SUMMARY) {
+	    # Scan hlp file and print summary of help blocks.
+	    if (hd_getname (hp, modnum, TY_HLP, Memc[fname], SZ_PATHNAME) == 0)
+		goto exit_
+	    iferr (call pr_summary (Memc[fname], ctrl))
+		call erract (EA_WARN)
+	    else
+		not_found = false
+
+	} else {
+	    # Get the package and module names.  If there is no package name
+	    # in the help directory, it is root help directory, and the module
+	    # name is actually the name of a package.
+
+	    if (HD_PAKNAME(hp) == 0) {
+		call strcpy (module, Memc[pakname], SZ_FNAME)
+		Memc[modname] = EOS
+	    } else {
+		call strcpy (Memc[HD_SBUF(hp) + HD_PAKNAME(hp)],
+		    Memc[pakname], SZ_FNAME)
+		call strcpy (module, Memc[modname], SZ_FNAME)
+	    }
+
+	    # Print full helpblock (default action).
+	    if (option == O_HELP || option == O_ALLDOC ||
+		option == O_REFERENCES) {
+
+		if (hd_getname (hp,modnum,TY_HLP,Memc[fname],SZ_PATHNAME) > 0) {
+		    ifnoerr {
+			call pr_helpblock (Memc[fname],
+			    Memc[pakname], Memc[modname], TY_HLP, ctrl)
+		    } then
+			not_found = false
+		}
+	    }
+
+	    # Print system documentation.
+	    if (option == O_SYSDOC || option == O_ALLDOC) {
+		if (hd_getname (hp,modnum,TY_SYS,Memc[fname],SZ_PATHNAME) > 0) {
+		    ifnoerr {
+			call pr_helpblock (Memc[fname],
+			    Memc[pakname], Memc[modname], TY_HLP, ctrl)
+		    } then
+			not_found = false
+		}
+	    }
+	}
+
+exit_
+	call flush (H_OUT(ctrl))
+	call sfree (sp)
+	if (not_found)
+	    return (ERR)
+	else
+	    return (OK)
+end
diff --git a/pkg/system/help/prhlpblk.x b/pkg/system/help/prhlpblk.x
new file mode 100644
index 00000000..833cc09b
--- /dev/null
+++ b/pkg/system/help/prhlpblk.x
@@ -0,0 +1,154 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+include	"help.h"
+
+# PR_HELPBLOCK -- Print a normal helpblock.  Open the help file and search
+# for the named help block.  If found and not printing single section or
+# parameter, print manpage title.  Process helpblock through Lroff.
+
+procedure pr_helpblock (fname, pakname, modname, blktype, ctrl)
+
+char	fname[ARB]
+char	pakname[ARB]
+char	modname[ARB]
+int	blktype
+pointer	ctrl
+
+bool	block_found, help_file, at_eof
+int	fd, lmarg, rmarg, soflag, foflag, i, nblocks
+pointer	sp, hb, block_name, ps
+
+bool	streq()
+int	open(), hb_getnextblk()
+pointer	ps_open()
+extern	hinput(), houtput()
+errchk	hb_getnextblk, pr_block_header
+
+begin
+	call smark (sp)
+	call salloc (hb, LEN_HBSTRUCT, TY_STRUCT)
+	call salloc (block_name, SZ_FNAME, TY_CHAR)
+
+	iferr (fd = open (fname, READ_ONLY, TEXT_FILE)) {
+	    call sfree (sp)
+	    call erract (EA_ERROR)
+	    return
+	}
+
+	# Search through the help file for the named help block, of type
+	# help or system.  Decode the help block header into the HB struct.
+
+	H_IN(ctrl) = fd
+	HB_LINENO(hb) = 0
+	H_EOF(ctrl) = NO
+	lmarg = H_LMARGIN(ctrl)
+	rmarg = H_RMARGIN(ctrl)
+	at_eof = false
+	help_file = false
+	nblocks = 0
+	call strcpy (modname, Memc[block_name], SZ_FNAME)
+
+	# Output standout mode control chars (soflag) only if told to do so.
+	# Only output forms mode control chars if using Manpage format output
+	# filter.
+		
+	soflag = H_SOFLAG(ctrl)
+	foflag = H_MANPAGE(ctrl)
+
+	# Process all help blocks in the file which list the current key
+	# (modname) in their title block, and which are of the desired
+	# type, i.e., hlp or sys.
+
+	while (!at_eof && H_QUIT(ctrl) != YES) {
+	    block_found = false
+	    repeat {
+		at_eof = (hb_getnextblk (hb, ctrl) == EOF)
+		if (!at_eof) {
+		    help_file = true
+
+		    if (HB_TYPE(hb) == blktype || blktype == NULL) {
+			# Search keyword list for the module name.  If no
+			# module name was given, match anything.
+
+			if (modname[1] == EOS) {
+			    block_found = true
+			    nblocks = nblocks + 1
+			} else {
+			    for (i=1;  i <= HB_NKEYS(hb);  i=i+1)
+				if (streq (HB_KEY(hb,i), modname)) {
+				    block_found = true
+				    nblocks = nblocks + 1
+				    break
+				}
+			}
+
+			# If no module name, use first key as help block name.
+			if (modname[1] == EOS)
+			    call strcpy (HB_KEY(hb,1), Memc[block_name],
+				SZ_FNAME)
+		    }
+		}
+	    } until (block_found || at_eof)
+
+	    if (at_eof) {
+		# If cannot find block, something is wrong, because the
+		# help directory said there was a block of the desired
+		# type in this file.  If no help blocks at all were found
+		# in the file, the best thing we can do is to simply print
+		# the file; hopefully it contains unformatted text.  If the
+		# file contains help blocks, but none with the correct key
+		# and type, then we print a warning message.
+
+		if (!help_file) {
+		    call close (fd)
+		    # Print header only if there is more than a single list
+		    # element.
+		    if (H_LENTL(ctrl) > 1 && H_OPTION(ctrl) != O_REFERENCES)
+			call pr_modname (ctrl, pakname, modname)
+		    call pr_file (fname, ctrl, pakname)
+		    call sfree (sp)
+		    return
+		} else if (nblocks == 0) {
+		    call eprintf ("Cannot find help block for `%s' in `%s'\n")
+			call pargstr (modname)
+			call pargstr (fname)
+		    break
+		} else
+		    break
+	    }
+
+	    # Clear the screen and print the block header, if not printing a
+	    # single section or parameter.
+
+	    H_EOF(ctrl) = NO
+	    if (H_SECNAME(ctrl) == EOS && H_PARNAME(ctrl) == EOS)
+		call pr_block_header (hb, Memc[block_name], ctrl)
+
+
+	    # Finally!!  Call Lroff to format the help text.
+	    iferr {
+		if (H_FORMAT(ctrl) == HF_TEXT) {
+		    call lroff (hinput, ctrl, houtput, ctrl, lmarg, rmarg,
+		        soflag, foflag)
+		} else if (H_FORMAT(ctrl) == HF_HTML) {
+		    call lroff2html (H_IN(ctrl), H_OUT(ctrl), Memc[block_name],
+			HB_SECTION(hb), HB_TITLE(hb), H_PARNAME(ctrl),
+			H_SECNAME(ctrl))
+		} else if (H_FORMAT(ctrl) == HF_PS) {
+		    ps = ps_open (H_OUT(ctrl), YES)
+		    call sprintf (Memc[block_name], SZ_LINE, "%s (%s)")
+			call pargstr (Memc[block_name])
+			call pargstr (HB_SECTION(hb))
+		    call ps_header (ps, Memc[block_name], HB_TITLE(hb), 
+			Memc[block_name])
+		    call lroff2ps (H_IN(ctrl), H_OUT(ctrl), ps,
+			H_PARNAME(ctrl), H_SECNAME(ctrl))
+		}
+	    } then
+		call erract (EA_WARN)
+	}
+
+	call close (fd)
+	call sfree (sp)
+end
diff --git a/pkg/system/help/prmodname.x b/pkg/system/help/prmodname.x
new file mode 100644
index 00000000..dbf2b939
--- /dev/null
+++ b/pkg/system/help/prmodname.x
@@ -0,0 +1,35 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"help.h"
+
+# PR_MODNAME -- Print the module name header "pakname.modname".  Omit the
+# package name if it begins with an underscore, unless there is no module
+# name.
+
+procedure pr_modname (ctrl, pakname, modname)
+
+pointer	ctrl			# help control parameters
+char	pakname[ARB]		# package name
+char	modname[ARB]		# module name
+
+pointer	sp, lbuf
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+
+	if (pakname[1] != EOS || modname[1] != EOS) {
+	    call houtput (ctrl, "\n")
+	    if (pakname[1] == '_' && modname[1] != EOS) {
+		call sprintf (Memc[lbuf], SZ_LINE, "%s:\n")
+		    call pargstr (modname)
+	    } else {
+		call sprintf (Memc[lbuf], SZ_LINE, "%s.%s:\n")
+		    call pargstr (pakname)
+		    call pargstr (modname)
+	    }
+	    call houtput (ctrl, Memc[lbuf])
+	}
+
+	call sfree (sp)
+end
diff --git a/pkg/system/help/prsummary.x b/pkg/system/help/prsummary.x
new file mode 100644
index 00000000..d030842d
--- /dev/null
+++ b/pkg/system/help/prsummary.x
@@ -0,0 +1,95 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"help.h"
+
+# PR_SUMMARY -- Print a summary of the headers of all help blocks in a
+# helpfile.  Open the help file and search for the successive blocks.
+# Print nice diagnostics describing the contents of the file. 
+
+define	TABSTOP		50
+
+procedure pr_summary (fname, ctrl)
+
+char	fname[ARB]
+pointer	ctrl
+
+int	fd, i, n, nlines
+pointer	sp, hb, lbuf
+int	open(), gstrcat(), hb_getnextblk(), strlen(), getline(), strmatch()
+
+begin
+	call smark (sp)
+	call salloc (hb, LEN_HBSTRUCT, TY_STRUCT)
+	call salloc (lbuf, SZ_LINE, TY_CHAR)
+
+	iferr (fd = open (fname, READ_ONLY, TEXT_FILE)) {
+	    call eprintf ("Cannot open helpfile `%s'\n")
+		call pargstr (fname)
+	    call sfree (sp)
+	    return
+	}
+	HB_LINENO(hb) = 0
+	H_IN(ctrl) = fd
+
+	# Search through the help file for the named help block, of type
+	# help or system.  Decode the help block header into the HB struct.
+
+	call houtput (ctrl, "\n")
+	call sprintf (Memc[lbuf], SZ_LINE, "==> %s <==\n")
+	    call pargstr (fname)
+	call houtput (ctrl, Memc[lbuf])
+
+	while (H_EOF(ctrl) != EOF && hb_getnextblk (hb, ctrl) != EOF) {
+	    # Describe help block.
+	    call sprintf (Memc[lbuf], SZ_LINE, "%6d ")
+		call pargi (HB_LINENO(hb))
+
+	    # Print keyword names.
+	    if (HB_NKEYS(hb) > 0)
+		n = gstrcat (HB_KEY(hb,1), Memc[lbuf], SZ_LINE)
+	    else
+		n = strlen (Memc[lbuf])
+
+	    for (i=2;  i <= HB_NKEYS(hb);  i=i+1) {
+		n = gstrcat (", ", Memc[lbuf], SZ_LINE)
+		n = gstrcat (HB_KEY(hb,i), Memc[lbuf], SZ_LINE)
+	    }
+
+	    # Advance to tabstop and print section label.
+	    repeat {
+		Memc[lbuf+n] = ' '
+		n = n + 1
+	    } until (n >= TABSTOP)
+	    Memc[lbuf+n] = EOS
+
+	    n = gstrcat (HB_SECTION(hb), Memc[lbuf], SZ_LINE)
+	    if (HB_TYPE(hb) != TY_HLP) {
+		call sprintf (Memc[lbuf+n], SZ_LINE-n, " (%s)")
+		    call pargstr (HB_TYPESTR(hb))
+	    }
+
+	    call strcat ("\n", Memc[lbuf], SZ_LINE)
+	    call houtput (ctrl, Memc[lbuf])
+
+	    # Print block title, if any, on next line.
+	    if (strlen (HB_TITLE(hb)) > 0) {
+		call sprintf (Memc[lbuf], SZ_LINE, "       %s\n")
+		    call pargstr (HB_TITLE(hb))
+		call houtput (ctrl, Memc[lbuf])
+	    }
+
+	    # Count and print number of lines in help block.
+	    for (nlines=0;  getline (fd, Memc[lbuf]) != EOF;  nlines=nlines+1)
+		if (strmatch (Memc[lbuf], "^.{endhelp}") > 0)
+		    break
+	    call sprintf (Memc[lbuf], SZ_LINE, "       block size %d lines\n")
+		call pargi (nlines)
+
+	    call houtput (ctrl, Memc[lbuf])
+	    call houtput (ctrl, "\n")
+	    HB_LINENO(hb) = HB_LINENO(hb) + nlines + 1
+	}
+	    
+	call close (fd)
+	call sfree (sp)
+end
diff --git a/pkg/system/help/t_hdbexamine.x b/pkg/system/help/t_hdbexamine.x
new file mode 100644
index 00000000..8046848e
--- /dev/null
+++ b/pkg/system/help/t_hdbexamine.x
@@ -0,0 +1,35 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<syserr.h>
+include	<error.h>
+include	"help.h"
+
+
+# HDBEXAMINE -- Examine the help database.
+
+procedure t_hdbexamine()
+
+pointer	sp, helpdb
+bool	streq(), clgetb()
+int	envgets()
+string	s_helpdb "helpdb"
+
+begin
+	call smark (sp)
+	call salloc (helpdb, SZ_HELPDB, TY_CHAR)
+
+	# Fetch the name of the help database file to be examined.  If the
+	# name is "helpdb", the filename is taken from the environment.
+
+	call clgstr (s_helpdb, Memc[helpdb], SZ_HELPDB)
+	if (streq (Memc[helpdb], s_helpdb))
+	    if (envgets (s_helpdb, Memc[helpdb], SZ_HELPDB) <= 0)
+		call syserrs (SYS_ENVNF, s_helpdb)
+
+	# Examine the structure and contents of the database, printing the
+	# results on the standard output.
+
+	call hdb_examine (STDOUT, Memc[helpdb], clgetb ("verbose"))
+
+	call sfree (sp)
+end
diff --git a/pkg/system/help/t_help.x b/pkg/system/help/t_help.x
new file mode 100644
index 00000000..2f840fb0
--- /dev/null
+++ b/pkg/system/help/t_help.x
@@ -0,0 +1,290 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<syserr.h>
+include	<finfo.h>
+include	<fset.h>
+include	<error.h>
+include	"help.h"
+
+.help help
+.nf __________________________________________________________________________
+HELP -- Print help documentation for the named modules (tasks, packages,
+or library procedures).  Given the package name and module name, the help
+database is used to locate the file(s) containing the desired help text.
+Documentation is organized by package, and is maintained in the package
+directory.  The Lroff text formatter is used to process help text containing
+formatting directives.  Text is output via the TTY interface, which processes
+standout mode control sequences, and clears the screen between pages.
+If the standard output is redirected, output is via putline, and pagination
+is disabled.  See Help.hlp for detailed documentation.
+.endhelp _____________________________________________________________________
+
+define	SZ_TEMPBUF	SZ_HELPDB
+
+
+# HELP -- The main procedure.  Fetch all parameters, decode the option
+# string, set up the control structure, and finally call process_template
+# to expand the module template and process the help text for each module.
+
+procedure t_help()
+
+int	ncols, nlines, list, out_device
+long	fi[LEN_FINFO], db_ctime
+char	helpdb[SZ_HELPDB], device[SZ_FNAME]
+bool	output_is_not_redirected, file_template
+pointer	sp, ctrl, option, tempbuf, db, tty, fname, tempdev
+
+long	clktime()
+pointer	ttyodes(), hdb_open()
+bool	strne(), streq(), clgetb()
+int	btoi(), stridxs(), finfo(), fntopnb(), fntgfnb(), strdic()
+int	clgeti(), get_option(), fstati(), envgets(), envgeti()
+
+data	tty /NULL/, db_ctime /0/
+string	s_helpdb "helpdb"
+define	forms_ 91
+
+begin
+	call smark (sp)
+	call salloc (ctrl, LEN_CTRLSTRUCT, TY_STRUCT)
+	call salloc (option, SZ_FNAME, TY_CHAR)
+	call salloc (tempbuf, SZ_TEMPBUF, TY_CHAR)
+	call salloc (tempdev, SZ_FNAME, TY_CHAR)
+	call salloc (fname, SZ_PATHNAME, TY_CHAR)
+
+	# If we were called without any arguments, do not query for the
+	# template, just set it to null and help will be given for the
+	# current package.
+
+	call aclri (Memi[ctrl], LEN_CTRLSTRUCT)
+	if (clgeti ("$nargs") == 0) {
+	    H_OPTION(ctrl) = O_MENU
+	    H_TEMPLATE(ctrl) = EOS
+	    H_PARNAME(ctrl) = EOS
+	    H_SECNAME(ctrl) = EOS
+	} else
+	    call clgstr ("template", H_TEMPLATE(ctrl), SZ_LINE)
+
+	# get the output device type.  If the device is 'gui' pass off to
+	# the XHELP code to execute the GUI, otherwise process the device
+	# type once we set up the normal HELP task structure.
+
+	call aclrc (Memc[tempdev], SZ_TEMPBUF)
+	call clgstr ("device", Memc[tempdev], SZ_FNAME)
+        out_device = strdic (Memc[tempdev], device, SZ_FNAME, HF_DEVICES)
+	if (out_device == HF_GUI) {
+	    call xhelp (H_TEMPLATE(ctrl))
+	    call sfree (sp)
+	    return
+	}
+
+	# Determine whether the template is a module name template or a
+	# file matching template.
+
+	file_template = clgetb ("file_template")
+	if (file_template) {
+	    H_OPTION(ctrl) = O_HELP
+	    goto forms_
+	}
+
+	# Fetch the name of the help database, i.e., "helpdb", "helpdir",
+	# or the name of a file.  The first time the process runs we open
+	# and read in the database.  The database remains in memory between
+	# calls to HELP, provided the process does not shutdown, provided
+	# the name of the database to be used does not change, and provided
+	# a new help database is not created.
+
+	call aclrc (Memc[tempbuf], SZ_TEMPBUF)
+	call clgstr (s_helpdb, Memc[tempbuf], SZ_TEMPBUF)
+	if (streq (Memc[tempbuf], s_helpdb))
+	    if (envgets (s_helpdb, Memc[tempbuf], SZ_TEMPBUF) <= 0)
+		call syserrs (SYS_ENVNF, s_helpdb)
+
+	# Check to see if any of the files in the list are newer than the
+	# time of the last hdb_open.
+
+	if (db_ctime > 0) {
+	    list = fntopnb (Memc[tempbuf], YES)
+	    while (fntgfnb (list, Memc[fname], SZ_PATHNAME) != EOF) {
+		if (finfo (Memc[fname], fi) == ERR) {
+		    call eprintf ("Cannot access help database file `%s'\n")
+			call pargstr (Memc[fname])
+		} else if (db != NULL && FI_CTIME(fi) > db_ctime) {
+		    call hdb_close (db)
+		    db = NULL
+		    break
+		}
+	    }
+	    call fntclsb (list)
+	} else
+	    db = NULL
+
+	# Reopen the help database if in-core copy is out of date.
+	if (db == NULL || strne (Memc[tempbuf], helpdb)) {
+	    call strcpy (Memc[tempbuf], helpdb, SZ_HELPDB)
+	    if (db != NULL)
+		call hdb_close (db)
+	    db = hdb_open (helpdb)
+	    db_ctime = clktime (long(0))
+	}
+
+	# Fetch the value of the ALL switch.  This determines whether help
+	# will stop after processing the first module matching the template,
+	# or process all modules in the database which match the template.
+	# Explicit use of a pattern matching character anywhere in the template
+	# enable allmodoules.
+
+	if (stridxs ("*?[],", H_TEMPLATE(ctrl)) > 0)
+	    H_ALLMODULES(ctrl) = YES
+	else
+	    H_ALLMODULES(ctrl) = btoi (clgetb ("all"))
+	
+	# If the FILTER_INPUT flag is set, only part of the input text will be
+	# processed.  Filtering is only done if printing a single section or
+	# parameter.
+
+	H_FILTER_INPUT(ctrl) = NO
+
+	# Determine whether or not text for a single section or parameter
+	# is to be output.  If the value of one of these strings is "all",
+	# all sections or all parameters are to be output.  If the "all"
+	# default is in effect, null the string as a flag to lower level
+	# code that all help text is to be processed.
+
+	if (H_OPTION(ctrl) == NULL) {
+	    call clgstr ("section", H_SECNAME(ctrl), SZ_SECNAME)
+	    if (streq (H_SECNAME(ctrl), "all")) {
+		H_SECNAME(ctrl) = EOS
+		call clgstr ("parameter", H_PARNAME(ctrl), SZ_PARNAME)
+		if (streq (H_PARNAME(ctrl), "all"))
+		    H_PARNAME(ctrl) = EOS
+	    }
+	    if (H_SECNAME(ctrl) != EOS || H_PARNAME(ctrl) != EOS)
+		H_FILTER_INPUT(ctrl) = YES
+	}
+
+	# Fetch and decode option string; abbreviations are permitted.
+	if (H_OPTION(ctrl) != O_MENU) {
+	    call clgstr ("option", Memc[option], SZ_FNAME)
+	    call strlwr (Memc[option])
+	    iferr (H_OPTION(ctrl) = get_option (Memc[option])) {
+		call erract (EA_WARN)
+		H_OPTION(ctrl) = O_HELP
+	    }
+	}
+forms_
+	# Pause between screens of output text only if the standard output
+	# is not redirected, and if enabled by the user.
+
+	H_IN(ctrl) = ERR
+	H_OUT(ctrl) = STDOUT
+	H_NLINES(ctrl) = -1
+	H_STATE(ctrl) = BOF
+	H_EOF(ctrl) = NO
+	H_QUIT(ctrl) = NO
+
+	# If the standard output is not redirected, i.e., if writing to the
+	# terminal, determine whether or not output is to be paginated (pause
+	# between pages).  If output is redirected, the pagination flag
+	# and help option controls whether or not manpage style output is
+	# enabled.  Manpage output formatting is desirable only when formatting
+	# help text or printing named files.
+
+	H_RAWOUT(ctrl) = NO
+	H_MANPAGE(ctrl) = NO
+	H_PAGINATE(ctrl) = NO
+	output_is_not_redirected = (fstati (STDOUT, F_REDIR) == NO)
+
+	if (out_device < HF_GUI) {
+	    if (output_is_not_redirected && out_device == HF_TERMINAL) {
+	        if (clgetb ("page")) {
+		    H_PAGINATE(ctrl) = YES
+		    call xttysize (ncols, nlines)
+	        }
+	    } else {
+	        if (clgetb ("page")) {
+		    switch (H_OPTION(ctrl)) {
+		    case O_HELP, O_SOURCE, O_SYSDOC, O_ALLDOC:
+		        H_MANPAGE(ctrl) = YES
+		        H_NLPP(ctrl) = clgeti ("nlpp")
+		        call man_init()
+		    default:
+		        H_RAWOUT(ctrl) = YES
+		    }
+	        } else
+		    H_RAWOUT(ctrl) = YES
+	    }
+	}
+
+	# Get the current package.  Normally this is specified as the null
+	# string or "AskCL", which causes a runtime query of the CL for the
+	# current package when a help template is expanded.  The reason for
+	# providing this parameter is to allow this query to be disabled
+	# (by setting the value to the name of an actual package such as
+	# system or clpackage) so that, e.g., the help task can be called
+	# as a host level task.
+
+	iferr (call clgstr ("curpack", H_CURPACK(ctrl), SZ_CURPACK))
+	    call strcpy ("AskCL", H_CURPACK(ctrl), SZ_CURPACK)
+
+	# Get output device (normally "terminal"), open TTY descriptor for
+	# the device.  The TTY descriptor is left open between calls to HELP
+	# provided the process does not shutdown and the device name or page
+	# size does not change.
+
+	if (out_device == HF_TERMINAL) {
+	    if (tty == NULL || strne (Memc[tempdev], device)) {
+                call strcpy (Memc[tempdev], device, SZ_FNAME)
+	        if (tty != NULL)
+	            call ttycdes (tty)
+	        tty = ttyodes (device)
+	    }
+	}
+
+	if (out_device == HF_HTML) {
+	    H_FORMAT(ctrl) = HF_HTML
+	    H_RAWOUT(ctrl) = YES
+	    H_SOFLAG(ctrl) = NO
+	    H_MANPAGE(ctrl) = NO
+	    tty = NULL
+	} else if (out_device == HF_PS || out_device == HF_POSTSCRIPT) {
+	    H_FORMAT(ctrl) = HF_PS
+	    H_RAWOUT(ctrl) = YES
+	    H_SOFLAG(ctrl) = NO
+	    H_MANPAGE(ctrl) = NO
+	    tty = NULL
+	} else if (out_device == HF_TEXT) {
+	    H_FORMAT(ctrl) = HF_TEXT
+	    H_SOFLAG(ctrl) = NO
+	} else {
+	    H_FORMAT(ctrl) = HF_TEXT
+	    H_SOFLAG(ctrl) = YES
+	}
+	H_TTY(ctrl) = tty
+
+	# Get left and right margins for output text.  Make sure right margin
+	# does not exceed screen width of output device.
+
+	if (H_FORMAT(ctrl) == HF_TEXT) {
+	    H_LMARGIN(ctrl) = max (1, clgeti ("lmargin"))
+	    H_RMARGIN(ctrl) = max (H_LMARGIN(ctrl) + 1, clgeti ("rmargin"))
+	    if (streq (device, "terminal"))
+	        H_RMARGIN(ctrl) = min (H_RMARGIN(ctrl), envgeti ("ttyncols"))
+	}
+
+	# Initialization is completed, control structure is completed.
+	# Format and output the help text.  If we have a module name template
+	# process the template against the help database, otherwise work
+	# directly out of the named files.
+
+	if (file_template)
+	    call do_file_template (H_TEMPLATE(ctrl), ctrl)
+	else
+	    call do_module_template (db, H_TEMPLATE(ctrl), ctrl)
+
+	# Finish last page and return buffer space if writing manpages.
+	if (H_MANPAGE(ctrl) == YES)
+	    call man_close (H_OUT(ctrl))
+
+	call sfree (sp)
+end
diff --git a/pkg/system/help/t_lroff.x b/pkg/system/help/t_lroff.x
new file mode 100644
index 00000000..9df65d43
--- /dev/null
+++ b/pkg/system/help/t_lroff.x
@@ -0,0 +1,35 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# LROFF -- Text process a file.
+
+procedure t_lroff()
+
+char	fname[SZ_FNAME]
+char	line[SZ_LINE]
+char	format[SZ_FNAME]
+int	fd, lmargin, rmargin
+int	open(), getline(), strmatch(), clgeti()
+extern	getline(), putline()
+
+begin
+	call clgstr ("input_file", fname, SZ_FNAME)
+	fd = open (fname, READ_ONLY, TEXT_FILE)
+
+	lmargin = clgeti ("lmargin")
+	rmargin = clgeti ("rmargin")
+	call clgstr ("format", format, SZ_FNAME)
+
+	while (getline (fd, line) != EOF) {
+	    if (strmatch (line, "^.help") > 0) {
+		if (format[1] == 't')
+		    call lroff (getline, fd, putline, STDOUT, lmargin, rmargin,
+		        YES, NO)
+		else if (format[1] == 'h')
+		    call lroff2html (fd, STDOUT, fname, "", "", "", "")
+		else if (format[1] == 'p')
+		    call lroff2ps (fd, STDOUT, NULL, "", "")
+	    }
+	}
+	
+	call close (fd)
+end
diff --git a/pkg/system/help/t_mkhelpdb.x b/pkg/system/help/t_mkhelpdb.x
new file mode 100644
index 00000000..b4cff7a7
--- /dev/null
+++ b/pkg/system/help/t_mkhelpdb.x
@@ -0,0 +1,76 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<syserr.h>
+include	<error.h>
+
+# MKHELPDB -- Make (compile) the help database.  The root help directory
+# is compiled, along with all sub help directories in the package hierarchy.
+# Each package in the system, including both CL packages and library packages,
+# has an associated help directory file.  Despite the fact that the package
+# structure is hierarchical, package names must be distinct because packages
+# are indexed in the database in a linear structure.
+
+procedure t_mkhelpdb()
+
+bool	verbose
+pointer	sp, helpdir, helpdb, tempname
+int	envgets(), access()
+bool	streq(), clgetb()
+string	s_helpdir "helpdir"
+string	s_helpdb "helpdb"
+
+begin
+	call smark (sp)
+	call salloc (helpdir,  SZ_FNAME, TY_CHAR)
+	call salloc (helpdb,   SZ_FNAME, TY_CHAR)
+	call salloc (tempname, SZ_FNAME, TY_CHAR)
+
+	# Fetch the names of the root help directory file and the new database
+	# file.  If the names are "helpdir" or "helpdb", the filenames are
+	# taken from the environment.
+
+	call clgstr (s_helpdir, Memc[helpdir], SZ_FNAME)
+	if (streq (Memc[helpdir], s_helpdir))
+	    if (envgets (s_helpdir, Memc[helpdir], SZ_FNAME) <= 0)
+		call syserrs (SYS_ENVNF, s_helpdir)
+
+	call clgstr (s_helpdb, Memc[helpdb], SZ_FNAME)
+	if (streq (Memc[helpdb], s_helpdb))
+	    if (envgets (s_helpdb, Memc[helpdb], SZ_FNAME) <= 0)
+		call syserrs (SYS_ENVNF, s_helpdb)
+
+	verbose = clgetb ("verbose")
+	if (verbose) {
+	    call printf ("helpdir = %s\n")
+		call pargstr (Memc[helpdir])
+	    call printf ("helpdb = %s\n")
+		call pargstr (Memc[helpdb])
+	}
+
+	# The database is compiled into a temporary file to protect the current
+	# database in the event the compile is aborted.  When the compile has
+	# successfully completed we will delete the original and replace it
+	# with the new database.  This also has the advantage that the switch
+	# takes very little time, making it less likely that anyone will notice
+	# when the database is updated.
+
+	call mktemp ("tmp$hdb", Memc[tempname], SZ_FNAME)
+
+	# Perform the compilation.
+	call hdb_compile (Memc[helpdir], Memc[tempname], verbose)
+
+	# Now attempt to delete the old database and replace it with the new.
+	# By default, the help database flle is maintained in dev$, a directory
+	# with global write permission, so that anyone can delete the file.
+	# If the old file cannot be deleted we delete the new database and
+	# abort.
+
+	if (access (Memc[helpdb],0,0) == YES)
+	    iferr (call delete (Memc[helpdb])) {
+		call delete (Memc[tempname])
+		call erract (EA_ERROR)
+	    }
+
+	call rename (Memc[tempname], Memc[helpdb])
+	call sfree (sp)
+end
diff --git a/pkg/system/help/tlist.x b/pkg/system/help/tlist.x
new file mode 100644
index 00000000..5bd61ab1
--- /dev/null
+++ b/pkg/system/help/tlist.x
@@ -0,0 +1,406 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+include	"help.h"
+
+.help tlist
+.nf __________________________________________________________________________
+TLIST -- Routines for processing the template list.
+Functions:
+	tl_open	    	Expand the template list into a list
+			    of unambiguous templates.
+	tl_read	    	Fetch next template from the list.
+	tl_close	Close the list.
+
+Do not confuse these routines with those used to expand and read from the
+template for a single package.  Our main function is to pass through the
+template list, adding package names where they were left out by the user.
+If no package name is given all packages in the database are searched.
+.endhelp ______________________________________________________________________
+
+define	SZ_TLSBUF	1024		# local string buffer
+define	INC_SZTLSBUF	512		# increment if overflow
+define	SZ_CURPACK	31		# allocation for curpack name string
+
+define	LEN_TLSTRUCT	7
+define	TL_CTRL		Memi[$1]	# string buffer
+define	TL_SBUF		Memi[$1+1]	# string buffer
+define	TL_SZSBUF	Memi[$1+2]	# size of string buffer
+define	TL_NEXTCH	Memi[$1+3]	# index of next char in sbuf
+define	TL_LISTPTR	Memi[$1+4]	# for fetching list elements
+define	TL_CURPACK	Memi[$1+5]	# offset of name of current package
+define	TL_LISTLEN	Memi[$1+6]	# number of elements in list
+
+
+# TL_OPEN -- Take the template list supplied by the user and produce as
+# output a list wherein each element has an explicit package name.  If an
+# element is a module name or a module matching template, it gets replaced
+# by a set of such templates, one for each package in the database.
+# If a package matching template is given, it is expanded into a list of
+# packages.
+
+pointer procedure tl_open (db, hp, tlist, ctrl)
+
+pointer	db			# database descriptor
+pointer	hp			# system package directory
+char	tlist[ARB]		# template list string
+pointer	ctrl
+
+int	ip, junk
+pointer	sp, op, tl, sbuf, template, pakstr, modstr
+int	tl_fetchelem(), tl_matchpak(), tl_putstr(), tl_getcurpack()
+errchk	tl_getcurpack, tl_ambiguous, tl_matchpak
+define	paknotfound_	91
+
+begin
+	call smark (sp)
+	call salloc (template, SZ_LINE, TY_CHAR)
+	call salloc (pakstr, SZ_FNAME, TY_CHAR)
+	call salloc (modstr, SZ_FNAME, TY_CHAR)
+
+	call calloc (tl, LEN_TLSTRUCT, TY_STRUCT)
+	call malloc (sbuf, SZ_TLSBUF, TY_CHAR)
+
+	TL_CTRL(tl)    = ctrl
+	TL_SBUF(tl)    = sbuf
+	TL_SZSBUF(tl)  = SZ_TLSBUF
+	TL_NEXTCH(tl)  = SZ_CURPACK + 2
+	TL_LISTPTR(tl) = SZ_CURPACK + 2
+	TL_LISTLEN(tl) = 0
+
+	# If null template list, set to "curpack.".  This will cause the
+	# help for the current package to be printed.  Otherwise expand
+	# the template.  Template expansion can produce a very big list.
+
+	for (ip=1;  IS_WHITE (tlist[ip]);  ip=ip+1)
+	    ;
+
+	if (tlist[ip] == EOS) {
+	    # Put current package name into list.
+	    op = template + tl_getcurpack (tl, Memc[template], SZ_LINE)
+	    Memc[op] = '.'
+	    Memc[op+1] = EOS
+	    junk = tl_putstr (tl, Memc[template])
+
+	} else {
+	    # Expand nonnull template list.  Each time through the loop
+	    # processes the next comma delimited template.
+
+	    while (tl_fetchelem (tlist,ip,Memc[pakstr],Memc[modstr]) > 0) {
+		if (Memc[pakstr] == EOS) {
+		    if (H_OPTION(ctrl) == O_DIR) {
+			# User wants to print a help directory, but left out
+			# the "." in the package name.  Assume the package
+			# name is an abbreviation and all all packages which
+			# match to the list.
+
+			if (tl_matchpak (tl, hp, Memc[modstr], "") == 0)
+			    goto paknotfound_
+		    } else
+			call tl_ambiguous (tl, db, hp, Memc[modstr])
+
+		} else if (tl_matchpak (tl,hp,Memc[pakstr],Memc[modstr]) == 0) {
+		    # Explicit package template was given by user, but no
+		    # matches were made to packages installed in the system
+		    # help directory.  Inform user.
+paknotfound_
+		    call eprintf ("Cannot find help for package `%s'\n")
+			call pargstr (Memc[pakstr])
+		}
+	    }
+	}
+
+	# Return unused space in string buffer.  Copy length of list to
+	# the main help structure.
+
+	call realloc (TL_SBUF(tl), TL_NEXTCH(tl), TY_CHAR)
+	TL_SZSBUF(tl) = TL_NEXTCH(tl)
+	H_LENTL(ctrl) = TL_LISTLEN(tl)
+
+	call sfree (sp)
+	return (tl)
+end
+
+
+# TL_CLOSE -- Close a template list opened with TL_OPEN.
+
+procedure tl_close (tl)
+
+pointer	tl
+
+begin
+	call mfree (TL_SBUF(tl), TY_CHAR)
+	call mfree (tl, TY_STRUCT)
+end
+
+
+# TL_READ -- Get the next template from the template list.
+# Returns EOF when end of list is reached, else nchars in template.
+
+int procedure tl_read (tl, outstr, maxch)
+
+pointer	tl
+char	outstr[ARB]
+int	maxch
+int	offset, nchars
+int	gstrcpy()
+
+begin
+	if (TL_LISTPTR(tl) >= TL_NEXTCH(tl))
+	    return (EOF)
+
+	offset = TL_LISTPTR(tl)
+	nchars = gstrcpy (Memc[TL_SBUF(tl)+offset], outstr, maxch)
+	TL_LISTPTR(tl) = offset + nchars + 1
+
+	return (nchars)
+end
+
+
+# TL_MATCHPACK -- Match package name against the system package list,
+# add each package matched to the list.
+
+int procedure tl_matchpak (tl, hp, package, module)
+
+pointer	tl
+pointer	hp
+char	package[ARB]
+char	module[ARB]
+
+int	paklen, junk, ngen, modnum
+pointer	sp, template, op, ml
+int	tl_putstr(), ml_read()
+pointer	ml_open()
+errchk	ml_open, ml_read, tl_putstr
+
+begin
+	call smark (sp)
+	call salloc (template, SZ_LINE, TY_CHAR)
+
+	ngen = 0
+	ml = ml_open (hp, package)
+
+	repeat {
+	    paklen = ml_read (ml, modnum, Memc[template], SZ_LINE)
+	    if (paklen == EOF)
+		break
+	    op = template + paklen
+	    Memc[op] = '.'
+	    call strcpy (module, Memc[op+1], SZ_LINE-paklen-1)
+	    junk = tl_putstr (tl, Memc[template])
+	    ngen = ngen + 1
+	}
+
+	call ml_close (ml)
+	call sfree (sp)
+	return (ngen)
+end
+
+
+# TL_AMBIGUOUS -- Called to expand an ambiguous template, i.e., one which did
+# not include the "." delimiter to denote the package and module fields.
+# We handle this by generating a series of module matching templates, one for
+# each package in the database, starting with the current package.  If the
+# ambiguous pattern contains any pattern matching metacharacters we match only
+# against the modules in the current package.  Otherwise (completely ambiguous
+# template) we match against:
+#
+#	[1] all modules in the current package
+#	[2] all modules in all packages, except the current package
+# 
+# For example, "alpha" might be expanded into the list
+#	curpack.alpha, pak1.alpha, pak2.alpha, ...
+#
+# The search [2] is a depth-first search of all packages in the root.  Since
+# each non-root package is a module of some other package, this search will
+# find all non-root packages as well as modules.  For the root packages to be
+# found they must reference themselves as modules, but NOT as subpackages,
+# otherwise recursion will occur.
+
+procedure tl_ambiguous (tl, db, hp, module)
+
+pointer	tl			# template list descriptor
+pointer	db			# database descriptor
+pointer	hp			# package directory
+char	module[ARB]		# module template
+
+int	paklen, junk, ip, pk
+pointer	sp, curpack, template, op
+bool	streq()
+int	tl_putstr(), stridx(), tl_getcurpack(), hd_getname()
+errchk	tl_putstr, tl_getcurpack, hd_getname
+
+begin
+	call smark (sp)
+	call salloc (template, SZ_LINE, TY_CHAR)
+	call salloc (curpack, SZ_FNAME, TY_CHAR)
+
+	# Output the template for the current package.  Save the name of
+	# the current package for later use.
+
+	op = template + tl_getcurpack (tl, Memc[template], SZ_LINE)
+	call strcpy (Memc[template], Memc[curpack], SZ_FNAME)
+	Memc[op] = '.'
+	call strcpy (module, Memc[op+1], SZ_LINE-(op-template)-1)
+	junk = tl_putstr (tl, Memc[template])
+
+	# Check for pattern matching metacharacters.  If present we match
+	# only against the current package, hence we are all done.
+
+	for (ip=1;  module[ip] != EOS;  ip=ip+1)
+	    if (stridx (module[ip], "*?[]") > 0) {
+		call sfree (sp)
+		return
+	    }
+
+	# Output a template for each remaining package in the database,
+	# excluding the current package which gets searched first and has
+	# already been put in the list.  Template is "pakname.module".
+
+	pk = 1
+	repeat {
+	    paklen = hd_getname (hp, pk, TY_MODNAME, Memc[template], SZ_LINE)
+	    pk = pk + 1
+	    if (paklen <= 0)
+		break
+	    if (streq (Memc[template], Memc[curpack]))
+		next
+	    op = template + paklen
+	    Memc[op] = '.'
+	    call strcpy (module, Memc[op+1], SZ_LINE-paklen-1)
+	    junk = tl_putstr (tl, Memc[template])
+	}
+
+	call sfree (sp)
+end
+
+
+# TL_FETCHELEM -- Extract next element from template list.  An element
+# consists of the package pattern, module pattern, both, or either.
+# Set output strings to the null string if field not present.  Return
+# the number of chars in the full element, zero when EOS is reached on
+# the template list.
+
+int procedure tl_fetchelem (tlist, ip, pakstr, modstr)
+
+char	tlist[ARB]
+int	ip
+char	pakstr[SZ_FNAME]
+char	modstr[SZ_FNAME]
+
+bool	package_seen
+char	ch
+int	op, ip_save
+
+begin
+	package_seen = false
+	pakstr[1] = EOS
+
+	# Skip whitespace, the comma delimiters between list elements,
+	# and any null list elements (i.e. ",,").
+
+	while (IS_WHITE (tlist[ip]) || tlist[ip] == ',')
+	    ip = ip + 1
+	ip_save = ip
+
+	# Extract the first field.  If period is seen, this is the package
+	# field, copy it to pakstr.  Remainder is the module name.
+	op = 1
+	ch = tlist[ip]
+	while (ch != ',' && ch != '\n' && ch != EOS) {
+	    modstr[op] = ch
+	    if (ch == '.') {
+		if (package_seen)
+		    call error (18, "Too many '.' delims in help template")
+		package_seen = true
+		modstr[op] = EOS
+		call strcpy (modstr, pakstr, SZ_FNAME)
+		op = 1
+	    } else
+		op = op + 1
+	    ip = ip + 1
+	    ch = tlist[ip]
+	}
+	modstr[op] = EOS
+
+	return (ip - ip_save)
+end
+
+
+# TL_GETCURPACK -- Get the name of the current package.  The hidden CL command
+# "_curpack" prints the name of the current package.
+
+int procedure tl_getcurpack (tl, pakname, maxch)
+
+pointer	tl			# template list
+char	pakname[maxch]		# package name (output)
+int	maxch
+
+pointer	ctrl
+int	nchars, junk
+int	gstrcpy(), fscan()
+bool	strne()
+errchk	clcmd
+
+begin
+	# The current package is read only once for each tlist expansion.
+	# The following uses the CL global parameter "list" to read the
+	# package list.  The CL interface is violated by sending an explicit
+	# command to the CL.  Poor practice, but the query can be disabled
+	# by setting the help.curpack parameter, if this causes a problem.
+
+	if (TL_CURPACK(tl) == NULL) {
+	    # Was the current package set via a help task parameter?
+	    ctrl = TL_CTRL(tl)
+	    if (H_CURPACK(ctrl) != EOS)
+		if (strne (H_CURPACK(ctrl), "AskCL")) {
+		    call strcpy (H_CURPACK(ctrl),
+			Memc[TL_SBUF(tl)+1], SZ_CURPACK)
+		    TL_CURPACK(tl) = 1
+		}
+
+	    if (TL_CURPACK(tl) == NULL) {
+		# Send "_curpack" command to the CL, read the response (a single
+		# line) from CLIN.
+
+		call clcmd ("_curpack")
+		junk = fscan (CLIN)
+		    call gargwrd (Memc[TL_SBUF(tl)+1], SZ_CURPACK)
+		TL_CURPACK(tl) = 1
+	    }
+	}
+
+	nchars = gstrcpy (Memc[TL_SBUF(tl) + TL_CURPACK(tl)], pakname, maxch)
+	return (nchars)
+end
+
+
+# TL_PUTSTR -- Put a string (incl EOS) in the string buffer at nextch.
+# If there is not enough space in the buffer, reallocate a larger buffer.
+# Return the index of the string in the string buffer.
+
+int procedure tl_putstr (tl, str)
+
+pointer	tl
+char	str[ARB]
+int	nextch, nchars, strlen()
+errchk	realloc
+
+begin
+	# Null strings are not stored and cause a null index to be returned.
+	nchars = strlen (str)
+	if (nchars == 0)
+	    return (0)
+
+	nextch = TL_NEXTCH(tl)
+	if (nextch + nchars + 1 > TL_SZSBUF(tl)) {
+	    TL_SZSBUF(tl) = TL_SZSBUF(tl) + INC_SZTLSBUF
+	    call realloc (TL_SBUF(tl), TL_SZSBUF(tl), TY_CHAR)
+	}
+
+	call strcpy (str, Memc[TL_SBUF(tl) + nextch], ARB)
+	TL_NEXTCH(tl) = nextch + nchars + 1
+	TL_LISTLEN(tl) = TL_LISTLEN(tl) + 1
+
+	return (nextch)
+end
diff --git a/pkg/system/help/xhelp/help.gui b/pkg/system/help/xhelp/help.gui
new file mode 100644
index 00000000..b3f21c07
--- /dev/null
+++ b/pkg/system/help/xhelp/help.gui
@@ -0,0 +1,3027 @@
+# XHELP.GUI -- Graphics user interface for the IRAF help browser.
+
+reset-server
+appInitialize xhelp XHelp {
+
+XHelp*objects:\
+	toplevel        Layout          helpLayout\
+	helpLayout	Group		menubarGroup\
+	helpLayout	Paned		helpPanes\
+	menubarGroup	Layout		menubarLayout\
+	menubarLayout	MenuButton	fileButton\
+	menubarLayout	MenuButton	optionsButton\
+	menubarLayout	Command		printButton\
+	menubarLayout	Command		findButton\
+	menubarLayout	Command		searchButton\
+	menubarLayout	MenuButton	historyButton\
+	menubarLayout	Command		helpButton\
+	menubarLayout	Command		quitButton\
+\
+	helpPanes	Group		topicGroup\
+	topicGroup	Layout		topicLayout\
+	topicLayout	Label		topicLabel\
+	topicLayout	Command		reloadButton\
+	topicLayout	Frame		topicFrame\
+	topicFrame	AsciiText	topicEntry\
+	topicLayout	Command		topicClear\
+	topicLayout	Frame		listFrame\
+	listFrame	Viewport	listView\
+	listView	List		topicList\
+\
+	helpPanes	Group		outputGroup\
+	outputGroup	Layout		outputLayout\
+	outputLayout	Command		htbButton\
+	outputLayout	Command		htfButton\
+	outputLayout	Command		htuButton\
+	outputLayout	Command		hthButton\
+	outputLayout	MenuButton	secButton\
+	outputLayout	MenuButton	parButton\
+        outputLayout    TextToggle      hlpOpt\
+        outputLayout    TextToggle      srcOpt\
+        outputLayout    TextToggle      sysOpt\
+        outputLayout    TextToggle      filOpt\
+	outputLayout	Frame		helpFrame\
+	helpFrame	HTML		helpText\
+\
+\
+        toplevel        TopLevelShell   printShell\
+        printShell      Layout          prntLayout\
+        prntLayout      Group           printGroup\
+        prntLayout      Group           printCmdGroup\
+\
+        printGroup      Layout          printLayout\
+        printLayout     Label           toLabel\
+        printLayout     TextToggle      toPrinter\
+        printLayout     TextToggle      toFile\
+        printLayout     Label           printLabel\
+        printLayout     Frame           printFrame\
+        printFrame      AsciiText       printEntry\
+        printLayout     Label           pageLabel\
+        printLayout     TextToggle      pageLetter\
+        printLayout     TextToggle      pageLegal\
+        printLayout     TextToggle      pageA4\
+        printLayout     TextToggle      pageB5\
+\
+        printCmdGroup   Layout          printCmdLayout\
+        printCmdLayout  Command         printOkay\
+        printCmdLayout  Command         printDismiss\
+\
+\
+        toplevel        TopLevelShell   searchShell\
+        searchShell     Group           searchGroup\
+        searchGroup     Layout          searchLayout\
+        searchLayout    Label           resLabel\
+	searchLayout	Frame		resFrame\
+	resFrame	HTML		resList\
+        searchLayout    Label           searchLabel\
+        searchLayout    Frame           searchFrame\
+        searchFrame     AsciiText       searchEntry\
+        searchLayout    Command         searchClear\
+        searchLayout    Command         searchOkay\
+        searchLayout    Label           searchStatus\
+        searchLayout    TextToggle      exactMatch\
+        searchLayout    Command         searchHelp\
+        searchLayout    Command         searchDismiss\
+\
+\
+        toplevel        TransientShell  findShell\
+        findShell       Layout          fsLayout\
+        fsLayout        Group           findGroup\
+        fsLayout        Group           findCmdGroup\
+\
+        findGroup       Layout          findLayout\
+        findLayout      Label           findLabel\
+        findLayout      Frame           findFrame\
+        findFrame       AsciiText       findEntry\
+        findLayout      TextToggle      findDir\
+        findLayout      TextToggle      findCase\
+\
+        findCmdGroup    Layout          findCmdLayout\
+        findCmdLayout   Command         findOkay\
+        findCmdLayout   Command         findClear\
+        findCmdLayout   Command         findDismiss\
+\
+\
+        toplevel        TopLevelShell   fileShell\
+        fileShell       Layout          flist\
+        flist           Group           flGroup\
+        flGroup         Layout          flFrame\
+        flFrame        	Label           flistLabel\
+        flFrame         Frame           flistFrame\
+        flistFrame      AsciiText       flistText\
+        flFrame        	Label           flpkgLabel\
+        flFrame        	Label           flpkgVal\
+        flFrame         Command         flDismiss\
+\
+\
+        toplevel        TopLevelShell   fileBrowser\
+        fileBrowser     Layout          fbLayout\
+        fbLayout        Group           fnavGroup\
+        fbLayout        Group           fbCmdGroup\
+        fbLayout        Parameter       directory\
+\
+        fnavGroup       Layout          fnavLayout\
+        fnavLayout     	Command         fnavHome\
+        fnavLayout     	Command         fnavUp\
+        fnavLayout     	Command         fnavRoot\
+        fnavLayout     	Command         fnavRescan\
+        fnavLayout     	Label           filterLabel\
+        fnavLayout     	Frame           filterFrame\
+        filterFrame     AsciiText       filterEntry\
+        fnavLayout      Command         filterClear\
+        fnavLayout     	Group      	dirGroup\
+        dirGroup        Frame           dirFrame\
+        dirFrame        Viewport        dirView\
+        dirView         List            dirList\
+        fnavLayout     	Group           fileGroup\
+        fileGroup       Frame           fgFrame\
+        fgFrame         Viewport        fileView\
+        fileView        List            fileList\
+        fnavLayout     	Label           curdirLabel\
+        fnavLayout     	Label           curdirVal\
+        fnavLayout     	Label           fnameLabel\
+        fnavLayout     	Frame           fnameFrame\
+        fnameFrame      AsciiText       fnameEntry\
+        fnavLayout     	Command         fnameClear\
+\
+	fnavLayout      Group           fmtGroup\
+	fmtGroup        Layout          fmtLayout\
+	fmtLayout       Label           fmtLabel\
+	fmtLayout       TextToggle      fmtSrc\
+	fmtLayout       TextToggle      fmtText\
+	fmtLayout       TextToggle      fmtHTML\
+	fmtLayout       TextToggle      fmtPS\
+	fmtLayout       Label           owLabel\
+	fmtLayout       TextToggle      overwrite\
+\
+        fbCmdGroup   	Layout          fbCmdLayout\
+        fbCmdLayout  	Command         fbcOkay\
+        fbCmdLayout  	Command         fbcHelp\
+        fbCmdLayout  	Command         fbcDismiss\
+\
+\
+        toplevel        TopLevelShell   doc_source\
+        doc_source      Layout          srcLayout\
+        srcLayout       Frame           srcMenuFrame\
+        srcMenuFrame    Layout          srcMenuBar\
+        srcMenuBar      Command         srcDismiss\
+        srcLayout       Frame           srcFrame\
+        srcFrame        AsciiText       srcText\
+\
+\
+        toplevel        TopLevelShell   hlpShell\
+        hlpShell        Layout          hlpLayout\
+        hlpLayout       Group           hlpMenuGroup\
+        hlpMenuGroup    Layout          hlpMenu\
+        hlpMenu   	Command         hlpBack\
+        hlpMenu   	Command         hlpForward\
+        hlpMenu   	Command         hlpHome\
+        hlpMenu   	Command         hlpTutorial\
+        hlpMenu   	Command         hlpDismiss\
+        hlpLayout       Frame           hlpTextFrame\
+        hlpTextFrame    HTML            hlpText\
+        hlpLayout       Label           hfLabel\
+        hlpLayout       Frame           hfFrame\
+        hfFrame         AsciiText       hfEntry\
+        hlpLayout       Command         hfFind\
+        hlpLayout       Command         hfClear\
+        hlpLayout       TextToggle      hfDir\
+        hlpLayout       TextToggle      hfCase\
+        hlpShell        Parameter       help\
+\
+\
+        toplevel        TransientShell  warning\
+        warning         Layout          warn\
+        warn            Frame           warnFrame\
+        warnFrame       Layout          WFlayout\
+        WFlayout        Icon            warnIcon\
+        WFlayout        TextBox         warnText\
+        warn            Frame           warnBtnFrame\
+        warnBtnFrame    Command         warnDismiss\
+\
+\
+        toplevel        TopLevelShell   tclShell\
+        tclShell        Layout          tclLayout\
+        tclLayout       Group           tclCmdGroup\
+        tclCmdGroup     Layout          tclCmd\
+        tclCmd          Command         tclClear\
+        tclCmd          Command         tclExecute\
+        tclCmd          Toggle          tclLogging\
+        tclCmd          Command         tclDismiss\
+        tclLayout       Frame           tclFrame\
+        tclFrame        AsciiText       tclEntry\
+\
+\
+        toplevel        Parameter       xhelp\
+        xhelp           Parameter       textout\
+        xhelp           Parameter       alert\
+        xhelp           Parameter       apropos\
+        xhelp           Parameter       pkglist\
+        xhelp           Parameter       helpres\
+        xhelp           Parameter       helpfiles\
+        xhelp           Parameter       printer\
+        xhelp           Parameter       curpack\
+        xhelp           Parameter       curtask\
+        xhelp           Parameter       history\
+        xhelp           Parameter       showtype\
+        xhelp           Parameter       type
+
+
+
+    !-------------------------------------------------------
+    ! Define some global resources for the main	menu panels.
+    !-------------------------------------------------------
+    *shadowWidth:			1
+    *background:			gray75
+    *Arrow.width:			16
+    *Arrow.height:			25
+    *Text*height:			21
+    *Command.height:			21
+    *Command.highlightThickness:	1
+    *Command.internalHeight:     	4
+    *MenuButton.height:			21
+    *MenuButton.highlightThickness:	1
+    *Label.borderWidth:			0
+    *Label.shadowWidth:			0
+    *TextButton.shadowWidth:		0
+    *TextButton.highlightThickness:	1
+    *TextToggle*borderWidth:		0
+    *TextToggle.highlightThickness:	0
+    *Toggle.highlightThickness:		1
+    *Group.shrinkToFit:			True
+
+    *Arrow.foreground:			gray72
+    *Arrow.background:			gray72
+    *Text*background:			gray72
+    *AsciiText*background:		gray72
+    *TextBox.background:		gray72
+    *MultiList*background:		gray72
+    *List*background:			gray72
+    *Slider2d*thumbColor:		gray75
+
+
+    !-------------------------------------------------------------
+    ! Define resources to take advantage of the	3D scrollbar look.
+    !-------------------------------------------------------------
+    *Scrollbar*background:		gray75
+    *Scrollbar*width:			15
+    *Scrollbar*height:			15
+    *Scrollbar*shadowWidth:		1
+    *Scrollbar*cursorName:		top_left_arrow
+    *Scrollbar*pushThumb:		true
+
+
+    !----------------------------------------
+    ! Menu resources giving a shadow effect.
+    !----------------------------------------
+    *SmeBSB.leftMargin:			10
+    *SmeBSB.rightMargin:		5
+    *SmeBSB.shadowWidth:		2
+    *SmeBSB*background:			SteelBlue
+    *SimpleMenu*background:		gray77
+    *SimpleMenu.borderWidth:		2
+    *SimpleMenu.borderColor:		Black
+    *SimpleMenu.shadowWidth:		2
+    *SimpleMenu.line1.foreground:	gray51
+    *SimpleMenu.line2.foreground:	gray91
+    *SimpleMenu.line3.foreground:	gray51
+    *SimpleMenu.line4.foreground:	gray91
+    *SimpleMenu.line5.foreground:	gray51
+    *SimpleMenu.line6.foreground:	gray91
+    *SimpleMenu.line7.foreground:	gray51
+    *SimpleMenu.line8.foreground:	gray91
+    *SimpleMenu.line9.foreground:	gray51
+    *SimpleMenu.line10.foreground:	gray91
+    *SimpleMenu.line11.foreground:	gray51
+    *SimpleMenu.line12.foreground:	gray91
+    *SimpleMenu.line13.foreground:	gray51
+    *SimpleMenu.line14.foreground:	gray91
+    *SimpleMenu.line15.foreground:	gray51
+    *SimpleMenu.line16.foreground:	gray91
+    *SimpleMenu.line17.foreground:	gray51
+    *SimpleMenu.line18.foreground:	gray91
+    *SimpleMenu.line19.foreground:	gray51
+    *SimpleMenu.line20.foreground:	gray91
+    *SimpleMenu.line21.foreground:	gray51
+    *SimpleMenu.line22.foreground:	gray91
+    *SimpleMenu.line23.foreground:	gray51
+    *SimpleMenu.line24.foreground:	gray91
+    *SimpleMenu.line25.foreground:	gray51
+    *SimpleMenu.line26.foreground:	gray91
+    *SimpleMenu.line27.foreground:	gray51
+    *SimpleMenu.line28.foreground:	gray91
+    *SimpleMenu.line29.foreground:	gray51
+    *SimpleMenu.line30.foreground:	gray91
+
+
+    !----------------------------------------
+    ! Define the default fonts to be used.
+    !----------------------------------------
+    *font: 				-adobe-times-medium-r-normal-*-14-*-*-*
+    *Command.font:                      -adobe-times-bold-i-normal-*-12-*-*-*
+    *MenuButton.font:                   -adobe-times-bold-i-normal-*-12-*-*-*
+    *Toggle.font:                       -adobe-times-bold-i-normal-*-12-*-*-*
+    *Label.font:                        -adobe-times-bold-i-normal-*-12-*-*-*
+    *TextToggle.font:                   -adobe-times-bold-i-normal-*-12-*-*-*
+    *Group*SimpleMenu*font:            	7x13
+    *HTML*SimpleMenu*font:             	7x13
+
+    *HTML.plainFont:                    -*-fixed-medium-r-*-*-13-*-*-*
+    *HTML.listingFont:                  -*-fixed-medium-r-*-*-13-*-*-*
+    *HTML.font: 			-adobe-times-medium-r-normal-*-14-*-*-*
+    *HTML.boldFont:                     -adobe-times-bold-r-normal-*-14-*-*-*
+    *HTML.plainboldFont:                -adobe-times-bold-r-normal-*-14-*-*-*
+    *HTML.fixedboldFont:                -adobe-times-bold-r-normal-*-14-*-*-*
+    *HTML.header1Font:                  -adobe-times-bold-r-normal-*-18-*-*-*
+    *HTML.header2Font:                  -adobe-times-bold-r-normal-*-14-*-*-*
+    *HTML.header3Font:                  -adobe-times-bold-r-normal-*-14-*-*-*
+    *HTML.header4Font:                  -adobe-times-bold-r-normal-*-14-*-*-*
+    *HTML.header5Font: 			-adobe-times-medium-i-normal-*-12-*-*-*
+    *HTML.header6Font:                  -adobe-times-bold-r-normal-*-10-*-*-*
+    *HTML.anchorUnderlines:             0
+    *HTML.visitedAnchorUnderlines:      0
+    *HTML.anchorColor:                  blue
+    *HTML.visitedAnchorColor:           red3
+    *HTML*background:			gray72
+    *HTML*Scrollbar.background:		gray75
+    *HTML*Scrollbar*width:		15
+    *HTML*Scrollbar*height:		15
+
+
+    !---------------------------
+    ! Set the default resources.
+    !---------------------------
+    *xhelp.title:                               IRAF Help Browser V1.0 - DEV
+    *xhelp.geometry:                            +0+0
+    *xhelp.width:                               625
+    *xhelp.minWidth:                            580
+    *xhelp.minheight:                           450
+    *xhelp.height:                              850
+
+    *helpLayout.geometry:			600x700+0+0
+    *helpLayout*borderWidth:			0
+    *helpLayout*Group.shrinkToFit:		True
+    *helpLayout*Frame.frameType:		sunken
+    *helpLayout*Frame.frameWidth:		1
+    *helpLayout*Command.internalHeight:		2
+    *helpLayout*MenuButton.internalHeight:	2
+    *helpLayout*Label*highlightThickness:	0
+    *helpLayout*List*shadeSurplus:		False
+    *helpLayout*List.internalWidth:             10
+    *helpLayout*Viewport.allowVert:             True
+    *helpLayout*Viewport.forceBars:             True
+    *helpLayout.width:				600
+    *helpLayout.layout: vertical { \
+    	menubarGroup < +inf -inf * > \
+	-1 \
+	horizontal { \
+	    -1 \
+	    helpPanes < +inf -inf * +inf -inf > \
+	    -1 \
+	} \
+	-1 \
+    }
+    *helpLayout*SimpleMenu.shadowWidth:		2
+    *helpLayout*SimpleMenu.borderColor:		Black
+    *helpLayout*SimpleMenu.borderWidth:		2
+
+    *menubarGroup.label:
+    *menubarGroup.outerOffset:			0
+    *menubarGroup.innerOffset:			5
+    *menubarGroup.frameType:			raised
+    *menubarGroup.frameWidth:			2
+    *menubarGroup*Command.shadowWidth:		0
+    *menubarGroup*MenuButton.shadowWidth:	0
+    *menubarLayout.layout: horizontal { \
+	fileButton     5 \
+	optionsButton  5 \
+	printButton    5 \
+    	findButton     5 \
+    	searchButton   5 \
+	historyButton  5 \
+    	10 < +inf -inf > \
+    	helpButton     5 \
+    	quitButton       \
+    }
+    *fileButton.label:				File
+    *fileButton.menuName:			fileMenu
+    *optionsButton.label:			Options
+    *optionsButton.menuName:			optsMenu
+    *printButton.label:				Print
+    *findButton.label:				Find
+    *searchButton.label:			Search
+    *historyButton.label:			History
+    *historyButton.menuName:			historyMenu
+    *helpButton.label:				Help
+    *quitButton.label:				Quit
+
+    *topicGroup.label:
+    *topicGroup.outerOffset:			2
+    *topicGroup.innerOffset:			2
+    *topicGroup*frameType:			chiseled
+    *topicGroup.frameWidth:			2
+    *topicGroup*Label.shadowWidth:		0
+    *topicGroup*Label.borderWidth:		1
+    *topicGroup*Label.highlightThickness:	1
+    *topicGroup*Viewport.useRight:		True
+    *topicGroup*Viewport.useBottom:		True
+    *topicLayout.layout: vertical { \
+    	2 < +2 -2 > \
+        horizontal { \
+    	    3 < +3 -3 > \
+    	    topicLabel 2 topicFrame < +inf * > \
+	    2 \
+	    topicClear 5 reloadButton \
+    	    1 \
+        } \
+    	4 \
+        horizontal { \
+	    vertical { \
+    	        -1 \
+    	        listFrame < +inf -inf * +inf -inf > \
+    	        -1 \
+	    } \
+        } \
+    	2 < +2 -2 > \
+    }
+    *reloadButton.label:			Reload
+    *topicLabel.label:				Topic:
+    *topicLabel.justify:			right
+    *topicEntry*width:                         	100
+    *topicEntry*editType:                      	edit
+    *topicEntry*font:                      	7x13
+    *topicEntry*displayCaret:                 	True
+    *topicClear.label:				Clear
+
+    *topicList.font:				7x13
+    *topicList.width:				500
+    *topicList.height:				100
+    *topicList.columnSpacing:			20
+    *topicList.verticalList:			True
+    !*topicList.defaultColumns:			6
+    !*topicList.forceColumns:			True
+
+    *outputGroup.label:
+    *outputGroup.outerOffset:			2
+    *outputGroup.innerOffset:			2
+    *outputGroup*frameType:			chiseled
+    *outputGroup.frameWidth:			2
+    *outputGrout*TextToggle*on:            	0
+    *outputGroup*TextToggle.frameWidth:         2
+    *outputGroup*TextToggle.frameType:          chiseled
+    *outputGroup*TextToggle.location:           0 0 65 25
+    *outputGroup*TextToggle.leftMargin:      	4
+    *outputLayout.layout: vertical { \
+	2 \
+	horizontal { \
+	    5 \
+	    htbButton 2 htfButton 2 htuButton 2 hthButton 2 \
+            10 < +inf -10 > \
+	    secButton 2 parButton     \
+            10 < +inf -10 > \
+	    hlpOpt -2 srcOpt -2 sysOpt 4 filOpt \
+	    5 \
+	} \
+	2 \
+	horizontal { \
+	    2 \
+	    helpFrame < +inf -inf * +inf -inf > \
+	    2 \
+	} \
+    }
+    *htbButton.label:		 		Back
+    *htbButton.sensitive:		 	False
+    *htfButton.label:		 		Forward
+    *htfButton.sensitive:		 	False
+    *htuButton.label:		 		Up
+    *htuButton.sensitive:		 	False
+    *hthButton.label:		 		Home
+    *printButton.label:				Print
+    *secButton.label:				Sections
+    *secButton.menuName:			secMenu
+    *parButton.label:				Parameters
+    *parButton.menuName:			parMenu
+    *hlpOpt*label:            			Help
+    *hlpOpt*on:            			1
+    *hlpOpt*onIcon:            			diamond1s
+    *hlpOpt*offIcon:            		diamond0s
+    *hlpOpt*highlightColor:            		green
+    *srcOpt*label:            			Source
+    *srcOpt*on:            			0
+    *srcOpt*onIcon:            			diamond1s
+    *srcOpt*offIcon:            		diamond0s
+    *srcOpt*highlightColor:            		green
+    *sysOpt*label:            			Sysdoc
+    *sysOpt*on:            			0
+    *sysOpt*onIcon:            			diamond1s
+    *sysOpt*offIcon:            		diamond0s
+    *sysOpt*highlightColor:            		green
+    *filOpt*label:            			Files
+    *filOpt*on:            			0
+    *filOpt*onIcon:            			square1s
+    *filOpt*offIcon:            		square0s
+    *filOpt*highlightColor:            		yellow
+
+    *helpText.width:				650
+    *helpText.height:				620
+    *helpText.anchorUnderlines:                 1
+    *helpText.visitedAnchorUnderlines:          1
+    *helpText.verticalScrollOnRight:            true
+    *helpText.translations:  			\
+            <Btn2Down>:  popup(secMenu)         \n\
+              <Btn2Up>:  popdown(secMenu)       \n\
+            <Btn3Down>:  popup(navMenu)         \n\
+              <Btn3Up>:  popdown(navMenu)       \n
+
+    *helpText*navMenu.foreground:          	Black
+    *helpText*navMenu.background:          	gray75
+    *helpText*secMenu.foreground:          	Black
+    *helpText*secMenu.background:          	gray75
+
+
+    !--------------------------+
+    ! Printer Shell resources. |
+    !--------------------------+
+    *printShell.title:                          Printer Selection
+    *printShell.width:                          300
+    *printShell.height:                         177
+    *printShell.minHeight:                      177
+    *printShell.maxHeight:                      177
+    *printShell*borderWidth:                    0
+    *printShell*Group.frameType:                chiseled
+    *printShell*Group.frameWidth:               2
+    *printShell*Group.innerOffset:              5
+    *printShell*Group.outerOffset:              2
+    *printShell*Command.internalheight:		4
+    *printShell*Text*editType:                  edit
+    *printShell*Text*height:                    25
+    *printShell*TextToggle.frameWidth:          0
+    *printShell*Group.label:
+    *prntLayout.layout: vertical { \
+        printGroup < +inf -inf * > \
+        printCmdGroup < +inf -inf * > \
+    }
+
+    *printLayout*location:            		0 0 70 25
+    *printLayout*offIcon:             		diamond0s
+    *printLayout*onIcon:              		diamond1s
+    *printLayout*highlightColor:      		yellow
+    *printLayout*Label.height:			35
+    *printLayout*Label.justify:			right
+    *printLayout*TextToggle.frameWidth:         0
+    *printLayout*TextToggle.leftMargin:         8
+    *printLayout*TextToggle*highlightColor:     yellow
+    *printLayout*TextToggle*onIcon:             square1s
+    *printLayout*TextToggle*offIcon:            square0s
+    *printLayout*TextToggle*alignment:		left
+    *printLayout.layout: vertical { \
+	0 < +0 >\
+	horizontal { toLabel 10 toPrinter 10 toFile 10 } \
+	5 < +inf -5 > \
+	horizontal { \
+            printLabel 5 printFrame < +inf -inf * > -1 \
+	} \
+	5 < +inf -5 > \
+	horizontal { \
+	    vertical { pageLabel 10 } \
+            12 \
+	    horizontal { \
+	        vertical { pageLetter -3  pageLegal } \
+		10 \
+	        vertical { pageA4     -3  pageB5 } \
+	    } \
+	} \
+	0 < +0 >\
+    }
+    *toLabel.label:                          	Print to:
+    *toPrinter.label:                           Printer
+    *toPrinter.on:                              True
+    *toFile.label:                              File
+    *printLabel.label:                          Printer:
+    *printFrame.frameType:                      sunken
+    *printFrame.frameWidth:			1
+    *printEntry*string:                     	printer
+    *pageLabel.label:				Page Size:
+    *pageLetter.label:				Letter
+    *pageLetter.on:				1
+    *pageLegal.label:				Legal
+    *pageA4.label:				A4
+    *pageB5.label:				B5
+
+    *printCmdLayout.layout: horizontal { \
+	3 \
+        printOkay 20 < +inf -20 > printDismiss \
+	3 \
+    }
+    *printOkay.label:                           Print
+    *printDismiss.label:                        Dismiss
+
+
+    !-------------------------+
+    ! File Browser resources. |
+    !-------------------------+
+    *fileBrowser.width:               		450
+    *fileBrowser.height:               		375
+    *fileBrowser.title:               		Open a New File...
+    *fileBrowser*borderWidth:               	0
+    *fileBrowser*Group.frameType:           	chiseled
+    *fileBrowser*Group.frameWidth:          	2
+    *fileBrowser*Group.innerOffset: 		3
+    *fileBrowser*Group.outerOffset:		3
+    *fileBrowser*Group.label:				
+
+    *fbLayout.layout: vertical { \
+	2 \
+        fnavGroup < +inf -inf * +inf -inf > \
+	-2 \
+	horizontal { \
+	    -5 \
+            fbCmdGroup < +inf -inf * > \
+	    -5 } \
+	-3 \
+    }
+
+    *fnavGroup*Frame.frameType:           	sunken
+    *fnavGroup*Frame.frameWidth:          	1
+    *fnavGroup*Text*editType:			edit
+    *fnavGroup*Text*height:			25
+    *fnavGroup*Text*font:			7x13
+    *fnavGroup*List.verticalList:              	True
+    *fnavGroup*List.defaultColumns:            	1
+    *fnavGroup*List.forceColumns:              	True
+    *fnavGroup*List.font:              		7x13
+    *fnavGroup*Label.justify:			left
+    *fnavGroup*Viewport.allowVert:             	True
+    *fnavGroup*Viewport.allowHoriz:            	False
+    *fnavGroup*Viewport.forceBars:             	True
+    *fnavGroup*Viewport.useRight:              	True
+    *fnavGroup*Group.outerOffset:		7
+    *fnavGroup*Group.innerOffset:		3
+    *fnavLayout.layout: vertical { \
+	5 \
+	vertical { \
+	    -1 \
+	    horizontal { \
+		5 \
+	        fnavHome   < +inf -inf * > 2 \
+	        fnavUp     < +inf -inf * > 2 \
+	        fnavRoot   < +inf -inf * > 2 \
+	        fnavRescan < +inf -inf * >   \
+		10 \
+	        filterLabel 2 filterFrame < +inf -inf * > \
+		2 \
+		filterClear \
+		5 \
+	    } \
+	    3 \
+	} \
+	5 \
+	horizontal { \
+	    -5 \
+	    dirGroup < +inf -inf * +inf - inf  > \
+	    -8 \
+	    fileGroup < +inf -inf * +inf - inf  > \
+	    -5 \
+	} \
+	-3 \
+	horizontal { \
+	    curdirLabel 5 curdirVal < +inf -inf * > 5 } \
+	5 \
+	horizontal { \
+	    fnameLabel 2 fnameFrame < +inf -inf * > 2 fnameClear 5\
+	} \
+	7 \
+	fmtGroup < +inf -inf * > \
+	-3 \
+    }
+    *fileBrowser*fnavGroup*dirGroup.label:	Directories
+    *fileBrowser*fnavGroup*fileGroup.label:	Files
+    *fileBrowser*fnavGroup*dirGroup.innerOffset:	3
+    *fileBrowser*fnavGroup*fileGroup.innerOffset:	3
+    *fileBrowser*fnavGroup*dirGroup.outerOffset:	7
+    *fileBrowser*fnavGroup*fileGroup.outerOffset:	7
+    *fileBrowser*fnavGroup*dirGroup.font:		7x13bold
+    *fileBrowser*fnavGroup*fileGroup.font:		7x13bold
+
+    *filterLabel.label:				Filter
+    *filterClear.label:				Clear
+    *curdirLabel.label:				Directory:
+    *curdirVal.label:				
+    *curdirVal.font:				7x13
+    *fnameLabel.label:				Selection\ \ 
+    *fnameClear.label:				Clear
+    *fnavHome.label:				Home
+    *fnavUp.label:				Up
+    *fnavRoot.label:				Root
+    *fnavRescan.label:				Rescan
+
+    *fmtGroup*Group.outerOffset:                3
+    *fmtGroup*Group.innerOffset:                3
+    *fmtLayout*TextToggle.frameWidth:           0
+    *fmtLayout*TextToggle.leftMargin:           4
+    *fmtLayout*TextToggle.alignment:		left
+    *fmtLayout*TextToggle*highlightColor:       yellow
+    *fmtLayout*TextToggle*onIcon:               square1s
+    *fmtLayout*TextToggle*offIcon:              square0s
+    *fmtLayout.layout: vertical { \
+	horizontal { 5 fmtLabel 10 fmtSrc 3 fmtText 3 fmtHTML 3 fmtPS 5 } \
+	horizontal { 50 owLabel 10 overwrite 5 < +inf > } \
+    }
+    *fmtLabel.label:                            Save As Format:
+    *fmtSrc.label:                              Source
+    *fmtSrc.on:                                 1
+    *fmtSrc.location:                           0 0 65 22
+    *fmtText.label:                             Text
+    *fmtText.location:                          0 0 65 22
+    *fmtHTML.label:                             HTML
+    *fmtHTML.location:                          0 0 65 22
+    *fmtPS.label:                               PostScript
+    *fmtPS.location:                            0 0 100 22
+    *owLabel.label:                             Options:
+    *overwrite.label:                       Allow overwrite of existing files?
+    *overwrite.location:                        0 0 200 22
+
+    *fbCmdLayout.outerOffset:			0
+    *fbCmdLayout.layout: horizontal { \
+	5 \
+	vertical { 2 fbcOkay 2 } \
+        20 < +inf -20 > \
+	vertical { 2 fbcHelp 2 } \
+	2 \
+	vertical { 2 fbcDismiss 2 } \
+	5 \
+    }
+    *fbcOkay.label:		 		Okay
+    *fbcHelp.label:		 		Help
+    *fbcDismiss.label:		 		Dismiss
+
+
+    !-----------------------+
+    ! Find Shell resources. |
+    !-----------------------+
+    *findShell.title:                           Find within a document...
+    *findShell.width:                           365
+    *findShell.height:                          130
+    *findShell*borderWidth:                     0
+    *findShell*Group.frameType:                 chiseled
+    *findShell*Group.frameWidth:                2
+    *findShell*Group.innerOffset:               5
+    *findShell*Group.outerOffset:               2
+    *findShell*Command.internalheight:		4
+    *findShell*Text*editType:                   edit
+    *findShell*Text*height:                     25
+    *findShell*TextToggle.frameWidth:           0
+    *findShell*Group.label:
+    *fsLayout.layout: vertical { \
+        findGroup    < +inf -inf * > \
+        findCmdGroup < +inf -inf * > \
+    }
+
+    *findLayout*location:            		0 0 120 25
+    *findLayout*offIcon:             		diamond0s
+    *findLayout*onIcon:              		diamond1s
+    *findLayout*highlightColor:      		yellow
+    *findLayout*Label.height:			35
+    *findLayout*Label.justify:			right
+    *findLayout*TextToggle.frameWidth:          0
+    *findLayout*TextToggle.leftMargin:          4
+    *findLayout*TextToggle*highlightColor:      yellow
+    *findLayout*TextToggle*onIcon:              square1s
+    *findLayout*TextToggle*offIcon:             square0s
+    *findLayout.layout: vertical { \
+	5 \
+	horizontal { \
+            findLabel 7 findFrame < +inf -inf * > -1 \
+	} \
+	5 \
+	horizontal { \
+            20 < +inf -20 > \
+            findDir 10 findCase \
+            20 < +inf -20 > \
+	} \
+    }
+    *findLabel.label:                           Find:
+    *findFrame.frameType:                       sunken
+    *findFrame.frameWidth:			1
+    *findEntry*string:                     	
+    *findDir.label:				Find Backwards
+    *findCase.label:				Case Sensitive
+
+    *findCmdLayout.layout: horizontal { \
+	3 \
+        findOkay \
+        20 < +inf -20 > \
+        findClear \
+        20 < +inf -20 > \
+        findDismiss \
+	3 \
+    }
+    *findOkay.label:                            Find
+    *findClear.label:                           Clear
+    *findDismiss.label:                         Dismiss
+
+
+    !-------------------------------------------+
+    ! Set the document source viewer resources. |
+    !-------------------------------------------+
+    *doc_source.title:                        	Page source
+    *doc_source.width:                          575
+    *doc_source.height:                         450
+    *srcLayout*borderWidth:                     0
+    *srcLayout.layout:  vertical { \
+        srcMenuFrame < +inf -inf * > \
+	-2 \
+        srcFrame < +inf -inf * +inf -inf > \
+	-2 \
+    }
+
+    *srcMenuBar.layout: 	horizontal { 50 < +inf -inf > srcDismiss 5 }
+    *srcMenuFrame.height:                       40
+    *srcMenuFrame.outerOffset:                  0
+    *srcMenuFrame.innerOffset:                  5
+    *srcMenuFrame.frameType:                    chiseled
+    *srcMenuFrame.frameWidth:                   2
+    *srcFrame.frameType:                     	sunken
+    *srcFrame.frameWidth:                    	1
+    *srcFrame.outerOffset:                    	5
+    *srcText*scrollVertical:                 	always
+    *srcText*scrollHorizontal:               	always
+    *srcText*Scrollbar.width:			15
+    *srcText*Scrollbar.height:			15
+    *srcText*background:			gray75
+    *srcText*font:				7x13
+    *srcText*editType:                       	read
+    *srcText*displayCaret:			False
+    *srcDismiss.label:                         	Dismiss
+    *srcDismiss.width:                         	150
+
+
+    !-------------------------+
+    ! Search Shell resources. |
+    !-------------------------+
+    *searchShell.title:                         Search for a topic...
+    *searchShell.width:                         600
+    *searchShell.height:                        250
+    *searchShell*borderWidth:                   0
+    *searchShell*Viewport.allowVert:            True
+    *searchShell*Viewport.allowHoriz:           True
+    *searchShell*Viewport.useBottom:            True
+    *searchShell*Viewport.useRight:           	False
+    *searchShell*Viewport.forceBars:            True
+    *searchGroup.frameType:               	chiseled
+    *searchGroup.frameWidth:              	2
+    *searchGroup.innerOffset:             	7
+    *searchGroup.outerOffset:             	7
+    *searchGroup.highlightThickness:		0
+    *searchGroup.label:                             
+    *searchLayout.layout: vertical { \
+	horizontal { \
+            45 < +45 -45 > \
+            resLabel < +inf -inf * > \
+            5 < +inf -inf > \
+            exactMatch    \
+	} \
+	2 < +2 - 2 > \
+        resFrame < +inf -inf * +inf -inf > \
+	5 < +5 - 5 > \
+	horizontal { \
+            searchLabel 5 searchFrame \
+	    5 \
+            searchClear 2 searchOkay    \
+            5 < +inf -inf > \
+            searchStatus    \
+            5 < +inf -inf > \
+            searchHelp  2 searchDismiss \
+	} \
+    }
+    *resLabel.label: Task            Package         Description
+    *resLabel.justify:				left
+    *resFrame.frameType:			sunken
+    *resFrame.frameWidth:			1
+    *resList.font:				7x13
+    *resList.width:				100
+    *resList.height:				100
+    *resList.marginWidth:			5
+    *resList.marginHeight:			5
+
+    *searchLabel.label:                         Topic:
+    *searchFrame.frameType:                     sunken
+    *searchFrame.frameWidth:                    1
+    *searchEntry*font:                      	7x13
+    *searchEntry*displayCaret:                 	True
+    *searchEntry*editType:                      edit
+    *searchEntry*height:                        25
+    *searchEntry*width:                         150
+    *searchClear.label:                         Clear
+    *searchOkay.label:                          Search
+    *searchStatus.label:                        
+    *exactMatch.label:                       	Require Exact Match
+    *exactMatch*on:                             1
+    *exactMatch*onIcon:                         diamond1s
+    *exactMatch*offIcon:                        diamond0s
+    *exactMatch*highlightColor:                 green
+    *exactMatch.frameWidth:         		2
+    *exactMatch.frameType:          		chiseled
+    *exactMatch.location:           		0 0 150 25
+    *exactMatch.leftMargin:         		4
+    *searchHelp.label:                       	Help
+    *searchDismiss.label:                       Dismiss
+
+
+    !----------------
+    !  Help Window.
+    !----------------
+    *hlpShell.title:                           Help
+    *hlpShell.width:                           500
+    *hlpShell.height:                          620
+    *hlpLayout*borderWidth:                    0
+    *hlpLayout*Frame*frameType:                sunken
+    *hlpLayout*Frame*frameWidth:               1
+   
+    *hlpMenuGroup.label:		
+    *hlpMenuGroup.outerOffset:			0
+    *hlpMenuGroup.innerOffset:			0
+    *hlpLayout.layout: vertical { \
+        hlpMenuGroup < +inf -inf * > \
+	-3 \
+        hlpTextFrame < +inf -inf * +inf -inf > \
+	horizontal { \
+	    5 \
+	    hfLabel 5 hfFrame < +inf -inf *> \
+	    2 \
+	    hfFind 2 hfClear 5 hfDir 5 hfCase \
+	    5 \
+	} \
+	2 \
+    }
+    *hlpLayout*TextToggle*location:             0 0 90 25
+    *hlpLayout*TextToggle*offIcon:              diamond0s
+    *hlpLayout*TextToggle*onIcon:               diamond1s
+    *hlpLayout*TextToggle*highlightColor:       yellow
+    *hlpLayout*TextToggle*frameType:		chiseled
+    *hlpLayout*TextToggle*frameWidth:		2
+    *hfEntry*editType:                      	edit
+    *hfEntry*font:                      	7x13
+    *hfEntry*displayCaret:                 	True
+    *hfLabel.label:				Find: 
+    *hfFind.label:				Find
+    *hfClear.label:				Clear
+    *hfDir.label:				Backwards
+    *hfCase.label:				Caseless
+    *hfCase.on:					true
+
+    *hlpMenu*Command.internalHeight:     	4
+    *hlpMenu*Command.highlightThickness: 	1
+    *hlpMenu*Command.height: 			20
+    *hlpMenu.layout: vertical { \
+        5 \
+        horizontal { \
+            5 \
+            hlpBack 2 hlpForward 2 hlpHome 2 hlpTutorial \
+            20 < +inf -20 > \
+            hlpDismiss \
+            5 \
+        } \
+        5 \
+    }
+    *hlpBack.label:                             Back
+    *hlpBack.sensitive:                         False
+    *hlpForward.label:                          Forward
+    *hlpHome.label:                             Home
+    *hlpTutorial.label:                         Tutorial
+    *hlpTutorial.sensitive:                     false
+    *hlpDismiss.label:                          Dismiss
+ 
+    *hlpTextFrame.outerOffset:			2
+    *hlpText.width:                             500
+    *hlpText.height:                            500
+    *hlpText.anchorUnderlines:                  1
+    *hlpText.visitedAnchorUnderlines:           1
+    *hlpText.verticalScrollOnRight:             true
+
+
+    !------------------+
+    ! File List dialog.
+    !------------------+
+    *fileShell.title:                          	Help Files
+    *fileShell.geometry:                        500x165
+    *fileShell*borderWidth:                     0
+    *fileShell*Command.width:                   90
+    *fileShell*Command.height:                  30
+    *fileShell*Frame.frameType:                 sunken
+    *fileShell*Frame.frameWidth:                1
+    *fileShell*Frame.innerOffset:               1
+    *fileShell*Text*font:			7x13
+    *flist.layout: vertical { \
+        1 \
+        horizontal { 1 flGroup < +inf -inf * +inf -inf> 1 } \
+        1 \
+    }
+
+    *flGroup.frameType:               		chiseled
+    *flGroup.frameWidth:              		2
+    *flGroup.innerOffset:             		5
+    *flGroup.outerOffset:             		5
+    *flGroup.label:                             
+    *flFrame.layout: vertical { \
+        5 \
+        horizontal { \
+            13 \
+            flistLabel < +inf -inf * > \
+            5 < +inf -5 > \
+        } \
+        2 \
+        horizontal { 1 flistFrame < +inf -inf * +inf -inf > 1 } \
+        7 \
+        horizontal { \
+            5 \
+            flpkgLabel 2 flpkgVal < +inf -inf * > \
+            5 < +inf -5 > \
+            flDismiss \
+            5 \
+        } \
+    }
+    *flDismiss.label:                           Dismiss
+    *flistLabel.label:    	     	Option       Status        Filename
+    *flistLabel.justify:			left
+    *flpkgLabel.label:    			Task: 
+    *flpkgLabel.justify:			left
+    *flpkgVal.label:				(Undefined)
+    *flpkgVal.justify:				left
+    *flpkgVal*font:				7x13
+    *flistText.label:
+    *flistText.scrollVertical:			Never
+    *flistText.scrollHorizontal:		whenNeeded
+    *flistText*displayCaret:			False
+    *flistText*editType:			edit
+
+
+    !----------------+
+    ! WARNING dialog.
+    !----------------+
+    *warning.geometry:                          +400+300
+    *warning*borderWidth:                       0
+    *warning*TextBox.frameWidth:                0
+    *warning*Command.width:                  	90
+    *warning*Command.height:                 	30
+    *warning*Frame.frameType:                 	sunken
+    *warning*Frame.frameWidth:                	1
+    *warning*Frame.innerOffset:                 3
+    *warn.layout: vertical { \
+        5 \
+        horizontal { \
+            5 \
+            warnFrame < +inf * +inf > \
+            5 \
+        } \
+        5 \
+        horizontal { \
+            5 < +inf -5 > \
+            warnBtnFrame \
+            5 < +inf -5 > \
+        } \
+        5 \
+    }
+
+    *WFlayout.layout: horizontal { \
+        5 \
+        vertical { \
+            5 < +inf -5 > \
+            warnIcon \
+            5 < +inf -5 > \
+        } \
+        5 \
+        warnText < +inf -inf * +inf -inf > \
+        5 \
+    }
+    *warnIcon.location:                         0 0 40 40
+    *warnIcon.image:                            WARNING
+    *warnText.label:                            generic warning text
+    *warnText.width:                            270
+    *warnText.height:                           60
+    *warnText*background:			gray75
+    *warnDismiss.label:                         Dismiss
+
+    !--------------------------
+    ! Define a Debug Tcl shell.
+    !--------------------------
+    *tclShell.width:				550
+    *tclShell.height:				180
+    *tclShell.title:				TCL Command Entry Shell
+    *tclLayout*borderWidth:			0
+    *tclLayout*Frame.frameType:			sunken
+    *tclLayout*Frame.frameWidth:		1
+    *tclLayout.layout:	vertical { \
+        tclCmdGroup   < +inf -inf * > \
+        tclFrame < +inf -inf * +inf -inf> \
+    }
+    *tclEntry*editType:				edit
+    *tclEntry*type:				string
+    *tclEntry*scrollVertical:			Always
+    *tclEntry*scrollHorizontal:			whenNeeded
+
+    *tclCmdGroup.label:		
+    *tclCmdGroup.outerOffset:			0
+    *tclCmdGroup.innerOffset:			0
+    *tclCmd.layout: vertical { \
+        5 \
+        horizontal { \
+            5 \
+            tclClear   3 tclExecute   \
+            10 < +inf -10>    \
+            tclLogging 3 tclDismiss   \
+            5 \
+        } \
+        5 \
+    }
+    *tclClear.label:				Clear
+    *tclExecute.label:				Execute
+    *tclLogging.label:				Enable Logging
+    *tclDismiss.label:				Dismiss
+}
+
+
+
+################################################################################
+
+createObjects
+
+# Define Bitmaps and Pixmaps to be used.
+createBitmap null 16 16 {
+   0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+   0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+   0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}
+
+createBitmap check 16 16 {
+   0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x40, 0x00, 0x60,
+   0x00, 0x30, 0x00, 0x18, 0x00, 0x0c, 0x08, 0x06, 0x18, 0x03, 0xb0, 0x01,
+   0xe0, 0x00, 0x40, 0x00, 0x00, 0x00, 0x00, 0x00}
+
+createBitmap arrow 16 16 {
+   0x00, 0x00, 0x00, 0x00, 0x00, 0x04, 0x00, 0x0c, 0x00, 0x14, 0xf8, 0x27,
+   0x08, 0x40, 0xf8, 0x27, 0x00, 0x14, 0x00, 0x0c, 0x00, 0x04, 0x00, 0x00,
+   0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}
+
+activate
+
+################################################################################
+
+
+################################################################################
+# Global variables. |
+################################################################################
+set version		"IRAF Help GUI V1.0"	;# version string
+set curpack		""			;# current values
+set curtask		""
+set helpType		"package"		;# type of help to get
+set helpOption		"help"			;# help option param
+set fileManaged 	0			;# is fileShell mapped?
+set pkgList		{ }
+set visited(0)		empty
+
+set listOrient		1 			;# options
+set showType		0
+set showFiles		1
+set exactMatch		1
+
+set MAX_MENU_SIZE	40			;# what it says it is
+
+# History array initialization.
+set HPkg(0) 		{Home}			;# package
+set HOpt(0) 		{help}			;# option
+set HTask(0) 		{Home}			;# task
+set HUrl(0) 		{}			;# url
+set HType(0) 		{task}			;# type (task|package|file)
+set HFile(0) 		{}			;# filename
+set htop		0			;# top of array
+set hcurrent		0			;# current page
+
+# Panel mapping flags.
+set search_mapped	0			;# searchShell mapped?
+
+
+################################################################################
+#   Utility Callbacks  
+################################################################################
+
+#  Procedures for sending client cursor commands.
+proc GKey { key args } { send client gkey $key   }
+proc GCmd { args }     { send client gcmd $args  }
+
+#  Procedures to test True/False strings in resources.
+proc true  { v } \
+   { expr { $v=="true"  || $v=="True"  || $v=="TRUE"  || $v==1 || $v=="yes" } }
+proc false { v } \
+   { expr { $v=="false" || $v=="False" || $v=="FALSE" || $v==0 || $v=="no"  } }
+
+#  No-op procedure for text widgets with no callbacks to swallow newline.
+proc noop { args } { }
+
+#  Common functions.
+proc min { a b } { expr {($a < $b) ? $a : $b} }
+proc max { a b } { expr {($a > $b) ? $a : $b} }
+
+
+#--------------------+
+# Debugging options. |
+#--------------------+
+set debug		0			;# debug flag
+send tclLogging set state [min 1 $debug]
+
+
+################################################################################
+# Initialize. |
+################################################################################
+proc Init args {
+    global history curpack curtask debug
+    global secMenuDescription parMenuDescription 
+    global HPkg HTask HUrl HOpt HFile HType
+
+    if {$debug == 1} { send tclShell map }
+
+    # Reinitialize global vars in case of a restart.
+    set curpack	   ""
+    set curtask	   ""
+    set helpType   "package"
+    set helpOption "help"
+
+    # Initialize the entry strings.
+    send printEntry set string ""
+    send topicEntry set string ""
+
+    # Initialize the various lists.
+    send topicList   setList "{ }"    resize
+
+    send secButton set sensitive False
+    send parButton set sensitive False
+    editMenu secMenu secButton $secMenuDescription
+    editMenu parMenu parButton $parMenuDescription
+
+} ; #send server postActivateCallback Init
+
+
+
+# Create the Navigation Menu.
+set navMenuDescription {
+    { "Back       "   f.exec 	{ Back 	  } 
+	      sensitive {([send htbButton isSensitive]==1) ? "true" : "false" }
+    }
+    { "Forward    "   f.exec 	{ Forward }
+	      sensitive {([send htfButton isSensitive]==1) ? "true" : "false" }
+    }
+    { "Up         "   f.exec 	{ Up	  }
+	      sensitive {([send htuButton isSensitive]==1) ? "true" : "false" }
+    }
+    { "Home       "        f.exec    { Home    } }
+    {                      f.dblline      	 }
+    { "Reload     "        f.exec    { Reload  } }
+    { "Open File... "      f.exec    { Open    } }
+    { "Save As...   "      f.exec    { SaveAs  } }
+    { "View Page Source  " f.exec    { srcOpen } }
+    {                      f.dblline             }
+    { "Find...    "        f.exec    { Find    } }
+    { "Search...  "        f.exec    { Search  } }
+    { "Print...   "        f.exec    { Print   } }
+} ; createMenu navMenu helpText $navMenuDescription
+
+# Create the default Section Menu.
+set secMenuDescription {
+    { " Top of Page       " f.exec    { send helpText gotoId 0 } }
+    {                       f.dblline          }
+    { "                   " f.exec    { noop } }
+    { " No Sections Found " f.exec    { noop } }
+    { "                   " f.exec    { noop } }
+} ; createMenu secMenu secButton $secMenuDescription
+
+# Create the default Parameter Menu.
+set parMenuDescription {
+    { " No Parameters Found" f.exec  { noop } }
+} ; createMenu parMenu parButton $parMenuDescription
+
+
+# Initialize.
+Init
+
+
+################################################################################
+# Menubar command callbacks.
+################################################################################
+
+# File Menu
+set fileMenuDescription {
+    { " Open File... "      f.exec {  Open    } }
+    { " Save As...   "      f.exec {  SaveAs  } }
+    { " Print...     "      f.exec {  Print   } }
+    {                       f.dblline      	   }
+    { " Reload...    "      f.exec {  Reload  } }
+    { " View Page Source  " f.exec {  srcOpen } }
+    { " Search...    "      f.exec {  Search  } }
+    { " Find...      "      f.exec {  Find    } }
+    {                       f.dblline       	}
+    { " Help         "      f.exec {  Help 	 } }
+    { " Quit         "      f.exec {  Quit 	 } }
+} ; createMenu fileMenu fileButton $fileMenuDescription
+
+
+# History Menu
+set historyMenuDescription {
+    { " Back         "	f.exec { Back	 }
+	      sensitive {([send htbButton isSensitive]==1) ? "true" : "false" }
+    }
+    { " Forward      "	f.exec { Forward }
+	      sensitive {([send htfButton isSensitive]==1) ? "true" : "false" }
+    }
+    { " Up           "	f.exec { Up  	 }
+	      sensitive {([send htuButton isSensitive]==1) ? "true" : "false" }
+    }
+    { " Home         "	f.exec { Home	 } }
+    {             	f.dblline   	   }
+    { " Clear History"  f.exec { histClear } }
+    {             	f.dblline   	   }
+} ; createMenu historyMenu historyButton $historyMenuDescription
+
+
+# Options Menu
+set optsMenuDescription {
+    { " Show task type          "  f.exec { setOpts showType  } 
+				bitmap {($showType==1)? "check" : "null"} }
+    { " Show missing files      "  f.exec { setOpts showFiles } 
+				bitmap {($showFiles==1)? "check" : "null"} }
+    {                  		     f.dblline   }
+    { " Vertical task listing   "  f.exec { setOpts verticalList True }
+				bitmap {($listOrient==1)? "check" : "null"}
+					    }
+    { " Horizontal task listing "  f.exec { setOpts verticalList False }
+				bitmap {($listOrient==0)? "check" : "null"}
+					    }
+    {                       	f.dblline       	}
+    { " Tcl Command Shell       "  f.exec {  tclOpen } }
+} ; createMenu optsMenu optionsButton $optsMenuDescription
+
+
+proc setOpts { opt args } {
+    global optsMenuDescription
+    global listOrient showType showFiles
+    global HTask HPkg HOpt hcurrent
+
+    switch $opt {
+	showType	{ set showType   [expr { ($showType == 1) ? 0 : 1 } ]
+			  GCmd type $showType
+			  set h $hcurrent
+			  GCmd help $HTask($h) $HPkg($h) $HOpt($h)
+			}
+	showFiles	{ set showFiles  [expr { ($showFiles == 1) ? 0 : 1 } ]
+			}
+	verticalList	{ set listOrient [expr { ($args == "False") ? 0 : 1 } ]
+			  send topicList set verticalList $args 
+			}
+    }
+    editMenu optsMenu optionsButton $optsMenuDescription
+}
+
+proc Print args {
+    send printShell map
+} ; send printButton addCallback Print
+
+proc Find { args } { 
+    send findShell popup
+} ; send findButton addCallback Find
+
+proc Search { args } { 
+    global search_mapped
+
+    send searchShell map
+    set search_mapped 1
+} ; send searchButton addCallback Search
+
+proc Reload args {
+    global HPkg HType HUrl HOpt HTask HFile hcurrent
+
+    if { $HType($hcurrent) == "file"} {
+        if {[info exists HFile($hcurrent)] == 1} {
+	    GCmd directory open $HFile($hcurrent)
+        } else { 
+	    setAlert param old [format "HFile param at %d not found" $hcurrent]
+        }
+    } else {
+        loadHistItem $hcurrent $HPkg($hcurrent) $HTask($hcurrent) \
+   	     $HOpt($hcurrent) $HType($hcurrent) $HUrl($hcurrent)
+    }
+} ; send reloadButton addCallback Reload
+
+proc Help args {
+    send hlpShell map
+} ; send helpButton addCallback Help
+
+proc Open args {
+    send fmtGroup unmap
+    send fmtGroup set height 0
+    send fbcOkay set label Load
+    send fileBrowser map
+}
+
+proc SaveAs args {
+    global format
+
+    # Reset the default format every time we open.
+    send $format set on 0
+    set format fmtSrc
+    send $format set on 1
+
+    # If there's no filename specified set one as a default.
+    setDefaultFname
+    
+    send fmtGroup map
+    send fmtGroup set height 65
+    send fbcOkay set label Save
+    send fileBrowser map
+}
+
+proc Quit args {
+    GCmd quit
+    deactivate unmap 
+}; send quitButton addCallback Quit
+
+
+
+################################################################################
+# Callbacks for client state variables (UI parameter objects).  When the
+# client's state changes it updates a UI parameter to reflect the change.
+# This produces a callback to one or more of the callbacks defined below,
+# used to update the GUI to reflect the changing state of the client.
+################################################################################
+
+proc setShowType { param old new } {
+    global showType listOrient showFiles optsMenuDescription
+    set showType $new
+    editMenu optsMenu optionsButton $optsMenuDescription
+}; send showtype addCallback setShowType
+
+proc setTopics { param old new } {
+    global pkgList
+    set pkgList $new
+    send topicList setList $new resize
+}; send pkglist addCallback setTopics
+
+proc appendHist { param old new } {
+    global helpType helpOption curtask curpack
+
+    if {$new == "package"} {
+        #addHistRecord $curpack $curpack "" $helpOption "" $helpType
+        addHistRecord $curpack $curpack "" $helpOption "" "package"
+    } elseif {$new == "append"} {
+        # We've got a result of some kind and all of the values have been
+        # set, so create a history record.
+        if {$curpack != "" && $curtask != ""} {
+            addHistRecord $curpack $curtask "" $helpOption "" $helpType
+        }
+    }
+}; send history addCallback appendHist
+
+proc setCurpack { param old new } {
+    global curpack
+
+    if { $new != [getPkgName $curpack] } {
+        if { $curpack != "" && \
+	     $curpack != "clpackage" && \
+	     [string match "root*" $new] != 1} {
+                 set curpack [ format "%s.%s" $curpack $new]
+	         send htuButton set sensitive true
+        } else {
+            set curpack $new
+        }
+    }
+}; send curpack addCallback setCurpack
+
+
+proc setCurtask { param old new } {
+    global curtask curpack helpType
+
+    if {$helpType == "package"} {
+       send topicEntry set string $curpack
+    } else {
+       send topicEntry set string [ format "%s.%s" $curpack $new ]
+    }
+   
+    # Update the printer dialog so the filename defaults to the curtask.
+    if { [send toFile get on] } {
+	if {$HType($hcurrent) == "file" || [send srcOpt get on] == 1} {
+            send printEntry set string [format "%s" [fileSource] ]
+	} else {
+            send printEntry set string [format "%s.ps" $new]
+	}
+    } else {
+        send printEntry set string "printer"
+    }
+    set curtask $new
+}; send curtask addCallback setCurtask
+
+
+proc getPkgName { pkg } {
+    set last [ string last "." $pkg ]
+    if { $last > -1 } {
+        return [ string range $pkg [incr last] end ]
+    } else {
+        return $pkg
+    }
+}
+
+proc getParentName { pkg } {
+    set last [ string last "." $pkg ]
+    if { $last > -1 } {
+        set root [ string range $pkg 0 [incr last -1] ]
+    } else {
+        set root $pkg
+    }
+    return [getPkgName $root]
+}
+
+
+# Topic list selection callback.
+proc topicSelect { widget event args } {
+    global htop hcurrent curpack curtask helpOption helpType
+    global visited
+
+    # If we're currently positioned somewhere in the middle of the 
+    # history menu, push the current page to the history list before
+    # getting the next result.
+    if {$htop != $hcurrent} {
+        addHistRecord $curpack $curtask "" $helpOption "" $helpType
+	send htbButton set sensitive True
+	send htuButton set sensitive True
+    }
+
+    set item [string trimright [send topicList getItem itemno] "."]
+    send topicList getItem itemno
+    if { $itemno != "none" } {
+	GCmd help $item $curpack $helpOption
+        set visited($item) 1
+    }
+} ; send topicList addEventHandler topicSelect buttonReleaseMask
+
+proc listViewResize { args } {
+    global pkglist
+    send topicList setList $pkglist resize
+} ; send listView addEventHandler listViewResize ResizeRedirectMask
+
+
+# Get help for a specific topic from the topic entry widget.
+proc getTopicHelp { widget mode topic args } {
+    global curpack helpOption
+
+    if { [string match "*\.*" $topic] == 1} {
+	set pkglist [ split $topic "." ]
+	set pack [lindex $pkglist [expr {[llength $pkglist] - 2}] ]
+	set task [lindex $pkglist [expr {[llength $pkglist] - 1}] ]
+	set curpack $pack
+        GCmd help $task $curpack $helpOption
+    } else {
+	set curtask ""
+	set curpack ""
+        GCmd help $topic $curpack $helpOption
+    }
+} ; send topicEntry addCallback getTopicHelp
+
+proc topicClear { args } {
+    send topicEntry set string ""
+} ; send topicClear addCallback topicClear
+
+
+
+################################################################################
+# Help text and HTML processing procedures. 
+################################################################################
+
+proc setHelpResult { param old new } {
+    global helpType helpOption curtask curpack
+    global secMenuDescription parMenuDescription
+    global pkgList debug
+
+    # Debug status
+    if {$debug == 1} {
+	send tclEntry append \
+	    [format "helpres: type=%1.1s opt=%4.4s curtask=%s curpack=%s\n" \
+	    $helpType $helpOption $curtask $curpack ]
+	print [format "helpres: type=%1.1s opt=%4.4s curtask=%s curpack=%s\n" \
+	    $helpType $helpOption $curtask $curpack ]
+    }
+
+    if { [string match "*<HTML>*" $new] == 1} {
+	# Strip the header table.
+	if {[ string match "*TABLE*" $new] == 1 } {
+            set new_start [expr [string first "</TABLE>" $new] + 12]
+            set text [ filterBraces [string range $new $new_start end] ]
+	} else {
+	    set text [ filterBraces $new ]
+	}
+
+	# Got HTML directly from the client.
+	send helpText setText [format "<HTML><BODY>\n%s" $text ]
+
+	# Save the source for the viewer
+	set docSrc \
+	    [ format "<HTML><BODY>\n%s\n</BODY></HTML>\n" $text]
+
+	# Parse the file for menu items.
+	setSectionMenu   $new
+	setParameterMenu $new
+
+    } else {
+	# Disable help page content buttons for plain text.
+        send parButton set sensitive False
+        send secButton set sensitive False
+        editMenu secMenu secButton $secMenuDescription
+        editMenu parMenu parButton $parMenuDescription
+
+	# Filter plaintext .men files into something with links, otherwise
+	# unescape the curly braces used to pass the text through Tcl.
+	if {$helpType == "package"} {
+	    set str [ filterLinks [ filterTcl $new ] ]
+	} else {
+	    set str [ filterTcl $new ]
+	}
+	
+	# Load the results
+	send helpText setText "<HTML><BODY><PRE>\n$str\n</PRE></BODY></HTML>"
+
+	# Save the source for the viewer
+	set docSrc \
+	    [ format "<HTML><BODY><PRE>\n%s\n</PRE></BODY></HTML>\n" $str]
+    }
+    send helpText retestAnchors
+    send srcText set string $docSrc
+
+    # See which files associated with this topic are available.  We turn off
+    # the option toggles first so they can be reset as needed by the client.
+    if {$helpOption != "sysdoc"} {send sysOpt "set sensitive False ; set on 0" }
+    if {$helpOption != "source"} {send srcOpt "set sensitive False ; set on 0" }
+
+    if {$helpType == "package"} {
+        set parent [getParentName $curpack]
+	if {$parent == "clpackage" || [string match "root*" $parent] == 1} {
+            GCmd files $curtask "clpackage"
+	} else {
+            GCmd files $curtask $parent
+	}
+    } else {
+        GCmd files $curtask [string range $curpack \
+		[expr {[string last "." $curpack ] + 1}] end]
+    }
+
+    # Highlight the package list item.
+    for {set i 0} {$i < [llength $pkgList]} {incr i} {
+	if {[lindex $pkgList $i] == $curtask } {
+	    send topicList highlight $i
+ 	    break
+	}
+    }
+    printHistStack "helpres "
+
+}; send helpres addCallback setHelpResult
+
+
+# Set an arbitrary TextToggle widget highlight color.
+proc setOptColor { widget color args } {
+    send $widget "set on 1 ; \
+		  set offIcon diamond0s ; \
+     		  set highlightColor $color ; \
+    		  set background gray75 ; \
+		  set onIcon diamond1s ; \
+     		  set highlightColor $color ; \
+    		  set background gray75"
+}
+
+
+# Set the file options for the files that were found to be valid (i.e.
+# they're listed in the files output and actually exist).
+proc setHelpFileOpts { param old new } {
+    global helpType helpOption curtask curpack
+    global showFiles
+
+    set opt  [lindex $new 0]
+    set val  [lindex $new 1]
+    if {$opt != "file"} { set stat [lindex $new 2] }
+
+    # Set the option toggles according to valid files.
+    if {$opt == "sys" && $stat == 0} { 
+        send sysOpt set sensitive True
+    } elseif {$opt == "sys" && $stat == 1 && $showFiles == 1} { 
+        setOptColor sysOpt yellow
+    }
+    if {$opt == "src" && $stat == 0} { 
+        send srcOpt set sensitive True
+    } elseif {$opt == "src" && $stat == 1 && $showFiles == 1} { 
+        setOptColor srcOpt yellow
+    }
+
+    # Update the help files panel text.
+    if {$opt == "file"} {
+        set pkg [ string trimright $val ":"]
+        send flpkgVal set label [ format "%s" $pkg ]
+	send toplevel set title [format "XHelp: %s" $pkg ]
+        send flistText set string ""
+    } else {
+        send flistText append [ format " %5.5s    %-7.7s %s\n" \
+	    $opt \
+	    [ expr { ($stat == 0) ? "Okay" : "Error" }] \
+	    $val]
+    }
+
+} ; send helpfiles addCallback setHelpFileOpts
+
+
+# Process an HREF link selection.  URLs are assumed to be of the form
+#
+#		<pkgname>.<task>
+#		<task>
+#		'#'<hname>
+#
+# If an internal link is found as in the last case we ignore any defined 
+# package/task given, otherwise load the selected page.
+
+proc textAnchorSelected {widget cbtype event text href args} {
+    global HPkg HType HUrl HOpt HTask HFile hcurrent htop
+    global curpack helpOption visited
+
+    set visited($href) 1
+    send helpText retestAnchors
+
+    if {[string match "*#*" $href] == 1} {
+	set link [string range $href [expr [string first "#" $href] + 1] end ]
+	set HUrl($hcurrent) $link
+	send helpText gotoId [ send helpText anchorToId $link ]
+    } else {
+        if { [string match "*\.*" $href] == 1} {
+            set pack [lindex [split $href "."] 0]
+            set task [lindex [split $href "."] 1]
+            set curpack $pack
+            GCmd help $task $curpack $helpOption
+
+        } else {
+            GCmd help $href $curpack $helpOption
+        }
+    }
+
+    if {$hcurrent <= $htop} {
+	set h $hcurrent
+    	addHistRecord $HPkg($h) $HTask($h) $HUrl($h) $HOpt($h) \
+	    HFile($h) $HType($h)
+    }
+
+}; send helpText addCallback textAnchorSelected anchor
+
+
+# Remove only the escaped curly braces.  Used to filter HTML text passed to
+# the GUI with escapes in it.
+
+proc filterBraces { istr args } {
+    if {$istr != ""} {
+        regsub -all {(\\\{)} $istr "\{"   v1
+        regsub -all {(\\\})} $v1   "\}"   results
+        return $results
+    }
+}
+
+# Remove the backslash escapes from source files, escape special chars for
+# presentation on an HTML widget.
+
+proc filterTcl { istr args } {
+    if {$istr != ""} {
+        regsub -all {(\\\{)} $istr "\{"   v1
+        regsub -all {(\\\})} $v1   "\}"   v2
+        regsub -all {(\<)}   $v2   "\\&lt;" v3
+        regsub -all {(\>)}   $v3   "\\&gt;" results
+        return $results
+    }
+}
+
+
+# Scan a plaintext doc to see if maybe this is a package help menu.  We use
+# some assumption that most of the lines will be of the form "task '-' desc"
+# then parse the file resetting all of the 'task' names as HREFs for other
+# tasks. 
+
+proc filterLinks { istr args } {
+
+    set lines [split $istr "\n"]
+    set blank " "
+    set results { }
+    lappend results ""
+    foreach i $lines {
+	set line [string trimleft [detab $i] ]
+        if {[regexp {(^[a-zA-Z0-9\ \_\(\)]+[\+|\-]*)} $line arg] == 1 &&
+	    ([string match "*\-*" $arg] == 1 ||
+	     [string first "\*" $line] > 0 || 
+	     [string match "*\+*" $arg] == 1) } {
+
+	    set task [string trim [string trimright $arg "-+"] ]
+	    set l [string first "-" $line]
+	    if {$l == -1} {
+		set l [string first "+" $line]
+	        if {$l == -1} {
+		    set l [string first "\*" $line]
+	        }
+	    }
+	    set desc [ string range $line $l end ]
+
+	    # We now have the task name and the description string, format
+	    # on output with the HREF defined.
+	    set nblanks [expr 13 - [string length $task] ]
+	    set fmtstr [format "%%%ds<A HREF=\"%%s\">%%s</A> %%s\n" $nblanks ]
+	    set ostr [format $fmtstr $blank $task $task $desc]
+
+	    lappend results $ostr
+
+	} elseif {[regexp {(^[a-zA-Z0-9\_]+\.[a-zA-Z0-9\_]+\:$)} $i val] == 1} {
+
+	    # Break out the task and package names.
+	    regsub -all {[\.:]} $i " " val
+	    scan $val "%s %s" parent child
+
+	    # Format a URL and append the results.
+	    if { [ string match "*root*" $parent] == 0} {
+	       set ref [format "<A HREF=\"%s\">%s</A>.<A HREF=\"%s\">%s</A>:\n"\
+		    $parent $parent $child $child ]
+	    } else {
+	       set ref [format "%s.<A HREF=\"%s\">%s</A>:\n"\
+		    $parent $child $child ]
+	    }
+	    lappend results $ref
+
+	} else {
+	    set nblanks [expr [string length $i] - [string length $line] ]
+	    set fmtstr [format "%%%ds%%s\n" $nblanks ]
+	    lappend results [format $fmtstr $blank $line ]
+	}
+    }
+    return [ join $results ]
+}
+
+
+# Generated a list of the lines and create the section menu.
+
+proc setSectionMenu { text args } {
+    global secMenuDescription
+
+    # Break out the table of contents from the string.  Note we're hard-
+    # wired here into the form of the comment string used to contain the
+    # section name.
+    set l [expr [string first "<! Contents: " $text] + 12]
+    if {$l == 0} { return }
+    set s [string range $text $l end]
+    set r [expr [string first ">" $s] - 4]
+    set t [string range $s 0 $r]
+    set lst [split $t '\'']
+
+    # Now take the list generated and create the menu.
+    set items { }
+    lappend items " \"Top of Page\" f.exec \{ send helpText gotoId 0 \}"
+    lappend items " f.dblline "
+    foreach i $lst {
+	if {$i != " "} { 
+	    set i [ string trimright $i ]
+	    regsub -all {[ ,.():;]} [string tolower $i] _ url
+	    lappend items "   \" $i  \" f.exec \{ jumpToName #s_$url \}"
+	}
+    }
+
+    if { [llength $items] == 3 } {
+       send secButton set sensitive False
+       editMenu secMenu secButton $secMenuDescription
+    } else {
+       editMenu secMenu secButton $items
+       send secButton set sensitive True
+    }
+}
+
+
+# Generated a list of the lines and create the parameter menu.
+proc setParameterMenu { text args } {
+    global parMenuDescription
+    set items { }
+    foreach i [split $text "\n"] {
+        if {[string match "\<\! Sec*PARAMETERS*Level=0*" $i] == 1} {
+    	    set l [expr [string first "Line='" $i] + 6]
+    	    set s [string range $i $l end]
+    	    set r [expr [string first "\'" $s] - 1]
+    	    set t [string range $s 0 $r]
+	    regsub -all {[\	]} $t " " d		;# remove tabs
+	    regsub -all {[\"]} $d "\\\"" entry
+
+    	    set l [expr [string first "Label='" $i] + 7]
+    	    set s [string range $i $l end]
+    	    set r [expr [string first "\'" $s] - 1]
+    	    set t [string range $s 0 $r]
+
+	    lappend items " \"$entry\" f.exec \{ jumpToName #l_$t \}"
+	}
+    }
+
+    if { [llength $items] == 0 } {
+       send parButton set sensitive False
+       editMenu parMenu parButton $parMenuDescription
+    } else {
+       editMenu parMenu parButton $items
+       send parButton set sensitive True
+    }
+}
+
+
+# Position the page to the requested href name.
+proc jumpToName { name } {
+    global curtask curpack helpType helpOption
+
+    send helpText gotoId [send helpText anchorToId $name]
+    send helpText retestAnchors
+
+    # Now add a history record for the jump
+    addHistRecord $curpack $curtask $name $helpOption "" $helpType
+}
+
+
+# Utility routine to 'detab' a line and preserve format.
+proc detab {str {tablen 8}} {
+    set a 0
+    set i [string first "\t" $str]
+    while {$i != -1} {
+        set m { }
+        set j $i
+        while {[incr j] % $tablen} { append m { } }
+        set str [string range $str $a \
+	    [expr {$i-1}]]$m[string range $str [incr i] end]
+        set i [string first "\t" $str]
+    }
+    return $str
+}
+
+
+################################################################################
+# Navigation and History Callbacks 
+################################################################################
+
+# Go back one page.
+proc Back args {
+    global curtask curpack helpType helpOption
+    global HPkg HType HUrl HOpt HTask HFile htop hcurrent
+
+    incr hcurrent -1
+    if {$hcurrent >= 0} {
+	set item $HTask($hcurrent)
+        set pkg  $HPkg($hcurrent)
+        set type $HType($hcurrent)
+	set h 	 $hcurrent
+
+	if { $item == "Home" } {
+	    loadHomePage
+	} else {
+    	    if { $HType($h) == "file"} {
+        	if {[info exists HFile($h)] == 1} {
+		    loadHistItem $h pkg $HFile($h) help file
+        	} else { 
+	    	    setAlert param old \
+			[format "HFile param at %d not found" $hcurrent]
+        	}
+    	    } else {
+	        loadHistItem $h $HPkg($h) $HTask($h) $HOpt($h) \
+		    $HType($h) $HUrl($h)
+	    }
+	}
+
+	if {$hcurrent == 0} {
+	    send htbButton set sensitive False
+	    send htuButton set sensitive False
+	}
+	if {$hcurrent >= 0} {
+	    send htfButton set sensitive True 
+	    send htuButton set sensitive True
+	}
+    } else {
+	set hcurrent 0
+    }
+    editHistoryMenu
+    printHistStack "Back "
+} ; send htbButton addCallback Back
+
+
+# Go forward one page.
+proc Forward args {
+    global curtask curpack helpType helpOption
+    global HPkg HType HUrl HOpt HTask HFile htop hcurrent
+
+    incr hcurrent
+    if {$hcurrent <= $htop} {
+	set item $HTask($hcurrent)
+        set pkg  $HPkg($hcurrent)
+        set type $HType($hcurrent)
+	set h 	 $hcurrent
+
+	if { $item == "Home" } {
+	    loadHomePage
+	} else {
+    	    if { $HType($h) == "file"} {
+        	if {[info exists HFile($h)] == 1} {
+		    loadHistItem $h pkg $HFile($h) help file
+        	} else { 
+	    	    setAlert param old \
+			[format "HFile param at %d not found" $hcurrent]
+        	}
+    	    } else {
+	        loadHistItem $h $HPkg($h) $HTask($h) $HOpt($h) \
+		    $HType($h) $HUrl($h)
+	    }
+	}
+
+	if {$hcurrent == $htop } {
+	    send htfButton set sensitive False
+	    send htbButton set sensitive True
+	} else {
+	    send htbButton set sensitive True
+	}
+    } else {
+	incr hcurrent -1
+    }
+    editHistoryMenu
+    printHistStack "Forward "
+} ; send htfButton addCallback Forward
+
+
+# Go up to previous package, skipping over pages inbetween.
+proc Up args {
+    global curtask curpack helpType helpOption
+    global HPkg HType HUrl HOpt HTask htop hcurrent
+
+    # From the current page go back until we find a package
+    if {$HType($hcurrent) == "package"} {
+	set i [expr {$hcurrent-1} ] 
+    } else {
+	set i $hcurrent
+    }
+    while {$HType($i) != "package" && $i >= 0} { 
+	incr i -1
+    }
+
+    # Found package, go get it.
+    set hcurrent $i
+    if {$i == 0} {
+	loadHomePage				;# push a history record??
+    } else {
+        GCmd load $HTask($i) [getPkgName $HPkg($i)] $HOpt($i)
+    }
+
+    set curtask    $HTask($hcurrent) 		;# update the state of things
+    set curpack    $HPkg($hcurrent)
+    set helpOption $HOpt($hcurrent)
+    set helpType   $HType($hcurrent)
+
+    send topicEntry set string $curpack 	;# update topic entry string
+
+    if {$hcurrent == 0} {			;# adjust navigation buttons
+        send htbButton set sensitive False
+        send htuButton set sensitive False
+    }
+    if {$hcurrent >= 0} {
+        send htfButton set sensitive True 
+        send htuButton set sensitive True
+    }
+    editHistoryMenu
+    printHistStack "Forward "
+} ; send htuButton addCallback Up
+
+
+# Go straight to the homepage.
+proc Home args {
+    global curtask curpack helpType helpOption
+    global HPkg HType HUrl HOpt HTask HFile htop hcurrent
+
+    # Load the homepage.
+    loadHomePage
+
+    # A Home command jumps over everything in the history list but we
+    # need to push a history record for it anyway.
+    addHistRecord $HPkg(0) $HTask(0) $HUrl(0) $HOpt(0) $HFile(0) $HType(0)
+
+    send topicEntry set string ""
+    set curtask ""
+    set curpack ""
+    set helpType "package"
+    set helpType "help"
+} ; send hthButton addCallback Home
+
+
+# Load the homepage.
+proc loadHomePage { args } {
+    global curtask curpack version showType
+
+    GCmd help Home
+
+    # Clean up.
+    set curtask ""
+    set curpack ""
+    send topicEntry set string ""
+    send toplevel set title $version
+}
+
+# Clear all the history information.
+proc histClear args {
+    global HPkg HType HUrl HOpt HTask HFile htop hcurrent
+    global visited
+
+    # Clear the visited anchors list.
+    foreach i [array names visited] {
+	unset visited($i)
+    }
+
+    # Clear the history stack.
+    for { set i [expr {$htop -1}] } { $i >= 0 } { incr i -1 } {
+	catch {
+	    unset HType($i)
+	    unset HOpt($i)
+	    unset HTask($i)
+	    unset HPkg($i)
+	    unset HUrl($i)
+	    unset HFile($i)
+	}
+    }
+
+    # Reinitialize, but save the current page as the new history stack.
+    catch {
+        set HPkg(0)     $HPkg($htop)	;# package
+        set HOpt(0)     $HOpt($htop)	;# option
+        set HTask(0)    $HTask($htop)	;# task
+        set HUrl(0)     $HUrl($htop)	;# url
+        set HType(0)    $HType($htop)	;# type
+        set HFile(0)    $HFile($htop)	;# filename
+    }
+    set htop	    0
+    set hcurrent    0
+
+    # Update navigation options and history menu.
+    send htbButton set sensitive False
+    send htfButton set sensitive False
+    send htuButton set sensitive False
+    editHistoryMenu
+}
+
+
+# Push an item on the history stack.
+proc addHistRecord { pkg task url opt file type } {
+    global HPkg HType HUrl HOpt HTask HFile htop hcurrent
+    global helpType helpOption
+    global historyMenuDescription
+
+    # Push a new history record to the top of the stack and make that the
+    # current record.
+    incr htop
+    set HPkg($htop) $pkg
+    set HTask($htop) $task
+    set HUrl($htop) $url
+    set HOpt($htop) $opt
+    set HFile($htop) [ expr {($file == "") ? "none" : $file } ]
+    set HType($htop) $type
+    set hcurrent $htop
+
+    # Activate the Back button.
+    if {$hcurrent == 1} { 
+	send htbButton set sensitive True 
+	if {$type == "package"} {
+	    send htuButton set sensitive True 
+	}
+    }
+
+    # Edit the history menu.
+    editHistoryMenu
+    printHistStack "addHistRecord"
+}
+
+
+# Edit the history menu to reflect the current state.
+proc editHistoryMenu { args } {
+    global HPkg HType HUrl HOpt HTask htop hcurrent
+    global helpType helpOption
+    global historyMenuDescription
+    global navMenuDescription
+    global MAX_MENU_SIZE
+
+    set items $historyMenuDescription
+    set nitems 0
+    if {$htop > $MAX_MENU_SIZE} {
+        set nstart [ min $htop [expr {$hcurrent + 3}] ]
+    } else {
+        set nstart $htop
+    }
+    for { set i $nstart } { $i >= 0 } { incr i -1 } {
+	set pkg  $HPkg($i)
+	set task $HTask($i)
+	set type $HType($i)
+	set opt  $HOpt($i)
+	set url  $HUrl($i)
+        if {$pkg != "" || $type == "file"} {
+	    if {$type == "task"} {
+		if {$url == ""} {
+		    set entry [format "%-22.22s %4s" $task \
+		    	[menuItemType $type $url $opt ] ]
+		} else {
+		    set entry [format "%-22.22s %4s" \
+			[ format "%s (%s)" $task [string trimleft $url "#"] ] \
+		    	    [menuItemType $type $url $opt ] ]
+		}
+	    } elseif {$type == "package"} {
+		set entry [format "%-22.22s %4s" [getPkgName $pkg] \
+		    [menuItemType $type $url $opt ] ]
+	    } elseif {$type == "file"} {
+		upvar #0 HFile file
+		set entry [format "%s" $file($i) ]
+	    } else {
+		setAlert param old [format "Unknown help type: %s" $type]
+	    }
+	   
+	    if {$type == "file"} {
+                lappend items " \" $entry \" f.exec \{ \
+		    loadHistItem $i pkg $entry help file \} \
+		    bitmap \{\($i==$hcurrent\) ? \"arrow\" : \"null\" \} "
+	    } else {
+                lappend items " \" $entry \" f.exec \{ \
+		    loadHistItem $i [getPkgName $pkg] $task $opt $type $url \} \
+		    bitmap \{\($i==$hcurrent\) ? \"arrow\" : \"null\" \} "
+	    }
+	}
+
+	incr nitems 1
+	if {$nitems > $MAX_MENU_SIZE} {
+            lappend items "f.dblline"
+            lappend items " \"    History truncated... \" f.exec \{ \} "
+	    break
+	}
+    }
+    editMenu historyMenu historyButton $items
+
+    # Edit the navigation menu to get the sensitivities right for
+    # the current state.
+    editMenu navMenu helpText $navMenuDescription
+}
+
+
+# Utility routine to set the history item entry type.
+proc menuItemType { type url opt } {
+    if {$url != ""} {
+	return [format "Link"]
+    }
+
+    switch $opt {
+      "help"     { return [format "%s" [expr {($type=="task")?"Task":"Pkg"} ]] }
+      "source"   { return [format "(src)"] }
+      "sysdoc"   { return [format "(sys)"] }
+    }
+}
+
+
+# Load a particular page/link from the history list.
+proc loadHistItem { itemno pkg task opt type args } {
+    global HPkg HType HUrl HOpt HTask htop hcurrent
+    global curtask curpack helpType helpOption hcurrent
+    global version
+
+    # Load the requested page. Check whether we just need to jump to
+    # the current page.
+    if {$task == "Home"} {
+	loadHomePage
+
+    } elseif {$type == "file"} {
+	GCmd directory open $task
+
+    } elseif {$itemno == $hcurrent || \
+       ($pkg  != $curpack    || \
+	$task != $curtask    || \
+	$type != $helpType   || \
+	$opt  != $helpOption) } {
+    	    GCmd load $task $pkg $opt
+    }
+
+    # If the history item included an internal link, jump to it.  The
+    # 'args' value will either be the URL or an empty string.
+    if {$args != ""} {
+        send helpText gotoId [send helpText anchorToId $args]
+        send helpText retestAnchors
+    }
+
+    # Update the topic entry string.
+    if { $type == "task" } {
+       send topicEntry set string [ format "%s.%s" $pkg $task ]
+    } elseif { $type == "file" } {
+       send topicEntry set string $task
+    } else {
+       send topicEntry set string $pkg
+    }
+
+    # Change the options button if needed.
+    if {$HOpt($itemno) != $helpOption} {
+	send [ getOptWidget $helpOption ] set on 0
+	setOptColor [ getOptWidget $HOpt($itemno) ] green
+    }
+
+    # Update the current entry.
+    set hcurrent $itemno
+
+    if { $type != "file" } {
+        set curtask    $HTask($hcurrent)
+        set curpack    $HPkg($hcurrent)
+        set helpOption $HOpt($hcurrent)
+        set helpType   $HType($hcurrent)
+    } else {
+        set helpOption "help"
+        set helpType   "file"
+    }
+
+    # Tweak the navigation buttons.
+    if {$hcurrent == 0} {
+        send htbButton set sensitive False
+        send htuButton set sensitive False
+    }
+    if {$hcurrent >= 0} {
+        send htfButton set sensitive True 
+        send htuButton set sensitive True
+    }
+    if {$hcurrent == $htop } {
+        send htfButton set sensitive False
+        send htbButton set sensitive True
+    }
+
+    # Edit the history menu.
+    editHistoryMenu
+}
+
+# Initialize the history menu.
+editHistoryMenu
+
+# Given the option type return the widget name.
+proc getOptWidget { opt } {
+    switch $opt {
+        "help"     { return "hlpOpt" }
+        "source"   { return "srcOpt" }
+        "sysdoc"   { return "sysOpt" }
+    }
+}
+
+# Debug utility to print the history stack.
+proc printHistStack { where args } {
+    global HPkg HType HUrl HOpt HTask HFile htop hcurrent
+    global debug
+
+    # Print the stack...
+    if {$debug > 0} {
+        print "_______________________________________________________________"
+        print $where
+        for { set i $htop } { $i >= 0 } { incr i -1 } {
+	    if {$HType($i) == "file"} {
+		upvar #0 HFile file
+            	print [format "%3s%d: type=%1.1s file=%s\n"\
+	            [ expr {($i==$hcurrent) ? ">>>" : "---"} ] \
+	            $i $HType($i) $file($i) ]
+	    } else {
+            	print [format "%3s%d: type=%1.1s opt=%4.4s task=%s pack=%s\n"\
+	            [ expr {($i==$hcurrent) ? ">>>" : "---"} ] \
+	            $i $HType($i) $HOpt($i) $HTask($i) $HPkg($i) $HUrl($i) ]
+	    }
+        }
+    }
+}
+
+# Test whether an anchor has been visited.
+proc testAnchor {widget cbtype href} {
+    global visited
+    return [info exists visited($href)]
+}
+send hlpText  addCallback testAnchor testAnchor
+send helpText addCallback testAnchor testAnchor
+send resList  addCallback testAnchor testAnchor
+
+
+################################################################################
+# Options Menu 
+################################################################################
+
+proc setType { param old new }  {
+    global helpType
+    set helpType [string tolower $new ]
+} ; send type addCallback setType
+
+proc selectOption { widget type value args } {
+    global curtask curpack helpOption
+
+    if { $curtask != "" } { 
+        foreach i { hlpOpt srcOpt sysOpt } { send $i set on 0 }
+        setOptColor $widget green
+        send $widget set on 1
+        switch $widget {
+	    hlpOpt   { set helpOption help
+		       send fmtText setSensitive true 
+		       send fmtPS setSensitive true 
+		     }
+	    srcOpt   { set helpOption source
+		       send fmtText setSensitive false 
+		       send fmtPS setSensitive false 
+		     }
+	    sysOpt   { set helpOption sysdoc
+		       send fmtText setSensitive true 
+		       send fmtPS setSensitive true 
+		     }
+        }
+	GCmd help $curtask [getParentName $curpack] $helpOption
+    }
+}; foreach i {hlpOpt srcOpt sysOpt } { send $i addCallback selectOption }
+
+proc toggleFileOption { args } {
+    if { [ send filOpt get on] == 1 } { 
+	send fileShell map
+    } else { 
+	send fileShell unmap
+    }
+} ; send filOpt addCallback toggleFileOption
+
+send flDismiss addCallback "send fileShell unmap ; send filOpt set on 0"
+
+
+################################################################################
+#  Procedure used by the printer prompt box. 
+################################################################################
+
+proc setPrinterName { param old new } {
+    send printEntry set string $new
+}; send printer addCallback setPrinterName
+
+set page_size	pageLetter
+
+proc pageRadio { widget type state args } {
+    global page_size
+
+    if {$state == 0} {
+	# Don't allow a button to be turned off.
+	send $widget set on 1
+    } else {
+	send $page_size set on 0
+	set page_size $widget
+    }
+} 
+foreach w {pageLetter pageLegal pageA4 pageB5} { send $w addCallback pageRadio }
+
+proc toPrinterToggle args {
+    global curtask HType hcurrent
+    if { [send toPrinter get on] } {
+	send toFile set on False
+	send printLabel set label "Printer: "
+        send printEntry set string "printer"
+    } else {
+	send toFile set on True
+	send printLabel set label "File Name: "
+	if {$HType($hcurrent) == "file" || [send srcOpt get on] == 1} {
+            send printEntry set string [format "%s" [fileSource] ]
+	} else {
+            send printEntry set string [format "%s.ps" $curtask]
+	}
+    }
+} ; send toPrinter addCallback toPrinterToggle
+
+proc toFileToggle args {
+    global curtask HType hcurrent
+    if { [send toFile get on] } {
+	send toPrinter set on False
+	send printLabel set label "File Name: "
+	if {$HType($hcurrent) == "file" || [send srcOpt get on] == 1} {
+            send printEntry set string [format "%s" [fileSource] ]
+	} else {
+            send printEntry set string [format "%s.ps" $curtask]
+	}
+    } else {
+	send toPrinter set on True
+	send printLabel set label "Printer: "
+        send printEntry set string "printer"
+    }
+} ; send toFile addCallback toFileToggle
+
+proc doPrintOkay { args } {
+    global curtask curpack HType hcurrent
+
+    set device [ send printEntry get string ]
+    if { [send toPrinter get on] } {
+        GCmd print $curtask $curpack $device
+    } else {
+	if {$HType($hcurrent) == "file" || [send srcOpt get on] == 1} {
+            GCmd directory save [fileSource] $fname 1 source
+	} else {
+            GCmd directory save [fileSource] $fname 1 postscript
+	}
+    }
+    send printShell unmap
+}
+send printOkay addCallback doPrintOkay
+send printEntry addCallback doPrintOkay
+
+send printDismiss addCallback "send printShell unmap "
+
+
+################################################################################
+#  Procedures used by the fileBrowser.
+################################################################################
+
+# File browsing globals
+set curdir	""				;# current directory
+set pattern	"*"				;# filename template
+set format	"fmtSrc"			;# SaveAs format
+
+
+# Browser selection callback.
+proc browserSelect { widget event args } {
+    global curdir helpOption
+
+    set opt [expr {$widget == "dirList" ? "dirlist" : "loadfile"}]
+    set item [send $widget getItem itemno]
+    send $widget getItem itemno
+    set mode [send fbcOkay get label]
+
+    if { $itemno != "none" } {
+        if { $mode == "Load"} {
+	    if {$opt != "dirlist"} {
+    	        addHistRecord "" "" "" help [format "%s%s" $curdir $item] "file"
+	    }
+            GCmd directory $opt $item
+	    send flistText set string [format "  file    Okay    %s%s\n" \
+		$curdir $item]
+        } else {
+	    if {$opt == "dirlist"} {
+                GCmd directory $opt $item
+	    } else {
+	        send fnameEntry set string [format "%s%s" $curdir $item]
+	    }
+        }
+    }
+}
+send dirList  addEventHandler browserSelect buttonReleaseMask
+send fileList addEventHandler browserSelect buttonReleaseMask
+
+
+# Client callback.
+proc browserListing { param old new } {
+    global curdir pattern
+
+    set option [ lindex $new 0 ]
+    switch $option {
+	dirlist	    { set list [lindex $new 1]
+		      send dirList  setList $list resize
+		    }
+	filelist    { set list [lindex $new 1]
+		      send fileList setList $list resize
+		    }
+	template    { set pattern [lindex $new 1]
+		      send filterEntry set string $pattern
+		    }
+	curdir      { set curdir [lindex $new 1]
+		      send curdirVal set label $curdir
+		    }
+	selection   { send fnameEntry set string [lindex $new 1] }
+    }
+} ; send directory addCallback browserListing
+
+
+# Set the filename matching template.
+proc setTemplate { widget mode pattern args } {
+    GCmd directory template $pattern
+} ; send filterEntry addCallback setTemplate
+
+
+# get the filename of the currently displayed page.
+proc fileSource { args } {
+    global helpOption
+
+    set str [ send flistText get string ]
+
+    set fname ""
+    for {set i 0} {$i < [llength $str]} {incr i 3} {
+        set j [expr {$i + 2} ]
+	if {($helpOption == "source" && [lindex $str $i] == "src") ||
+	    ($helpOption == "sysdoc" && [lindex $str $i] == "sys") ||
+	    ($helpOption == "help" && [lindex $str $i] == "hlp") ||
+	    ($helpOption == "file" && [lindex $str $i] == "file")} {
+	        set fname [lindex $str $j]
+		break
+        }
+    }
+    return $fname
+}
+
+# Open a specific file, either to load a new page or save the current page.
+proc openFile { widget args } {
+    global curdir helpOption
+
+    set fname [send fnameEntry get string]
+
+    if {$fname == ""} {
+	setAlert param old "No filename specified"
+    } else {
+        if { [send fbcOkay get label] == "Load"} {
+    	    addHistRecord "" "" "" "" [format "%s%s" $curdir $fname] "file"
+            GCmd directory open $fname
+	    send flistText set string [format " file    Okay    %s\n" $fname]
+        } else {
+	    set page 	[fileSource]
+	    set ow [send overwrite get on]
+	    if {[send fmtSrc get on] == 1} {
+                GCmd directory save $page $fname $ow source
+	    } elseif {[send fmtText get on] == 1} {
+                GCmd directory save $page $fname $ow text
+	    } elseif {[send fmtHTML get on] == 1} {
+                GCmd directory save $page $fname $ow html
+	    } elseif {[send fmtPS get on] == 1} {
+                GCmd directory save $page $fname $ow postscript
+	    }
+        }
+    }
+} ; send fbcOkay addCallback openFile
+
+
+# Make the SaveAs formats a radio box.
+proc fmtRadio { widget type state args } {
+    global format
+
+    if {$state == 0} {
+	# Don't allow a button to be turned off.
+	send $widget set on 1
+    } else {
+	send $format set on 0
+	set format $widget
+    }
+
+    # If there's no filename specified set one as a default.
+    setDefaultFname
+
+} ; foreach w {fmtSrc fmtText fmtHTML fmtPS} { send $w addCallback fmtRadio }
+
+
+# Set a default filename based on the selected format and task name
+proc setDefaultFname args {
+    global format curtask curdir
+
+    set fname [send fnameEntry get string]
+    if {$curtask != ""} {
+        switch $format {
+	    fmtSrc  { send fnameEntry \
+			set string [format "%s%s" $curdir $curtask]  }
+	    fmtText { send fnameEntry \
+			set string [format "%s%s.txt" $curdir $curtask]  }
+	    fmtHTML { send fnameEntry \
+			set string [format "%s%s.html" $curdir $curtask] }
+	    fmtPS   { send fnameEntry \
+			set string [format "%s%s.ps" $curdir $curtask]   }
+        }
+    }
+}
+
+proc browserHelp args {
+    if { [send fbcOkay get label] == "Load"} {
+	showHelp lfiles
+    } else {
+	showHelp sfiles
+    }
+} ; send fbcHelp addCallback browserHelp
+
+send fnavHome 	 addCallback "GCmd directory home"
+send fnavUp   	 addCallback "GCmd directory up"
+send fnavRoot 	 addCallback "GCmd directory root"
+send fnavRescan  addCallback "GCmd directory rescan"
+send fnameClear  addCallback "send fnameEntry set string \"\""
+send filterClear addCallback "send filterEntry set string \"\""
+send fbcDismiss  addCallback "send fileBrowser unmap"
+
+
+
+################################################################################
+# Procedures used by the find box. 
+################################################################################
+
+proc doFindOkay args {
+    set dir     forward
+    set case    caseless
+    set phrase  [send findEntry get string]
+
+    if { $phrase != "" } {
+	if { [send findDir get on] }  { set dir backward  }
+	if { [send findCase get on] } { set case caseSensitive }
+
+	if { [send helpText searchText $phrase start end $dir $case] > 0 } { 
+	    set elid   [lindex [lindex $start 0] 0]
+	    set id     [max 1 [expr $elid - 10] ]
+
+	    send helpText gotoId $id
+	    send helpText setSelection $start $end
+	} else {
+            send warnText set label "Search string not found."
+            send warning map
+	}
+    } else {
+        send warnText set label "Warning: No search phrase entered."
+        send warning map
+    }
+} ; foreach w { findOkay findEntry } { send $w addCallback doFindOkay }
+
+send findClear   addCallback { send findEntry set string "" }
+send findDismiss addCallback { send findShell popdown	    }
+
+
+################################################################################
+# Procedures used by the apropos prompt box. 
+################################################################################
+
+proc doSearchOkay args {
+    set phrase  [send searchEntry get string]
+    if { $phrase != "" } {
+	send searchStatus set label "Searching..."
+        GCmd search [send exactMatch get on] $phrase
+    } else {
+        send warnText set label "Warning: No search phrase entered."
+        send warning map
+    }
+} ; foreach w { searchOkay searchEntry } { send $w addCallback doSearchOkay }
+
+proc searchResults { param old new } {
+    global search_mapped
+
+    if {$search_mapped == 0} {
+	Search
+    }
+    send resList setText $new
+    send resList retestAnchors
+    send searchStatus set label ""
+} ; send apropos addCallback searchResults
+
+# Selection callback.
+proc searchAnchorSelected {widget cbtype event text href args} {
+    global helpOption helpType curpack curtask
+    global visited
+
+    # Break out the task and package names.
+    set pack [lindex [split $href "."] 0]
+    set task [lindex [split $href "."] 1]
+
+    # Set the state and load the page.
+    if {$helpOption != "help"} {
+	send [getOptWidget $helpOption] set on 0
+	setOptColor hlpOpt green
+    }
+    set curtask    $task
+    set curpack    $pack
+    set helpOption "help"
+    set helpType   [expr {($pack == $task) ? "package" : "task"}]
+    GCmd load $curtask $curpack $helpOption
+
+    # Add the history record, one for the package and one for the task.
+    addHistRecord $curpack $curpack "" $helpOption "" "package"
+    if {$pack != $task} {
+        addHistRecord $curpack $curtask "" $helpOption "" "task"
+    }
+    send htbButton set sensitive True 
+    send htuButton set sensitive True 
+    editHistoryMenu
+    printHistStack "searchAnchorSelected"
+
+    # Update the topic entry string.
+    if { $task == $pack } {
+       send topicEntry set string $pack
+    } else {
+       send topicEntry set string [ format "%s.%s" $pack $task ]
+    }
+
+    set visited($href) 1
+    send resList retestAnchors
+
+} ; send resList addCallback searchAnchorSelected anchor
+
+proc doSearchClear args {
+    send searchEntry set string ""
+} ; send searchClear addCallback doSearchClear
+
+proc doSearchDismiss args {
+    global search_mapped
+
+    set search_mapped 0
+    send searchShell unmap
+} ; send searchDismiss addCallback doSearchDismiss
+
+send searchHelp addCallback { showHelp search }
+
+
+################################################################################
+#     Define procedures for the help panel    
+################################################################################
+
+# Stuff for keeping track of visited anchors.
+set h_links 		{ 0 }
+set h_linkIndex 	  0
+
+proc getHelpText { param old new } {
+    send hlpText setText $new
+}; send help addCallback getHelpText
+
+proc anchorSelected {widget cbtype event text href args} {
+    global visited h_links h_linkIndex
+    set anchID [send hlpText anchorToId $href]
+    set visited($href) 1
+    if {$h_linkIndex == 0} {
+	send hlpBack set sensitive True
+        if {[lindex $h_links 1] != $anchID} {
+	    set h_links { 0 }
+	    send hlpForward set sensitive False
+        } 
+    } 
+    if {$h_linkIndex > 0 && \
+	[lindex $h_links [expr $h_linkIndex + 1]] != $anchID} {
+	    #set h_links [lrange $h_links 0 $h_linkIndex]
+	    set pos [send hlpText positionToId 0 0]
+	    set h_links [lreplace $h_links $h_linkIndex end $pos]
+    }
+    if {[lindex $h_links [expr $h_linkIndex + 1]] != $anchID} {
+        lappend h_links $anchID
+        incr h_linkIndex
+    } else {
+        send hlpForward set sensitive False
+        incr h_linkIndex
+    }
+    if {$h_linkIndex == [expr [llength $h_links] - 1]} {
+	send hlpForward set sensitive False
+    }
+    send hlpText gotoId $anchID
+    send hlpText retestAnchors
+}; send hlpText addCallback anchorSelected anchor
+
+
+# Callbacks to position forwards and backwards in link list.
+proc hlpForward args {
+    global h_links h_linkIndex
+    incr h_linkIndex
+    if {$h_linkIndex <= [llength $h_links]} {
+        set anchID [lindex $h_links $h_linkIndex] 
+        send hlpText gotoId $anchID
+        send hlpText retestAnchors
+	if {$h_linkIndex == [expr [llength $h_links] - 1]} {
+	    send hlpForward set sensitive False
+	    send hlpBack set sensitive True
+	} else {
+	    send hlpBack set sensitive True
+	}
+    } else {
+	incr h_linkIndex -1
+    }
+}; send hlpForward addCallback hlpForward
+
+proc hlpBack args {
+    global h_links h_linkIndex
+    incr h_linkIndex -1
+    if {$h_linkIndex >= 0} {
+        set anchID [lindex $h_links $h_linkIndex] 
+        send hlpText gotoId $anchID
+        send hlpText retestAnchors
+	if {$h_linkIndex == 0} { send hlpBack set sensitive False   }
+	if {$h_linkIndex >= 0} { send hlpForward set sensitive True }
+    } else {
+        incr h_linkIndex 1
+    }
+}; send hlpBack addCallback hlpBack
+
+proc hlpHome args {
+    global h_links h_linkIndex
+    set h_links { 0 }
+    set h_linkIndex  0
+    send hlpText gotoId 0
+    send hlpForward set sensitive False
+    send hlpBack set sensitive False
+}; send hlpHome addCallback hlpHome
+
+proc hlpTutorial args {
+    showHelp tutorial
+}; send hlpTutorial addCallback hlpTutorial
+send hlpTutorial unmap				;# NO TUTORIAL AT THE MOMENT
+
+proc showHelp {name args} {
+    anchorSelected widget cbtype event text #$name
+    send hlpShell map
+}
+
+proc hlpFind args {
+    set phrase  [send hfEntry get string]
+    set dir  forward
+    set case caseless
+
+    if { $phrase != "" } {
+	if { [send hfDir get on] }  { set dir backward  }
+	if { [send hfCase get on] } { set case caseSensitive  }
+	if {[send hlpText searchText $phrase start end $dir $case ] > 0} { 
+	    set elid   [lindex [lindex $start 0] 0]
+	    set id     [max 1 [expr $elid - 10] ]
+	    send hlpText gotoId $id
+	    send hlpText setSelection $start $end
+	} else {
+            send warnText set label "Search string not found."
+            send warning map
+	}
+    } else {
+        send warnText set label "Warning: No search phrase entered."
+        send warning map
+    }
+} ; foreach w { hfEntry hfFind } { send $w addCallback hlpFind }
+
+send hfClear addCallback { send hfEntry set string "" }
+
+send hlpDismiss addCallback "send hlpShell unmap"
+
+
+################################################################################
+#  Document source viewer procedures. 
+################################################################################
+
+proc srcOpen { args } { send doc_source map } 
+send srcDismiss addCallback "send doc_source unmap"
+
+
+################################################################################
+#  Define some TCL debug procedures 
+################################################################################
+
+proc tclOpen {} { send tclShell map }
+
+proc tclCommandClear { widget args } { 
+    send tclEntry set string "" 
+} ; send tclClear addCallback tclCommandClear
+
+proc tclCommandExecute { widget args } { \
+	send server [send tclEntry {get string}] 
+} ; send tclExecute addCallback tclCommandExecute
+
+proc tclCommand { widget mode command args } { 
+    send server $command 
+} ; send tclEntry addCallback tclCommand
+
+proc tclToggleLogging args {
+    global debug
+    if { [ send tclLogging get state] } { 
+        set debug 1 
+        send tclLogging set label "Disable Logging"
+    } else { 
+        set debug 0 
+        send tclLogging set label "Enable Logging"
+    }
+} ; tclToggleLogging
+send tclLogging addCallback tclToggleLogging
+
+send tclDismiss addCallback	"send tclShell unmap"
+
+# Connect the 'textout' parameter so it appends messages from
+# the client to the Tcl text window.
+proc tclLogMessages { param old new } {
+    global debug
+    if {$debug == 1} { send tclEntry append [format "%s\n" $new ] }
+} ; send textout addCallback tclLogMessages
+
+
+################################################################################
+# Warning dialog.  This pops up a dialog box with the given warning message.
+################################################################################
+
+proc warnOkay {widget args} {
+    global curpack curtask
+
+    set label [send warnText get label]
+    if {[string match "No*" $label ] == 1} {
+	set topic [send topicEntry get string]
+	set last [string last "." $topic ]
+        send topicEntry set string [ string range $topic 0 [ incr last -1 ] ]
+    }
+    send warning unmap
+}; send warnDismiss addCallback warnOkay
+
+
+# The parameter "alert" is used to forward alerts from the client.  The 
+# special 'dismiss' value can be used to shut down the alert from the
+# client, the special "pop" value pops the last history elements from the
+# stack (used in case of an error loading a file).
+
+proc setAlert {param old new} {
+    global HPkg HType HUrl HOpt HTask HFile htop hcurrent
+
+    if {$new == "dismiss"} {
+        send warning unmap
+    } elseif {$new == "pop"} {
+        catch {
+            unset HType($hcurrent)
+            unset HOpt($hcurrent)
+            unset HTask($hcurrent)
+            unset HPkg($hcurrent)
+            unset HUrl($hcurrent)
+            unset HFile($hcurrent)
+        }
+        incr hcurrent -1
+    } else {
+        send searchStatus set label ""
+        send warnText set label $new
+        send warning map
+    }
+}; send alert addCallback setAlert
+
+
diff --git a/pkg/system/help/xhelp/mkpkg b/pkg/system/help/xhelp/mkpkg
new file mode 100644
index 00000000..a12b2788
--- /dev/null
+++ b/pkg/system/help/xhelp/mkpkg
@@ -0,0 +1,28 @@
+# Make the GUI part of the HELP task.
+
+$checkout libpkg.a ../
+$update   libpkg.a
+$checkin  libpkg.a ../
+$exit
+
+zzdebug:
+	$omake	zzdebug.x
+	$link	zzdebug.o -l xtools
+	;
+
+libpkg.a:
+	xhelp.x		../help.h xhelp.h
+	xhcmds.x	../help.h xhelp.h
+	xhdir.x		xhelp.h <ctype.h> <diropen.h> <finfo.h>
+	xhfiles.x	../help.h xhelp.h <ctype.h> <fset.h> 
+	xhhelp.x	../help.h xhelp.h <error.h> <finfo.h> <fset.h> 
+	xhinit.x	xhelp.h
+	xhofile.x	xhelp.h <fset.h>
+	xhpkg.x		<error.h> <fset.h> ../helpdir.h ../help.h xhelp.h
+	xhprint.x	xhelp.h ../help.h <ctype.h> <error.h> <ttyset.h> 
+	xhqref.x	../help.h xhelp.h <ctype.h> <finfo.h> 
+	xhroot.x	xhelp.h
+	xhsave.x	../help.h xhelp.h <ctype.h> 
+	xhsearch.x	xhelp.h <ctype.h> <fset.h> 
+	xhsort.x	xhelp.h <ctype.h> 
+	;
diff --git a/pkg/system/help/xhelp/xhcmds.x b/pkg/system/help/xhelp/xhcmds.x
new file mode 100644
index 00000000..2c98af79
--- /dev/null
+++ b/pkg/system/help/xhelp/xhcmds.x
@@ -0,0 +1,185 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<syserr.h>
+include	"../help.h"
+include	"xhelp.h"
+
+
+# XHELP callback commands.
+define  XC_COMMANDS "|help|load|print|quit|search|files|directory|type|package|"
+define  CMD_HELP        1
+define  CMD_LOAD        2
+define  CMD_PRINT       3
+define  CMD_QUIT        4
+define  CMD_SEARCH      5
+define  CMD_FILES       6
+define  CMD_DIRECTORY   7
+define  CMD_TYPE   	8
+define  CMD_PACKAGE     9
+
+
+# XH_COMMAND_LOOP -- Process the GUI command loop.
+
+procedure xh_command_loop (xh)
+
+pointer	xh					#I task descriptor
+
+pointer	sp, cmd, name, pkg, pat, opt, dev
+real    x, y
+int     wcs, key, exact_match
+char	str[SZ_FNAME]
+
+bool	streq()
+int	strdic(), clgcur()
+int     xh_pkglist()
+
+begin
+	call smark (sp)
+	call salloc (name, SZ_FNAME, TY_CHAR)
+	call salloc (pkg, SZ_FNAME, TY_CHAR)
+	call salloc (opt, SZ_FNAME, TY_CHAR)
+	call salloc (dev, SZ_FNAME, TY_CHAR)
+	call salloc (cmd, SZ_FNAME, TY_CHAR)
+	call salloc (pat, SZ_FNAME, TY_CHAR)
+
+
+	# Enter the command loop.
+        while (clgcur ("coords", x, y, wcs, key, str, SZ_FNAME) != EOF) {
+
+	    # Skip any non-colon commands.
+	    if (key != ':') 			
+		next
+
+	    # Get the colon command string.
+	    call sscan (str)
+	        call gargwrd (Memc[cmd], SZ_FNAME)
+
+	    switch (strdic (Memc[cmd], Memc[cmd], SZ_FNAME, XC_COMMANDS)) {
+	    case CMD_HELP:
+	        call gargwrd (Memc[name], SZ_FNAME)
+
+	        # Get help on the requested topic, updates package list 
+	        # if necesary.
+	        if (streq(Memc[name],"Home")) {
+	    	    call xh_init (xh, NO, YES) 			
+	        } else {
+	            call gargwrd (Memc[pkg], SZ_FNAME)		# curpack
+		    call gargwrd (Memc[opt], SZ_FNAME)		# option
+		    call xh_cmd_help (xh, Memc[name], Memc[pkg], Memc[opt])
+		}
+
+	    case CMD_FILES:
+	        call gargwrd (Memc[name], SZ_FNAME)		# task name
+		call gargwrd (Memc[pkg], SZ_FNAME)		# parent package
+		call xh_files (xh, Memc[name], Memc[pkg])
+
+	    case CMD_LOAD:
+		# Load a requested page from the history.
+	        call gargwrd (Memc[name], SZ_FNAME)		# task name
+        	call gargwrd (Memc[pkg], SZ_FNAME)		# curpack
+        	call gargwrd (Memc[opt], SZ_FNAME)		# help option
+        	if (xh_pkglist (xh, Memc[name], HELPDB(xh), LIST(xh)) != 0)
+        	    call gmsg (XH_GP(xh), "pkglist", LIST(xh))
+        	call xh_help (xh, Memc[name], Memc[pkg], Memc[opt])
+
+	    case CMD_PRINT:
+		# Print the current results.
+	        call gargwrd (Memc[name], SZ_FNAME)		# task name
+        	call gargwrd (Memc[pkg], SZ_FNAME)		# curpack
+		call gargwrd (Memc[dev], SZ_FNAME)		# printer name
+		call xh_print_help (xh, Memc[name], Memc[pkg], Memc[dev])
+
+	    case CMD_QUIT:
+		# Quit the task.
+		break
+
+	    case CMD_SEARCH:
+		# Get the results of the keyword search.
+		call gargi (exact_match)
+	        call gargstr (Memc[pat], SZ_FNAME)
+		call xh_search (xh, exact_match, Memc[pat])
+
+	    case CMD_DIRECTORY:
+		# Process the directory browsing command.
+	        call gargwrd (Memc[opt], SZ_FNAME)
+		call xh_directory (xh, Memc[opt])
+
+	    case CMD_TYPE:
+		# Get the showtype value from the GUI
+		call gargi (XH_SHOWTYPE(xh))
+
+	    case CMD_PACKAGE:
+		# For the given item return the package in which it
+		# was found.  [DEBUG ROUTINE.]
+	        call gargwrd (Memc[name], SZ_FNAME)
+		call xh_pkgpath (xh, Memc[name], CURPACK(xh), Memc[pkg])
+		call printf ("%s => %s\n")
+		    call pargstr (Memc[name])
+		    call pargstr (Memc[pkg])
+		call flush(STDOUT)
+	    }
+	}
+
+	call sfree (sp)
+end
+
+
+# XH_CMD_HELP -- Process a help command.
+
+procedure xh_cmd_help (xh, topic, curpack, option)
+
+pointer	xh					# task descriptor
+char	topic[ARB]				# requested topic
+char	curpack[ARB]				# current package
+char	option[ARB]				# option (help|source|sysdoc)
+
+int	len
+
+bool    streq()
+int	strncmp()
+int	xh_pkgname(), xh_pkglist()
+
+begin
+	if (streq (option, "help")) {
+	    # No package name given, find one and load it.
+	    if (streq (curpack, "{}")) {
+		curpack[1] = EOS
+		len = 0
+	    	if (xh_pkgname (xh, topic, curpack) == OK)
+		    len = xh_pkglist (xh, curpack, HELPDB(xh), LIST(xh))
+
+		if (len != 0 && 
+		    strncmp(curpack, "root", 4) != 0 &&
+		    strncmp(curpack, "clpack", 6) != 0) {
+		        call gmsg (XH_GP(xh), "pkglist", LIST(xh))
+		        call strcpy (curpack, CURPACK(xh), SZ_FNAME)
+    		        call gmsg (XH_GP(xh), "curpack", curpack)
+			call gmsg (XH_GP(xh), "history", "package")
+		}
+	    }
+
+	    if (xh_pkglist (xh, topic, HELPDB(xh), LIST(xh)) != 0) {
+		# Got a package listing....
+		call gmsg (XH_GP(xh), "pkglist", LIST(xh))
+		call strcpy (topic, CURPACK(xh), SZ_FNAME)
+    		call gmsg (XH_GP(xh), "curpack", topic)
+	    }
+	}
+
+	if (streq (topic, CURPACK(xh))) {
+	    call gmsg (XH_GP(xh), "type", "package")
+	    call gmsg (XH_GP(xh), "curtask", topic)
+	    if (streq (option, "help"))
+		call xh_help (xh, "", CURPACK(xh), option)
+	    else
+		call xh_help (xh, topic, curpack, option)
+	    call strcpy (CURPACK(xh), CURTASK(xh), SZ_FNAME)
+
+	} else {
+	    call gmsg (XH_GP(xh), "type", "task")
+	    call gmsg (XH_GP(xh), "curtask", topic)
+	    call xh_help (xh, topic, CURPACK(xh), option)
+	    call strcpy (topic, CURTASK(xh), SZ_FNAME)
+	}
+	call gmsg (XH_GP(xh), "history", "append")
+end
diff --git a/pkg/system/help/xhelp/xhdir.x b/pkg/system/help/xhelp/xhdir.x
new file mode 100644
index 00000000..3c951afa
--- /dev/null
+++ b/pkg/system/help/xhelp/xhdir.x
@@ -0,0 +1,567 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include <diropen.h>
+include	<ctype.h>
+include	<finfo.h>
+
+include "xhelp.h"
+
+
+# Pattern matching definitions.
+define  PATCHARS        "*?[" 		
+
+# Browsing command dictionary.
+define	DIR_CMDS "|dirlist|loadfile|open|template|home|up|root|rescan|save|"
+define	DIRLIST		1		# get the directory listing
+define	LOADFILE	2		# load the requested file
+define	OPEN		3		# load the requested file
+define	TEMPLATE	4		# filename matching template
+define	HOME		5		# goto the user's home$
+define	UP		6		# go up one directory
+define	ROOT		7		# go to the root directory
+define	RESCAN		8		# rescan current directory
+define	SAVE		9		# save to the requested file
+
+
+# XH_DIRECTORY -- Process the directory browsing command.
+
+procedure xh_directory (xh, command)
+
+pointer	xh					#i task descriptor
+char	command[ARB]				#i command option
+
+pointer	sp, dir, file, pattern, path, fmt
+pointer task, pkg, opt, type
+int	ncmd, overwrite
+int	strdic(), strcmp(), envgets()
+
+begin
+	# Allocate working space and clear it.
+	call smark (sp)
+	call salloc (dir, SZ_PATHNAME, TY_CHAR) 	
+	call salloc (path, SZ_PATHNAME, TY_CHAR) 	
+	call salloc (file, SZ_FNAME, TY_CHAR)
+	call salloc (pattern, SZ_FNAME, TY_CHAR)
+	call salloc (fmt, SZ_FNAME, TY_CHAR)
+	call salloc (task, SZ_FNAME, TY_CHAR)
+	call salloc (pkg, SZ_FNAME, TY_CHAR)
+	call salloc (opt, SZ_FNAME, TY_CHAR)
+	call salloc (type, SZ_FNAME, TY_CHAR)
+
+	call aclrc (Memc[dir], SZ_FNAME)
+	call aclrc (Memc[path], SZ_FNAME)
+	call aclrc (Memc[file], SZ_FNAME)
+	call aclrc (Memc[fmt], SZ_FNAME)
+	call aclrc (Memc[pattern], SZ_FNAME)
+	call aclrc (Memc[task], SZ_FNAME)
+	call aclrc (Memc[pkg], SZ_FNAME)
+	call aclrc (Memc[opt], SZ_FNAME)
+	call aclrc (Memc[type], SZ_FNAME)
+
+	ncmd = strdic (command, command, SZ_LINE, DIR_CMDS)
+	switch (ncmd) {
+	    case DIRLIST:
+		call gargwrd (Memc[dir], SZ_PATHNAME) 	# get the dirname
+		if (strcmp ("../", Memc[dir]) == 0) {
+		    call xh_updir (xh)
+		} else {
+		    call sprintf (Memc[path], SZ_PATHNAME, "%s%s")
+		        call pargstr (CURDIR(xh))
+		        call pargstr (Memc[dir])
+		    call xh_set_curdir (xh, Memc[path])
+		    call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+	            call xh_selection (xh, CURDIR(xh))
+		}
+
+	    case LOADFILE:
+		call gargwrd (Memc[file], SZ_FNAME) 	# get the filename
+		call sprintf (Memc[path], SZ_PATHNAME, "%s%s")
+		    call pargstr (CURDIR(xh))
+		    call pargstr (Memc[file])
+		call xh_open_file (xh, "helpres", Memc[path], YES, YES)
+	        call xh_selection (xh, Memc[path])
+
+	    case OPEN:
+		call gargwrd (Memc[file], SZ_FNAME) 	# get the filename
+		call xh_ldfile (xh, Memc[file])
+
+	    case TEMPLATE:
+		call gargwrd (Memc[pattern], SZ_FNAME) 	# set the template
+		call xh_set_pattern (xh, Memc[pattern])
+		call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+
+	    case HOME:
+		if (envgets ("home", Memc[dir], SZ_PATHNAME) != EOF) {
+		    call xh_set_curdir (xh, Memc[dir])
+		    call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+	   	}
+
+	    case UP:
+		call xh_updir (xh)
+
+	    case ROOT:
+		call xh_set_curdir (xh, "/")
+		call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+
+	    case RESCAN:
+		call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+
+	    case SAVE:
+		call gargwrd (Memc[path], SZ_FNAME) 	# get the filename
+		call gargwrd (Memc[file], SZ_FNAME) 	# get the output fname
+		call gargi (overwrite) 			# get the overwrite flag
+		call gargwrd (Memc[fmt], SZ_FNAME) 	# get the save format
+		call xh_save_file (xh, Memc[path], Memc[file], Memc[fmt],
+		    overwrite)
+	}
+
+	call sfree (sp)
+end
+
+
+# XH_DIRLIST -- Given the directory name and a file template return the
+# directory contents.
+
+procedure xh_dirlist (xh, directory, pattern)
+
+pointer	xh					#i task descriptor
+char	directory[ARB]				#i directory to read
+char	pattern[ARB]				#i matching template
+
+pointer	sp, path, fname, patbuf
+pointer	dp, fp, ip, op, ep, sym
+bool	match_extension
+int	dd, n, patlen
+int	nfiles, ndirs, lastch
+
+pointer	stopen(), stenter()
+int	diropen(), xh_isdir(), strncmp(), stridxs()
+int	patmake(), patmatch(), strlen(), getline()
+
+begin
+	call smark (sp)
+	call salloc (path, SZ_PATHNAME, TY_CHAR)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+	call salloc (patbuf, SZ_LINE, TY_CHAR)
+
+	call aclrc (Memc[patbuf], SZ_LINE)
+
+	# If this isn't a directory just return silently.
+	if (xh_isdir (directory, Memc[path], SZ_PATHNAME) == 0) {
+	    call sfree (sp)
+	    return
+	}
+	
+	# Open the requested directory
+	dd = diropen (directory, PASS_HIDDEN_FILES)
+
+	# Set up the pattern matching code.  We recognize selecting all files
+	# with a particular extension as a special case, since this case is
+	# very common and can be done much more efficiently if we don't use
+	# the general pattern matching code.  If we have no pattern set the
+	# length to zero to indicate that everything will match.
+
+	if (pattern[1] == EOS) {
+	    patlen = 0
+	} else {
+	    match_extension = (strncmp (pattern, "*.", 2) == 0 &&
+		stridxs (PATCHARS, pattern[3]) <= 0)
+	    if (match_extension)
+		patlen = strlen (pattern)
+	    else {
+		# Convert file matching pattern into general pattern string.
+		Memc[fname] = '^'
+		op = fname + 1
+		lastch = 0
+		for (ip=1;  pattern[ip] != EOS;  ip=ip+1) {
+		    if (pattern[ip] == '*' && lastch != '?' && lastch != ']') {
+			Memc[op] = '?'
+			op = op + 1
+		    }
+		    lastch = pattern[ip]
+		    Memc[op] = lastch
+		    op = op + 1
+		}
+		Memc[op] = '$'
+		op = op + 1
+		Memc[op] = EOS
+			
+		# Compile the pattern.
+		patlen = patmake (pattern, Memc[patbuf], SZ_LINE)
+	    }
+	}
+
+	# Initialize counters.
+	ndirs  = 0
+	nfiles = 0
+	dp = NULL
+	fp = NULL
+
+	# Accumulate the contents into the directory and files lists.  We
+	# match files against the given template, all directories are
+	# matched regardless.
+	for (n=0; n != EOF; ) {
+	    n = getline (dd, Memc[fname])
+	    if (n < 1)
+		break
+	    n = n - 1
+	    Memc[fname+n] = EOS			# stomp the newline
+
+	    # See if this is a directory.
+	    call sprintf (Memc[path], SZ_PATHNAME, "%s%s")
+		call pargstr (CURDIR(xh))
+		call pargstr (Memc[fname])
+	    if (xh_isdir (Memc[path], Memc[path], SZ_PATHNAME) > 0) {
+		ndirs = ndirs + 1
+
+		# If this is the first directory initialize the symbol table.
+		if (ndirs == 1)
+		    dp = stopen ("dirlist", LEN_INDEX, LEN_STAB, SZ_SBUF)
+
+		# Enter the directory name into the symbol table.
+		call strcat ("/", Memc[fname], SZ_FNAME)
+		sym = stenter (dp, Memc[fname], strlen(Memc[fname])+1)
+
+	    } else {
+		# Check if the file matches the given pattern.
+		if (patlen > 0) {
+		    if (match_extension) {
+			if (n < patlen)
+			    next
+			ep = fname + n - 1
+			for (ip=patlen; ip > 2; ip=ip-1) {
+			    if (Memc[ep] != pattern[ip])
+				break
+			    ep = ep - 1
+			}
+			if (pattern[ip] != '.' || Memc[ep] != '.')
+			    next
+		    } else if (patmatch (Memc[fname], Memc[patbuf]) <= 0)
+			next
+		}
+
+		# We have a match.
+		nfiles = nfiles + 1
+
+		# If this is the first file initialize the symbol table.
+		if (nfiles == 1)
+		    fp = stopen ("filelist", LEN_INDEX, LEN_STAB, SZ_SBUF)
+
+		# Enter the directory name into the symbol table.
+		sym = stenter (fp, Memc[fname], strlen(Memc[fname])+1)
+	    }
+	}
+
+	# Send the results to the GUI.
+	call xh_putlist (xh, dp, "directory", "dirlist")
+	call xh_putlist (xh, fp, "directory", "filelist")
+
+	# Clean up.
+	if (dp != NULL)
+	    call stclose (dp)
+	if (fp != NULL)
+	    call stclose (fp)
+	call close (dd)
+	call sfree (sp)
+end
+
+
+# XH_PUTLIST -- Given the symtab for the directory contents construct a
+# list suitable for a message to the GUI.  The 'arg' parameter is passed
+# to indicate which type of list this is.
+
+procedure xh_putlist (xh, stp, param, arg)
+
+pointer	xh					#i task descriptor
+pointer	stp					#i symtab ptr for list
+char	param[ARB]				#i GUI param to notify
+char	arg[ARB]				#i GUI param arg
+
+pointer	sp, list, msg, sym, name, ip
+int	nchars
+
+pointer	sthead(), stnext(), stname()
+int	stsize(), gstrcpy(), strcmp(), strlen()
+
+begin
+	# Return if there is no symtab information.
+	if (stp == NULL) {
+	    call smark (sp)
+	    call salloc (msg, SZ_FNAME , TY_CHAR)
+	    call sprintf (Memc[msg], SZ_FNAME, "%s { }")
+	        call pargstr (arg)
+
+	    call gmsg (XH_GP(xh), param, Memc[msg]) 	# send it to the GUI
+	    call gflush (XH_GP(xh))
+	    call sfree (sp)
+	    return
+	}
+
+	# Allocate space for the list.
+	nchars =  stsize (stp) + 1
+
+	call smark (sp)
+	call salloc (list, nchars , TY_CHAR)
+	call aclrc (Memc[list], nchars)
+	ip = list
+
+	# Build the list from the symtab.
+        for (sym = sthead (stp); sym != NULL; sym = stnext (stp,sym)) {
+	    name = stname(stp,sym)
+	    if (strcmp (Memc[name], "./") != 0) {
+	        ip = ip + gstrcpy (Memc[name], Memc[ip], SZ_FNAME)
+	        ip = ip + gstrcpy (" ", Memc[ip], SZ_FNAME)
+	    }
+	}
+
+	# Sort the list.
+	call xh_sort_list (Memc[list])
+
+	# Allocate space for the message buffer.  The "+ 6" is space for
+	# the brackets around the list in the message created below.
+	nchars = nchars + strlen (arg) + 6
+	call salloc (msg, nchars, TY_CHAR)
+	call aclrc (Memc[msg], nchars)
+	ip = msg
+
+	# Begin the message by adding the arg and make a Tcl list of the
+	# contents.
+	call sprintf (Memc[msg], nchars, "%s { ")
+	    call pargstr (arg)
+	call strcat (Memc[list], Memc[msg], nchars)
+	call strcat (" }", Memc[msg], nchars)
+
+	# Finally, send it to the GUI.
+	call gmsg (XH_GP(xh), param, Memc[msg])
+	call gflush (XH_GP(xh))
+
+	call sfree (sp)
+end
+
+
+# XH_LDFILE -- Load the requested file.  If this is a file display it's
+# contents and update the browser with it's directory, otherwise if it's
+# a directory jump the browser to that directory.
+
+procedure xh_ldfile (xh, file)
+
+pointer	xh					#i task descriptor
+char	file[ARB]				#i requested file/dir
+
+pointer	sp, ip, dir, parent, path
+int	nchars
+int	access(), strlen(), xh_isdir()
+
+begin
+	call smark (sp)
+	call salloc (dir, SZ_PATHNAME, TY_CHAR)
+	call salloc (path, SZ_PATHNAME, TY_CHAR)
+	call salloc (parent, SZ_PATHNAME, TY_CHAR)
+
+	# Expand the current directory to a host path.
+	call fdirname (file, Memc[dir], SZ_PATHNAME)
+
+	if (xh_isdir (Memc[dir], Memc[path], SZ_PATHNAME) > 0) {
+	    # Set the curdir and load it's contents.
+	    call xh_set_curdir (xh, Memc[path])
+	    call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+	    call xh_selection (xh, Memc[path])
+	} 
+
+	if (access(file,0,0) == YES && 
+	    xh_isdir (file, Memc[path], SZ_PATHNAME) == 0) {
+	        # Work backwards to the parent '/', be sure to skip the trailing
+	        # backslash already in the dirname.
+	        ip = dir + strlen (Memc[dir]) - 2
+	        while (Memc[ip] != '/' && ip > dir) 
+	            ip = ip - 1
+
+	        nchars = ip - dir
+	        if (nchars > 0)
+	            call strcpy (Memc[dir], Memc[parent], nchars)
+	        else
+	            call strcpy ("/", Memc[parent], nchars)
+
+	        # Set the parent dir and load it's contents.
+	        call xh_set_curdir (xh, Memc[parent])
+	        call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+
+	        # Now load the file itself.
+	        call xh_open_file (xh, "helpres", file, YES, YES)
+	        call xh_selection (xh, file)
+	}
+
+	call sfree (sp)
+end
+
+
+# XH_UPDIR -- Go up to the parent directory and return contents.
+
+procedure xh_updir (xh)
+
+pointer	xh					#i task descriptor
+
+pointer	sp, ip, dir, parent
+int	nchars, strlen()
+
+begin
+	call smark (sp)
+	call salloc (dir, SZ_PATHNAME, TY_CHAR)
+	call salloc (parent, SZ_PATHNAME, TY_CHAR)
+
+	# Expand the current directory to a host path.
+	call fdirname (CURDIR(xh), Memc[dir], SZ_PATHNAME)
+
+	# Work backwards to the parent '/', be sure to skip the trailing
+	# backslash already in the dirname.
+	ip = dir + strlen (Memc[dir]) - 2
+	while (Memc[ip] != '/' && ip > dir) 
+	    ip = ip - 1
+
+	nchars = ip - dir
+	if (nchars > 0)
+	    call strcpy (Memc[dir], Memc[parent], nchars)
+	else
+	    call strcpy ("/", Memc[parent], nchars)
+
+	# Set the parent dir and load it's contents.
+	call xh_set_curdir (xh, Memc[parent])
+	call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+	call xh_selection (xh, CURDIR(xh))
+
+	call sfree (sp)
+end
+
+
+# XH_SET_CURDIR -- Set the filename matching template pattern.
+
+procedure xh_set_curdir (xh, dir)
+
+pointer	xh					#i task descriptor
+char	dir[ARB]				#i current directory
+
+pointer	sp, dirbuf
+int	strlen()
+
+begin
+	call smark (sp)
+	call salloc (dirbuf, SZ_PATHNAME, TY_CHAR)
+
+	call strcpy (dir, CURDIR(xh), SZ_PATHNAME)
+	if (dir[strlen(dir)] != '/')
+	    call strcat ("/", CURDIR(xh), SZ_PATHNAME)
+
+	call sprintf (Memc[dirbuf], SZ_PATHNAME, "curdir %s")
+	    call pargstr (CURDIR(xh))
+
+	call gmsg (XH_GP(xh), "directory", Memc[dirbuf])
+	call gflush (XH_GP(xh))
+
+	call sfree (sp)
+end
+
+
+# XH_SET_PATTERN -- Set the filename matching template pattern.
+
+procedure xh_set_pattern (xh, pattern)
+
+pointer	xh					#i task descriptor
+char	pattern[ARB]				#i template pattern
+
+pointer	sp, patbuf
+
+begin
+	call smark (sp)
+	call salloc (patbuf, SZ_FNAME, TY_CHAR)
+
+	call sprintf (Memc[patbuf], SZ_FNAME, "template %s")
+	    call pargstr (pattern)
+
+	call strcpy (pattern, PATTERN(xh), SZ_FNAME)
+	call gmsg (XH_GP(xh), "directory", Memc[patbuf])
+	call gflush (XH_GP(xh))
+
+	call sfree (sp)
+end
+
+
+# XH_SELECTION -- Set the selected filename.
+
+procedure xh_selection (xh, selection)
+
+pointer	xh					#i task descriptor
+char	selection[ARB]				#i selection
+
+pointer	sp, buf
+
+begin
+	call smark (sp)
+	call salloc (buf, SZ_FNAME, TY_CHAR)
+
+	call sprintf (Memc[buf], SZ_FNAME, "selection %s")
+	    call pargstr (selection)
+
+	call gmsg (XH_GP(xh), "directory", Memc[buf])
+	call gflush (XH_GP(xh))
+	call sfree (sp)
+end
+
+
+# XH_ISDIR -- Test whether the named file is a directory.  Check first to
+# see if it is a subdirectory of the current directory. If VFN is a directory,
+# return the OS pathname of the directory in pathname, and the number of
+# chars in the pathname as the function value.  Otherwise return 0.
+
+int procedure xh_isdir (vfn, pathname, maxch)
+
+char	vfn[ARB]		# name to be tested
+char	pathname[ARB]		# receives path of directory
+int	maxch			# max chars out
+
+bool	isdir
+pointer	sp, fname, op
+int	ip, fd, nchars, ch
+long	file_info[LEN_FINFO]
+int	finfo(), diropen(), gstrcpy(), strlen()
+
+begin
+	call smark (sp)
+	call salloc (fname, SZ_PATHNAME, TY_CHAR)
+
+	# Copy the VFN string, minus any whitespace on either end.
+	op = fname
+	for (ip=1;  vfn[ip] != EOS;  ip=ip+1) {
+	    ch = vfn[ip]
+	    if (!IS_WHITE (ch)) {
+		Memc[op] = ch
+		op = op + 1
+	    }
+	}
+	Memc[op] = EOS
+
+	isdir = false
+	if (finfo (Memc[fname], file_info) != ERR) {
+	    isdir = (FI_TYPE(file_info) == FI_DIRECTORY)
+
+	    if (isdir) {
+		call fdirname (Memc[fname], pathname, maxch)
+		nchars = strlen (pathname)
+	    }
+
+	} else {
+	    # If we get here, the VFN is the name of a new file.
+	    ifnoerr (fd = diropen (Memc[fname], 0)) {
+		call close (fd)
+		isdir = true
+	    }
+	    nchars = gstrcpy (Memc[fname], pathname, maxch)
+	}
+
+	call sfree (sp)
+	if (isdir)
+	    return (nchars)
+	else {
+	    pathname[1] = EOS
+	    return (0)
+	}
+end
diff --git a/pkg/system/help/xhelp/xhelp.h b/pkg/system/help/xhelp/xhelp.h
new file mode 100644
index 00000000..24bffc71
--- /dev/null
+++ b/pkg/system/help/xhelp/xhelp.h
@@ -0,0 +1,89 @@
+# XHELP.H -- Include file for the XHELP GUI task.
+
+# Help database header structure.  Stored at the beginning of a help
+# database file.  This information is taken from the help$helpdb.x source.
+
+define  LEN_HDBHEADER   14
+define  HDB_MAGICVAL    110104B
+
+define  HDB_MAGIC       Memi[$1]        # helpdb file type code
+define  HDB_RAW         Memi[$1+1]      # access compiled or raw database
+define  HDB_RHD         Memi[$1+2]      # if raw, HP of root help directory
+define  HDB_INDEX       Memi[$1+3]      # index of root help directory
+define  HDB_CRDATE      Meml[$1+4]      # creation date
+define  HDB_NENTRIES    Memi[$1+5]      # number of help directories in db
+define  HDB_MAXENTRIES  Memi[$1+6]      # maximum no. of help directories in db
+define  HDB_NMODULES    Memi[$1+7]      # count of the total number of modules
+define  HDB_INDEXOFFSET Meml[$1+8]      # file offset of index, chars
+define  HDB_INDEXPTR    Memi[$1+9]      # pointer to loaded index, ty_struct
+define  HDB_INDEXLEN    Memi[$1+10]     # length of index structure, su
+define  HDB_DATAOFFSET  Meml[$1+11]     # file offset of data area, chars
+define  HDB_DATAPTR     Memi[$1+12]     # pointer to loaded data area, ty_struct
+define  HDB_DATALEN     Memi[$1+13]     # length of data area, struct units
+
+# Index structure.  Identifies the contents of the database and tells where
+# they are stored.  There is one index entry for each help directory, i.e.,
+# for each package.
+
+define  LEN_HDBINDEX    34
+define  SZ_DBIKEY       63
+define  LEN_DBIDATA     2
+
+define  DBI_KEY         Memc[P2C($1)]   # entry name
+define  DBI_OFFSET      Memi[$1+32]     # offset of entry into data area, su
+define  DBI_MTIME       Meml[$1+33]     # modification date of entry
+
+define  MAX_ENTRIES     100             # initial max db entries
+define  INC_ENTRIES     50              # increment if overflow
+define  MAX_DEPTH       20              # max nesting of packages
+define  MAX_MENUSIZE    500             # max modules in a table
+define  MAX_NAMELEN     20              # max chars in a module name in table
+
+
+# XHELP Macro definitions.
+define	SZ_HELPLIST	20480
+define	SZ_XHELPSTRUCT	45
+
+define	XH_GP		Memi[$1]	# graphics descriptor
+define	XH_LPTR		Memi[$1+1]	# ptr for pkg list
+define	XH_TEMPLATE	Memi[$1+2]	# initial help topic
+define	XH_OPTION	Memi[$1+3]	# help option
+define	XH_PRINTER	Memi[$1+4]	# printer name
+define	XH_CURTASK	Memi[$1+5]	# current task name
+define	XH_CURPACK	Memi[$1+6]	# current package name
+define	XH_QUICKREF	Memi[$1+7]	# quick-reference filen
+define	XH_HOMEPAGE	Memi[$1+8]	# startup page
+define	XH_CURDIR	Memi[$1+9]	# current directory
+define	XH_PATTERN	Memi[$1+10]	# current filename template
+define	XH_HELPDB	Memi[$1+11]	# help database string
+define	XH_SHOWTYPE	Memi[$1+12]	# indicate packages in list
+define	XH_STP		Memi[$1+13]	# package list symtab ptr
+
+# Helpful macros
+define	LIST		Memc[XH_LPTR($1)]
+define	TEMPLATE	Memc[XH_TEMPLATE($1)]
+define	OPTION		Memc[XH_OPTION($1)]
+define	PRINTER		Memc[XH_PRINTER($1)]
+define	CURTASK		Memc[XH_CURTASK($1)]
+define	CURPACK		Memc[XH_CURPACK($1)]
+define	QUICKREF	Memc[XH_QUICKREF($1)]
+define	HOMEPAGE	Memc[XH_HOMEPAGE($1)]
+define	CURDIR		Memc[XH_CURDIR($1)]
+define	PATTERN		Memc[XH_PATTERN($1)]
+define	HELPDB		Memc[XH_HELPDB($1)]
+
+define	WIDE_PAGE	100		# needed by the print routines
+define	SZ_DDSTR	256
+
+
+# Filenames.
+define  HELP      "lib$scr/help.html"	# default help file
+define  PKGFILE   "uparm$help.pkgs"     # default package list symtab
+define  QREFFILE  "uparm$quick.ref"     # default references file
+
+
+# Symbol table definitions.
+define  LEN_INDEX       10              # length of symtab index
+define  LEN_STAB        32              # initial length of symtab
+define  SZ_SBUF         32              # initial size of symtab string buffer
+
diff --git a/pkg/system/help/xhelp/xhelp.x b/pkg/system/help/xhelp/xhelp.x
new file mode 100644
index 00000000..95a2f250
--- /dev/null
+++ b/pkg/system/help/xhelp/xhelp.x
@@ -0,0 +1,167 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<syserr.h>
+include	"../help.h"
+include	"xhelp.h"
+
+
+# XHELP -- The main task procedure.  XHELP is a GUI client program for
+# browsing the IRAF help system.  As much as possible it uses the existing
+# help database code but provides a friendlier interface, allowing users to
+# browse packages for help pages in the same way they would browse packages
+# in the CL.  It provides an HTML converter for LROFF sources for better
+# presentation in the GUI, as well as Postscript generation for better
+# looking hardcopy.  XHelp acts as a server for the help system, merely
+# returning any output that the GUI has requested.  Navigation is done in
+# the GUI code, this program maintains just the state of the last page
+# returned and knows nothing about how it got there.  See the xhelp.hlp for
+# detailed documentation.
+
+procedure xhelp (topic)
+
+char	topic[ARB]					#I help template
+
+pointer	xh
+char	uifname[SZ_FNAME]
+int     search, template
+
+pointer	xh_open(), gopenui()
+bool    clgetb()
+int     btoi()
+
+begin
+	# Open structure and allocate pointers.
+	xh = xh_open ()
+        call strcpy (topic, TEMPLATE(xh), SZ_FNAME)
+
+        # Load the task parameters.
+        call clgstr ("option",  OPTION(xh),  SZ_FNAME)
+        call clgstr ("printer", PRINTER(xh), SZ_FNAME)
+        call clgstr ("quickref", QUICKREF(xh), SZ_FNAME)
+	XH_SHOWTYPE(xh) = btoi (clgetb("showtype"))
+	search = btoi (clgetb("search"))
+	template = btoi (clgetb("file_template"))
+
+        # Fetch the name of the help database.
+	call xh_ghelpdb (xh)
+
+        # Open the GUI.
+        call clgstr ("uifname", uifname, SZ_FNAME)
+        XH_GP(xh) = gopenui ("stdgraph", NEW_FILE, uifname, STDGRAPH)
+	call gflush (XH_GP(xh))
+
+	# Initialize the task and send topic list to the GUI.
+	call xh_init (xh, template, search)
+
+	# Initialize the task and send topic list to the GUI.
+	call xh_command_loop (xh)
+
+	# Clean up.
+	call gclose (XH_GP(xh))
+	call xh_close (xh)
+end
+
+
+# XH_OPEN -- Open and allocate the XHELP task structure.
+
+pointer procedure xh_open ()
+
+pointer	xh					# task descriptor
+errchk	calloc
+
+begin
+        iferr (call calloc (xh, SZ_XHELPSTRUCT, TY_STRUCT))
+            call error (0, "Error opening task structure.")
+
+	iferr {
+            call calloc (XH_LPTR(xh), SZ_HELPLIST, TY_CHAR)
+            call calloc (XH_TEMPLATE(xh), SZ_FNAME, TY_CHAR)
+            call calloc (XH_OPTION(xh), SZ_FNAME, TY_CHAR)
+            call calloc (XH_PRINTER(xh), SZ_FNAME, TY_CHAR)
+            call calloc (XH_CURTASK(xh), SZ_FNAME, TY_CHAR)
+            call calloc (XH_CURPACK(xh), SZ_FNAME, TY_CHAR)
+            call calloc (XH_QUICKREF(xh), SZ_FNAME, TY_CHAR)
+            call calloc (XH_HOMEPAGE(xh), SZ_FNAME, TY_CHAR)
+            call calloc (XH_CURDIR(xh), SZ_PATHNAME, TY_CHAR)
+            call calloc (XH_PATTERN(xh), SZ_FNAME, TY_CHAR)
+            call calloc (XH_HELPDB(xh), SZ_HELPDB, TY_CHAR)
+	} then
+	    call error (0, "Error allocating structure pointers.")
+
+	return (xh)
+end
+
+
+# XH_CLOSE -- Close the XHELP task structure.
+
+procedure xh_close (xh)
+
+pointer	xh					# task descriptor
+
+begin
+        call mfree (XH_TEMPLATE(xh), TY_CHAR)
+        call mfree (XH_OPTION(xh), TY_CHAR)
+        call mfree (XH_PRINTER(xh), TY_CHAR)
+        call mfree (XH_CURTASK(xh), TY_CHAR)
+        call mfree (XH_CURPACK(xh), TY_CHAR)
+        call mfree (XH_QUICKREF(xh), TY_CHAR)
+        call mfree (XH_HOMEPAGE(xh), TY_CHAR)
+        call mfree (XH_CURDIR(xh), TY_CHAR)
+        call mfree (XH_PATTERN(xh), TY_CHAR)
+        call mfree (XH_HELPDB(xh), TY_CHAR)
+        call mfree (XH_LPTR(xh),  TY_CHAR)
+
+        call mfree (xh, TY_STRUCT)
+end
+
+
+# XH_GHELPDB -- Fetch the name of the help database, i.e., "helpdb",
+# "helpdir",  or the name of a file.   If the helpdb string is a list check
+# for the existance of each file in the list to ensure the final list
+# contains only valid help databases.
+
+procedure xh_ghelpdb (xh)
+
+pointer	xh					# task descriptor
+
+pointer	sp, hdb, hdbstr, name
+int	list
+int	fntopnb(), fntgfnb()
+int	access(), envgets()
+bool	streq()
+
+begin
+	call smark (sp)
+	call salloc (name, SZ_FNAME, TY_CHAR)
+	call salloc (hdb, SZ_HELPDB, TY_CHAR)
+	call salloc (hdbstr, SZ_HELPDB, TY_CHAR)
+
+	# Clear the working memory.
+	call aclrc (Memc[name], SZ_FNAME)
+	call aclrc (Memc[hdb], SZ_HELPDB)
+	call aclrc (Memc[hdbstr], SZ_HELPDB)
+
+	# Get the parameter value.
+        call clgstr ("helpdb", Memc[hdbstr], SZ_HELPDB)
+        if (streq (Memc[hdbstr], "helpdb"))
+            if (envgets ("helpdb", Memc[hdbstr], SZ_HELPDB) <= 0)
+                call syserrs (SYS_ENVNF, "helpdb")
+
+	# Open the list.
+	list = fntopnb (Memc[hdbstr], YES)
+
+	# Copy each of the existing files in the list to the output database
+	# string to be used by the task.
+	while (fntgfnb(list, Memc[name], SZ_FNAME) != EOF) {
+	    if (access (Memc[name], 0, 0) == YES) {
+		if (Memc[hdb] != EOS)
+		    call strcat (",", Memc[hdb], SZ_HELPDB)
+		call strcat (Memc[name], Memc[hdb], SZ_HELPDB)
+	    }
+	}
+	call strcpy (Memc[hdb], HELPDB(xh), SZ_HELPDB)
+
+	# Clean up.
+	call fntclsb (list)
+	call sfree (sp)
+end
diff --git a/pkg/system/help/xhelp/xhfiles.x b/pkg/system/help/xhelp/xhfiles.x
new file mode 100644
index 00000000..53597eda
--- /dev/null
+++ b/pkg/system/help/xhelp/xhfiles.x
@@ -0,0 +1,89 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<fset.h>
+include <ctype.h>
+include "../help.h"
+include	"xhelp.h"
+
+
+# XH_FILES -- Get the files associated with the requested help topic,
+# i.e. do a "help <topic> opt=files" request.
+
+procedure xh_files (xh, topic, curpack)
+
+pointer xh                                      #i task struct pointer
+char    topic[ARB]                           	#i help topic
+char    curpack[ARB]                           	#i current package
+
+pointer sp, bp, buf
+pointer	opt, val, line
+int	fd
+long	fsize
+char	fname[SZ_FNAME]
+
+long	fstatl()
+int	getline(), access(), open(), stridxs(), strlen()
+
+begin
+	if (topic[1] == EOS && curpack[1] == EOS)
+	    return
+
+	call smark (sp)
+	call salloc (line, SZ_LINE, TY_CHAR)
+	call salloc (buf, SZ_FNAME, TY_CHAR)
+	call salloc (opt, SZ_FNAME, TY_CHAR)
+	call salloc (val, SZ_FNAME, TY_CHAR)
+
+	# Get a temp file name.
+	call mktemp ("tmp$xhelpi", fname, SZ_FNAME)
+
+	# Open a temp file with the help information found.
+	fd = open (fname, NEW_FILE, TEXT_FILE)
+	call xh_get_help (fd, topic, curpack, "", HF_HTML, HELPDB(xh),
+	    "all", "files")
+	call close (fd)
+
+	# Open the results file for reading.
+	fd = open (fname, READ_ONLY, TEXT_FILE)
+	fsize = fstatl (fd, F_FILESIZE)
+
+	# If we got a result parse the lines for "opt=file" information.
+	if (fsize != 0) {
+
+            while (getline (fd, Memc[line]) != EOF) {
+
+		# Stomp the newline.
+	  	Memc[line+strlen(Memc[line])-1] = EOS
+
+		# Extract the option type and filename.
+		if (stridxs ("=", Memc[line]) > 0) {
+		    for (bp=line; IS_WHITE(Memc[bp]); bp=bp+1)
+			;
+		    call strcpy (Memc[bp], Memc[opt], 3)
+		    for (; Memc[bp] != '='; bp=bp+1)
+			;
+		    call strcpy (Memc[bp+1], Memc[val], SZ_FNAME)
+
+		    # See if the file exists.
+		    call sprintf (Memc[buf], SZ_FNAME, "%s %s \0")
+		        call pargstr (Memc[opt])
+		        call pargstr (Memc[val])
+		    if (access (Memc[val],0,0) == YES)
+		        call strcat (" 0", Memc[buf], SZ_FNAME)
+		    else
+		        call strcat (" 1", Memc[buf], SZ_FNAME)
+		    call gmsg (XH_GP(xh), "helpfiles", Memc[buf])
+
+		} else if (stridxs (":", Memc[line]) > 0) {
+		    call xh_pkgpath (xh, topic, curpack, Memc[line])
+		    call sprintf (Memc[buf], SZ_FNAME, "file %s")
+		        call pargstr (Memc[line])
+		    call gmsg (XH_GP(xh), "helpfiles", Memc[buf])
+		}
+	    }
+	}
+
+	call close (fd)		# clean up 
+	call delete (fname)
+	call sfree (sp)
+end
diff --git a/pkg/system/help/xhelp/xhhelp.x b/pkg/system/help/xhelp/xhhelp.x
new file mode 100644
index 00000000..708a00ce
--- /dev/null
+++ b/pkg/system/help/xhelp/xhhelp.x
@@ -0,0 +1,276 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<fset.h>
+include	<finfo.h>
+include	<error.h>
+include	"../help.h"
+include	"xhelp.h"
+
+
+# XH_HELP -- Get the requested help topic and send the result to the GUI.
+
+procedure xh_help (xh, topic, curpack, opt)
+
+pointer xh                                      #i task struct pointer
+char    topic[ARB]                           	#i help topic
+char    curpack[ARB]                           	#i current package
+char    opt[ARB]                           	#i help option
+
+pointer helpstr
+int	ip, fdi
+long	fsize
+char	ch, fname[SZ_FNAME], err[SZ_LINE]
+
+long	fstatl()
+char	getc()
+int	open()
+bool	strne()
+
+begin
+	# Get a temp file name.
+	call mktemp ("tmp$xhelpi", fname, SZ_FNAME)
+
+	# Open a temp file with the help information found.
+	fdi = open (fname, NEW_FILE, TEXT_FILE)
+	call xh_get_help (fdi, topic, curpack, "", HF_HTML, HELPDB(xh),
+	    "all", opt)
+	call close (fdi)
+
+	# Open the results file for reading.
+	fdi = open (fname, READ_ONLY, TEXT_FILE)
+	fsize = fstatl (fdi, F_FILESIZE)
+
+	# If no results try using the topic name as a curpack param.
+	if (fsize == 0) {
+	    call close (fdi)		# clean up from before
+	    call delete (fname)
+
+	    # Open a temp file with the help information found.
+	    fdi = open (fname, NEW_FILE, TEXT_FILE)
+	    call xh_get_help (fdi, topic, topic, "", HF_HTML, HELPDB(xh),
+		"all", opt)
+	    call close (fdi)
+
+	    # Open the results file for reading.
+	    fdi = open (fname, READ_ONLY, TEXT_FILE)
+	    fsize = fstatl (fdi, F_FILESIZE)
+
+	    # If we still have nothing then punt...
+	    if (fsize == 0 && topic[1] != EOS) {
+	        if (strne (opt, "help")) {
+	            call sprintf (err, SZ_LINE,
+		        "No '%s' option help available\n for `%s'.")
+		            call pargstr (opt)
+		            call pargstr (topic)
+	        } else {
+	            call sprintf (err, SZ_LINE, "No help available for\n`%s'.")
+		        call pargstr (topic)
+	        }
+	        call gmsg (XH_GP(xh), "alert", err)
+	        call close (fdi)
+	        call delete (fname)
+	        return
+	    }
+	}
+
+	# Now filter the file to escape the curly braces so they pass thru
+	# to the Tcl cleanly.  Put the result in the string sent to the GUI.
+	call calloc (helpstr, fsize + SZ_LINE, TY_CHAR)
+	ip = helpstr
+	repeat {
+	    ch = getc (fdi, ch)
+	    if (ch == '{' || ch == '}') {
+	        Memc[ip] = '\\'
+		ip = ip + 1
+	    }
+	    Memc[ip] = ch
+	    ip = ip + 1
+	} until (ch == EOF)
+	Memc[ip-1] = EOS
+
+	# Clean up.
+	call close (fdi)
+	call delete (fname)
+
+	# Send the help text to the GUI who will display it.
+	call gmsg (XH_GP(xh), "helpres", Memc[helpstr])
+	call mfree (helpstr, TY_CHAR)
+end
+
+
+# XH_GET_HELP -- The main work procedure, i.e. a rip-off of the t_help()
+# procedure.  Decode the option string, set up the control structure, and 
+# finally call process_template to expand the module template and process
+# the help text for each module.  The output is written to a temp file
+# opened by the calling procedure which may optionally sort it or display
+# it as is.
+
+procedure xh_get_help (fd, topic, curpack, file, format, helpdb, section, opt)
+
+int	fd					#i file descriptor of result
+char	topic[ARB]				#i topic
+char    curpack[ARB]                           	#i current package
+char    file[ARB]                           	#i file template
+int	format					#i output format (text|html|ps)
+char	helpdb[ARB]				#i help database
+char	section[ARB]				#i section on which to get help
+char	opt[ARB]				#i type of help
+
+int	list
+long	fi[LEN_FINFO], db_ctime
+pointer	sp, ctrl, optn, db, fname
+
+long	clktime()
+pointer	hdb_open()
+bool	strne(), streq()
+int	stridxs(), finfo(), fntopnb(), fntgfnb(), get_option()
+
+errchk	hdb_open
+
+data	db_ctime /0/
+define	forms_ 91
+
+begin
+	call smark (sp)
+	call salloc (ctrl, LEN_CTRLSTRUCT, TY_STRUCT)
+	call salloc (optn, SZ_FNAME, TY_CHAR)
+	call salloc (fname, SZ_PATHNAME, TY_CHAR)
+
+	# If we were called without any arguments, do not query for the
+	# template, just set it to null and help will be given for the
+	# current package.
+
+	call aclri (Memi[ctrl], LEN_CTRLSTRUCT)
+	if (topic[1] == EOS) {
+	    if (file[1] == EOS) {
+	        H_OPTION(ctrl) = O_MENU
+	        H_TEMPLATE(ctrl) = EOS
+	        H_PARNAME(ctrl) = EOS
+	        H_SECNAME(ctrl) = EOS
+	    } else
+	        call strcpy (file, H_TEMPLATE(ctrl), SZ_LINE)
+	} else {
+	    call strcpy (topic, H_TEMPLATE(ctrl), SZ_LINE)
+	}
+
+
+	# Check to see if any of the files in the list are newer than the
+	# time of the last hdb_open.  The first time the process runs we open
+	# and read in the database.  The database remains in memory between
+	# calls to HELP, provided the process does not shutdown, provided
+	# the name of the database to be used does not change, and provided
+	# a new help database is not created.
+
+	if (db_ctime > 0) {
+	    list = fntopnb (helpdb, YES)
+	    while (fntgfnb (list, Memc[fname], SZ_PATHNAME) != EOF) {
+                if (finfo (Memc[fname], fi) != ERR) {
+		    if (db != NULL && FI_CTIME(fi) > db_ctime) {
+		        call hdb_close (db)
+		        db = NULL
+		        break
+		    }
+		}
+	    }
+	    call fntclsb (list)
+	} else
+	    db = NULL
+
+	# Reopen the help database if in-core copy is out of date.
+	if (db == NULL) {
+	    db = hdb_open (helpdb)
+	    db_ctime = clktime (long(0))
+	}
+
+	# Fetch the value of the ALL switch.  This determines whether help
+	# will stop after processing the first module matching the template,
+	# or process all modules in the database which match the template.
+	# Explicit use of a pattern matching character anywhere in the template
+	# enable allmodoules.
+
+	if (stridxs ("*?[],", H_TEMPLATE(ctrl)) > 0)
+	    H_ALLMODULES(ctrl) = YES
+	else
+	    H_ALLMODULES(ctrl) = NO
+	
+	# If the FILTER_INPUT flag is set, only part of the input text will be
+	# processed.  Filtering is only done if printing a single section or
+	# parameter.
+
+	H_FILTER_INPUT(ctrl) = NO
+
+	# Determine whether or not text for a single section or parameter
+	# is to be output.  If the value of one of these strings is "all",
+	# all sections or all parameters are to be output.  If the "all"
+	# default is in effect, null the string as a flag to lower level
+	# code that all help text is to be processed.
+
+	if (H_OPTION(ctrl) == NULL) {
+	    call strcpy (section, H_SECNAME(ctrl), SZ_SECNAME)
+	    if (streq (H_SECNAME(ctrl), "all")) {
+		H_SECNAME(ctrl) = EOS
+		H_PARNAME(ctrl) = EOS
+	    }
+	    if (H_SECNAME(ctrl) != EOS || H_PARNAME(ctrl) != EOS)
+		H_FILTER_INPUT(ctrl) = YES
+	}
+
+	# Fetch and decode option string; abbreviations are permitted.
+	if (H_OPTION(ctrl) != O_MENU) {
+	    call strcpy (opt, Memc[optn], SZ_FNAME)
+	    call strlwr (Memc[optn])
+	    iferr (H_OPTION(ctrl) = get_option (Memc[optn])) {
+		H_OPTION(ctrl) = O_HELP
+		call erract (EA_WARN)
+	    }
+	}
+
+forms_
+	# Pause between screens of output text only if the standard output
+	# is not redirected, and if enabled by the user.
+
+	H_IN(ctrl) = ERR
+	H_OUT(ctrl) = fd
+	H_NLINES(ctrl) = -1
+	H_STATE(ctrl) = BOF
+	H_EOF(ctrl) = NO
+	H_QUIT(ctrl) = NO
+
+	# If the standard output is not redirected, i.e., if writing to the
+	# terminal, determine whether or not output is to be paginated (pause
+	# between pages).  If output is redirected, the pagination flag
+	# and help option controls whether or not manpage style output is
+	# enabled.  Manpage output formatting is desirable only when formatting
+	# help text or printing named files.
+
+	H_RAWOUT(ctrl) = YES
+	H_MANPAGE(ctrl) = NO
+	H_PAGINATE(ctrl) = NO
+	H_SOFLAG(ctrl) = NO
+	H_FORMAT(ctrl) = format
+
+	# We don't produce output to a screen so shut off the tty.
+
+	H_TTY(ctrl) = NULL
+
+	# Set left and right margins for output text format.
+
+	H_LMARGIN(ctrl) = 1
+	H_RMARGIN(ctrl) = 72
+
+	# Copy the current package to the control struct.
+	if (strne(curpack,"Home") && strne(curpack,""))
+            call strcpy (curpack, H_CURPACK(ctrl), SZ_CURPACK)
+
+	# Initialization is completed, control structure is completed.
+	# Format and output the help text.  If we have a module name template
+	# process the template against the help database, otherwise work
+	# directly out of the named files.
+
+	if (file[1] == EOS)
+            call do_module_template (db, H_TEMPLATE(ctrl), ctrl)
+	else
+            call do_file_template (H_TEMPLATE(ctrl), ctrl)
+
+	call sfree (sp)
+end
diff --git a/pkg/system/help/xhelp/xhinit.x b/pkg/system/help/xhelp/xhinit.x
new file mode 100644
index 00000000..41ff5b65
--- /dev/null
+++ b/pkg/system/help/xhelp/xhinit.x
@@ -0,0 +1,77 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"xhelp.h"
+
+
+# XH_INIT -- Initialize the task and the GUI.
+
+procedure xh_init (xh, file_template, search)
+
+pointer	xh				#i task struct pointer
+int	file_template			#i is topic a file template?
+int	search				#i doing a search?
+
+char	curdir[SZ_FNAME]
+int	fd
+
+pointer	strestore()
+int	envgets(), open()
+bool	streq()
+errchk	open
+
+begin
+	# Update the quickref file.
+	call xh_updt_quickref (xh)
+
+	# If starting up with a search, get that information first.
+ 	if (search == YES && TEMPLATE(xh) != EOS)  {
+	    call xh_search (xh, NO, TEMPLATE(xh))
+	    call strcpy ("", TEMPLATE(xh), SZ_FNAME)
+	    call strcpy ("", LIST(xh), SZ_LINE)
+	} 
+
+	# Create the root package and send the results to the GUI.
+	call xh_root_pkg (xh)
+	call gmsg (XH_GP(xh), "pkglist", LIST(xh))
+
+	# Initialize the package list symtab.
+	iferr (fd = open (PKGFILE, READ_ONLY, BINARY_FILE)) 
+	    call error (0, "Cannot open package list symtab")
+	XH_STP(xh) = strestore (fd)
+	call close (fd)
+	call gmsgi (XH_GP(xh), "showtype", XH_SHOWTYPE(xh))
+
+
+	if (TEMPLATE(xh) != EOS && file_template == NO) {
+	    # If we're given an initial help topic, get the page and load it.
+	    #call xh_help (xh, TEMPLATE(xh), TEMPLATE(xh), OPTION(xh))
+	    call xh_cmd_help (xh, TEMPLATE(xh), "{}", OPTION(xh))
+	    call strcpy ("", TEMPLATE(xh), SZ_FNAME)
+
+	} else if (TEMPLATE(xh) != EOS && file_template == YES) {
+	    # Load a user defined page.
+	    call xh_open_file (xh, "helpres", TEMPLATE(xh), YES, YES)
+	    call strcpy ("", TEMPLATE(xh), SZ_FNAME)
+
+	} else {
+	    # Load either a user defined homepage or the task help.
+	    call clgstr ("home", HOMEPAGE(xh), SZ_FNAME)
+	    if (streq ("", HOMEPAGE(xh)))
+	        call strcpy (HELP, HOMEPAGE(xh), SZ_FNAME)
+	    call xh_open_file (xh, "helpres", HOMEPAGE(xh), NO, YES)
+	}
+
+	# Set the printer to be used.
+	call gmsg (XH_GP(xh), "printer", PRINTER(xh))
+
+	# Initialize the online help doc.
+	call xh_open_file (xh, "help", HELP, NO, YES)
+
+	# Initialize the file browsing parameters.  Since we can't
+	# get the current directory for the session use home$.
+	if (envgets ("home", curdir, SZ_FNAME) != EOF) {
+	    call xh_set_pattern (xh, "*")
+	    call xh_set_curdir (xh, curdir)
+	    call xh_dirlist (xh, CURDIR(xh), PATTERN(xh))
+	}
+end
diff --git a/pkg/system/help/xhelp/xhofile.x b/pkg/system/help/xhelp/xhofile.x
new file mode 100644
index 00000000..c12ea7d3
--- /dev/null
+++ b/pkg/system/help/xhelp/xhofile.x
@@ -0,0 +1,188 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<fset.h>
+include	"xhelp.h"
+
+
+# XH_OPEN_FILE -- Open the named file and send it to the GUI.  If this is
+# a help document (i.e. it contains a ".help" block) we first convert it
+# to HTML, otherwise send it as is.
+
+procedure xh_open_file (xh, parameter, filename, check_for_help, warn)
+
+pointer	xh					# task descriptor
+char	parameter[ARB]				# GUI parameter to notify
+char	filename[ARB]				# file to open
+int	check_for_help				# check file for help block?
+int	warn					# warn if not present?
+
+pointer	sp, ip, buf, out, text
+int	fdi, fdo
+long	fsize
+bool	has_help
+
+int	access(), open(), getline()
+int	strmatch(), gstrcpy()
+long	fstatl()
+errchk	open
+
+define	err_	99
+
+begin
+	call smark (sp)
+	call salloc (buf, SZ_LINE, TY_CHAR)
+	call salloc (out, SZ_FNAME, TY_CHAR)
+
+	# Make sure the file exists.
+        if (access (filename, 0, 0) == NO) {
+	    if (warn == YES) {
+                call sprintf (Memc[buf], SZ_LINE, "File does not exist:\n`%s'.")
+                    call pargstr (filename)
+		call gmsg (XH_GP(xh), "alert", Memc[buf])
+	    }
+	    goto err_
+	} else if (access (filename, 0, BINARY_FILE) == YES) {
+	    if (warn == YES) {
+                call sprintf (Memc[buf], SZ_LINE, 
+		    "Attempt to load binary file:\n`%s'.")
+                        call pargstr (filename)
+		call gmsg (XH_GP(xh), "alert", "pop")
+		call gmsg (XH_GP(xh), "alert", Memc[buf])
+	    }
+	    goto err_
+	}
+
+	# If we're told not to look for help simply open the file and send
+	# it to the GUI (e.g. used for homepage and online help).
+	if (check_for_help == NO) {
+	    call xh_load_file (xh, parameter, filename)
+	    call sfree (sp)
+	    return
+	}
+
+	# Open the file.
+	iferr (fdi = open (filename, READ_ONLY, TEXT_FILE)) {
+	    if (warn == YES) {
+                call sprintf (Memc[buf], SZ_LINE, "Cannot open file\n`%s'.")
+                    call pargstr (filename)
+		call gmsg (XH_GP(xh), "alert", Memc[buf])
+	    }
+	    goto err_
+	}
+
+	# Allocate an array the length of the file, if this isn't a help file
+	# we use this as the message buffer and send it to the GUI.
+	fsize = fstatl (fdi, F_FILESIZE)
+	call salloc (text, fsize+1, TY_CHAR)
+	call aclrc (Memc[text], fsize+1)
+
+	# See whether this is a help file
+	has_help = FALSE
+	ip = text
+	while (getline (fdi, Memc[buf]) != EOF) {
+            if (strmatch (Memc[buf], "^.help") > 0) {
+		has_help = TRUE
+		break
+	    }
+	    ip = ip + gstrcpy (Memc[buf], Memc[ip], SZ_LINE)
+	}
+	Memc[ip] = EOS
+
+
+	# If the file was found to have a .help block we're positioned at
+	# the beginning of the block.  Convert the remainder to an HTML
+	# temp file and send that to the GUI, otherwise we already have the
+	# contents of the file in the text buffer so send that.
+	if (has_help) {
+	    # Create an output filename and open it for writing.
+	    call mktemp ("tmp$xhelpi", Memc[out], SZ_FNAME)
+	    fdo = open (Memc[out], NEW_FILE, TEXT_FILE)
+
+	    # Convert the remainder to HTML and send it to the GUI.
+	    if (fdo != ERR) {
+        	call lroff2html (fdi, fdo, filename, "", "", "", "")
+		call close (fdo)
+
+		call xh_load_file (xh, "helpres", Memc[out])
+		call delete (Memc[out])
+	    }
+
+	} else {
+	    # No help was found, send the contents straight to the display.
+            call xh_text_msg (XH_GP(xh), "helpres", Memc[text])
+	}
+
+
+err_	if (fdi != ERR)
+	    call close (fdi)
+	call sfree (sp)
+end
+
+
+# XH_LOAD_FILE -- Load the named file in the GUI.
+
+procedure xh_load_file (xh, parameter, filename)
+
+pointer	xh				# task descriptor
+char	parameter[ARB]			# GUI parameter to notify
+char	filename[ARB]			# file to display
+
+pointer	sp, ip, line, text
+int	fd, open(), getline(), gstrcpy()
+long	fsize, fstatl()
+errchk	open
+
+begin
+	call smark (sp)
+	call salloc (line, SZ_LINE, TY_CHAR)
+
+	# Open the file and send it to the display.
+	fd = open (filename, READ_ONLY, TEXT_FILE)
+	if (fd != ERR) {
+	    fsize = fstatl (fd, F_FILESIZE)
+	    call salloc (text, fsize+1, TY_CHAR)
+	    call aclrc (Memc[text], fsize+1)
+
+	    for (ip=text; getline (fd, Memc[line]) != EOF; )
+	        ip = ip + gstrcpy (Memc[line], Memc[ip], SZ_LINE)
+
+	    Memc[ip] = EOS
+	    call close (fd)
+
+            call xh_text_msg (XH_GP(xh), parameter, Memc[text])
+	}
+
+	call sfree (sp)
+end
+
+
+# XH_TEXT_MSG -- Send a text message to a named UI parameter but first
+# escape all curly braces so it passes through the Tcl correctly.
+
+procedure xh_text_msg (gp, param, msg)
+
+pointer gp
+char	param[ARB], msg[ARB]
+
+pointer	buf, ip
+int	i, nchars
+int	strlen()
+
+begin
+	nchars = strlen (msg)
+	call calloc (buf, nchars + SZ_LINE, TY_CHAR)
+
+	ip = buf
+	for (i=1; i < nchars; i=i+1) {
+	    if (msg[i] == '{' || msg[i] == '}') {
+		Memc[ip] = '\\'
+		ip = ip + 1
+	    }
+	    Memc[ip] = msg[i]
+	    ip = ip + 1
+	}
+
+	call gmsg (gp, "type", "file")
+	call gmsg (gp, param, Memc[buf])
+	call mfree (buf, TY_CHAR)
+end
diff --git a/pkg/system/help/xhelp/xhpkg.x b/pkg/system/help/xhelp/xhpkg.x
new file mode 100644
index 00000000..44580f15
--- /dev/null
+++ b/pkg/system/help/xhelp/xhpkg.x
@@ -0,0 +1,192 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<fset.h>
+include	<error.h>
+include	"../help.h"
+include	"../helpdir.h"
+include	"xhelp.h"
+
+
+# XH_GPKGLIST -- Get the requested package list as a sorted array of pointers.
+# This is essentially a "help <pkg>" request,  the caller passes the sorted
+# list returned in 'pkglist" required.
+
+int procedure xh_pkglist (xh, topic, helpdb, pkglist)
+
+pointer xh                		#i task descriptor pointer
+char    topic[ARB]                	#i search key
+char    helpdb[ARB]             	#i filename of database to be examined
+char    pkglist[ARB]             	#o package list
+
+int     i, m
+pointer	sp, pknm, hp, pp, sym
+pointer db, ixoff, ix
+
+bool    strne()
+int     gstrcpy(), strsearch(), hd_getname()
+pointer hdb_open(), hdb_load(), stfind()
+errchk  hdb_open, hdb_printpack, hdb_load
+
+begin
+        call smark (sp)
+        call salloc (pknm, MAX_MENUSIZE, TY_POINTER)
+
+        db = hdb_open (helpdb)
+        ixoff = HDB_INDEXPTR(db)
+	pp = NULL
+
+	# Search for the right topic.
+        do i = 1, HDB_NENTRIES(db) {
+            ix = ixoff + (i - 1) * LEN_HDBINDEX
+            if (strne (DBI_KEY(ix), "_index") &&
+                strsearch (DBI_KEY(ix),topic) != 0) {
+
+	        iferr (hp = hdb_load (db, DBI_KEY(ix))) {
+      	            call sfree (sp)
+      	            call erract (EA_WARN)
+                    return (0)
+        	}
+
+		# If this isn't the package we're after then move on.
+        	if (HD_PAKNAME(hp) == 0 ||
+		    strne (topic, Memc[HD_SBUF(hp)+HD_PAKNAME(hp)]))
+			next
+
+	        # Extract the names of the modules in the package.  Save the
+	        # pointers in an array for the table print routine.
+
+		pp = 1
+	        for (m=0;  m < MAX_MENUSIZE;  m=m+1) {
+	            call salloc (Memi[pknm+m], MAX_NAMELEN, TY_CHAR)
+	            if (hd_getname (hp, m+1, TY_MODNAME, Memc[Memi[pknm+m]],
+	                MAX_NAMELEN) <= 0)
+                    	    break
+
+		    # Copy the names to the output array.
+		    pp = pp + gstrcpy (Memc[Memi[pknm+m]], pkglist[pp], ARB)
+		    if (XH_SHOWTYPE(xh) == YES && XH_STP(xh) != NULL) {
+		        sym = stfind (XH_STP(xh), Memc[Memi[pknm+m]])
+		        if (sym != NULL)
+		            pp = pp + gstrcpy (".", pkglist[pp], ARB)
+		    }
+		    pp = pp + gstrcpy (" ", pkglist[pp], ARB)
+        	}
+		break
+            }
+        }
+	if (pp != NULL)
+	    pkglist[pp] = EOS
+
+        call hdb_free (db, hp)
+        call hdb_close (db)
+        call sfree (sp)
+	return (pp)
+end
+
+
+# XH_PKGPATH -- Get the package path associated with a particular task.
+# If we're given a parent package follow it back so we get the correct
+# path for a task that may be defined multiple places (e.g. SPLOT).
+
+procedure xh_pkgpath (xh, topic, curpack, path)
+
+pointer xh                                      #i task struct pointer
+char    topic[ARB]                           	#i help topic
+char    curpack[ARB]                           	#i help topic
+char    path[ARB]                           	#o package path
+
+pointer	sp, pkg, task, buf
+int	strncmp(), xh_pkgname()
+bool	streq()
+
+begin
+	call smark (sp)
+	call salloc (pkg, SZ_FNAME, TY_CHAR)
+	call salloc (task, SZ_FNAME, TY_CHAR)
+	call salloc (buf, SZ_FNAME, TY_CHAR)
+
+	if (curpack[1] == EOS || 
+	    streq (topic, curpack) ||
+	    strncmp ("root", curpack, 4) == 0 || 
+	    streq ("clpackage", curpack)) {
+	        call strcpy (topic, Memc[task], SZ_FNAME)
+	        call strcpy (topic, path, SZ_FNAME)
+	} else {
+	    call strcpy (curpack, Memc[task], SZ_FNAME)
+	    call sprintf (path, SZ_PATHNAME, "%s.%s")
+		call pargstr (curpack)
+		call pargstr (topic)
+	}
+
+	Memc[pkg] = EOS
+	while (xh_pkgname (xh, Memc[task], Memc[pkg]) == OK) {
+	    if (strncmp ("root", Memc[pkg], 4) == 0 || 
+		streq (Memc[task], Memc[pkg]) ||
+		streq ("clpackage", Memc[pkg]))
+		    break
+	    else {
+	        call sprintf (Memc[buf], SZ_PATHNAME, "%s.%s")
+		    call pargstr (Memc[pkg])
+		    call pargstr (path)
+		call strcpy (Memc[buf], path, SZ_PATHNAME)
+	    }
+	    call strcpy (Memc[pkg], Memc[task], SZ_FNAME)
+	    Memc[pkg] = EOS
+	}
+
+	call sfree (sp)
+end
+
+
+# XH_PKGNAME -- Get the package name associated with a particular task.
+
+int procedure xh_pkgname (xh, topic, pack)
+
+pointer xh                                      #i task struct pointer
+char    topic[ARB]                           	#i help topic
+char    pack[ARB]                           	#o package
+
+pointer sp, line, fname
+long	fsize, fstatl()
+int	fd, status, getline(), open(), stridxs()
+
+begin
+	call smark (sp)
+	call salloc (line, SZ_LINE, TY_CHAR)
+	call salloc (fname, SZ_LINE, TY_CHAR)
+
+	status = ERR
+
+	# Get a temp file name.
+	call mktemp ("tmp$xhelpi", Memc[fname], SZ_FNAME)
+
+	# Open a temp file with the help information found.
+	fd = open (Memc[fname], NEW_FILE, TEXT_FILE)
+	call xh_get_help (fd, topic, "", "", HF_HTML, HELPDB(xh),
+	    "all", "files")
+	call close (fd)
+
+	# Open the results file for reading.
+	fd = open (Memc[fname], READ_ONLY, TEXT_FILE)
+	fsize = fstatl (fd, F_FILESIZE)
+
+	# Search the results for the package line.
+        if (fsize != 0) {
+	    status = OK
+            while (getline (fd, Memc[line]) != EOF) {
+
+	        # Extract the package name.
+	        if (stridxs (":", Memc[line]) > 0) {
+		    Memc[line+stridxs(".",Memc[line])-1] = EOS
+		    call strcpy (Memc[line], pack, SZ_FNAME)
+	            break
+	        }
+	    }
+	}
+
+	call close (fd)				# clean up 
+	call delete (Memc[fname])
+	call sfree (sp)
+
+	return (status)
+end
diff --git a/pkg/system/help/xhelp/xhprint.x b/pkg/system/help/xhelp/xhprint.x
new file mode 100644
index 00000000..4e273848
--- /dev/null
+++ b/pkg/system/help/xhelp/xhprint.x
@@ -0,0 +1,151 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<syserr.h>
+include	<error.h>
+include <ctype.h>
+include	<ttyset.h>
+include "../help.h"
+include	"xhelp.h"
+
+
+# XH_PRINT_HELP -- Print the requested help topic.
+
+procedure xh_print_help (xh, topic, curpack, name)
+
+pointer xh                                      #i task struct pointer
+char    topic[ARB]                           	#i help topic
+char    curpack[ARB]                           	#i current package
+char    name[ARB]                           	#i printer name
+
+int	fd
+char	fname[SZ_FNAME]
+
+int	open()
+errchk	open
+
+begin
+	# Open a temp file with the help information.
+	call mktemp ("tmp$xhelpi", fname, SZ_FNAME)
+	fd = open (fname, NEW_FILE, TEXT_FILE)
+	call xh_get_help (fd, topic, curpack, "", HF_PS, HELPDB(xh),
+	    "all", OPTION(xh))
+	call close (fd)
+
+	call strcpy (name, PRINTER(xh), SZ_FNAME)
+	call xh_lprint (fname, PRINTER(xh))
+
+	# Clean up.
+	call delete (fname)
+end
+
+
+# XH_LPRINT -- Print a file or files on the lineprinter.  TTYPUTLINE is used
+# to output lines to the printer file, so that formfeeds, standout mode, etc.,
+# may be properly translated for the indicated device.
+
+procedure xh_lprint (fname, device)
+
+char	fname[ARB]
+char	device[ARB]
+
+int	out
+pointer	sp, lp, ddstr
+
+pointer	ttyodes()
+bool	streq()
+int	envgets(), lpopen(), ttygets()
+string	printer "printer"
+errchk	clgfil, xh_print_file
+
+begin
+	call smark (sp)
+	call salloc (ddstr,  SZ_DDSTR, TY_CHAR)
+
+	# Get device name.  Default is "printer", which means that the actual
+	# device name is given by the environment variable "printer".
+
+	if (streq (device, printer))
+	    if (envgets (printer, device, SZ_FNAME) <= 0)
+		call syserrs (SYS_ENVNF, printer)
+
+	# Open TTY descriptor and printer file.  We deal only with character
+	# data, so open printer as a text file.  Output the files to the line
+	# printer device.
+
+	lp = ttyodes (device)
+	if (ttygets (lp, "DD", Memc[ddstr], SZ_DDSTR) <= 0)
+	    call error (1, "missing 'DD' parameter in termcap entry")
+
+	out = lpopen (Memc[ddstr], APPEND, TEXT_FILE)
+
+	# Output file, followed by a form feed.
+	iferr {
+	    call xh_print_file (out, lp, fname)
+	    	call flush (out)
+	} then
+	    call erract (EA_WARN)
+
+	call close (out)
+	call ttycdes (lp)
+	call sfree (sp)
+end
+
+
+# XH_PRINT_FILE -- Open and print the named text file on the output file,
+# under control of the tty device descriptor LP.  Print a header on each
+# page if enabled, and formfeed at the end of the file.  The number of lines
+# per page is given by the tty descriptor.
+
+procedure xh_print_file (out, lp, fname)
+
+int	out
+pointer	lp
+char	fname[ARB]
+
+bool	one_tab_in
+int	lineno, maxlines, status, in
+pointer	sp, ip, lbuf
+int	open(), getline(), ttystati()
+errchk	salloc, open, ttystati, getline, ttyputline
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE+1, TY_CHAR)
+
+	in = open (fname, READ_ONLY, TEXT_FILE)
+	maxlines = ttystati (lp, TTY_NLINES)
+
+	# If printer page is very wide, set page in one tabstop from left
+	# margin.
+
+	one_tab_in = (ttystati (lp, TTY_NCOLS) > WIDE_PAGE)
+	if (one_tab_in) {
+	    Memc[lbuf] = '\t'
+	    ip = lbuf + 1
+	} else
+	    ip = lbuf
+
+
+	# Format and write each page of output.  If headers are enabled,
+	# output formfeed and header between pages.
+
+	status = OK
+	while (status != EOF) {
+	    # Output one page of text.  Each output line is processed by
+	    # ttyputline to expand tabs, underline, etc.
+
+	    for (lineno=1;  lineno <= maxlines;  lineno=lineno+1) {
+		status = getline (in, Memc[ip])
+		if (status == EOF)
+		    break
+		if (Memc[ip] == '\f') {
+		    call ttyputline (out, lp, "\f", YES)
+		    call strcpy (Memc[ip+1], Memc[ip], SZ_LINE)
+		}
+		call ttyputline (out, lp, Memc[lbuf], YES)
+	    }
+	}
+
+	call close (in)
+	call sfree (sp)
+end
diff --git a/pkg/system/help/xhelp/xhqref.x b/pkg/system/help/xhelp/xhqref.x
new file mode 100644
index 00000000..51912169
--- /dev/null
+++ b/pkg/system/help/xhelp/xhqref.x
@@ -0,0 +1,250 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<syserr.h>
+include	<finfo.h>
+include	<ctype.h>
+include	"../help.h"
+include	"xhelp.h"
+
+define	MAX_PKGS	200
+define	SZ_PKG		32
+define	EXTPKG		"hlib$extern.pkg"
+
+
+# XH_UPDT_QUICKREF -- Update a quick reference file for the help database
+# if needed.
+
+procedure xh_updt_quickref (xh)
+
+pointer	xh
+
+long	fiq[LEN_FINFO], fie[LEN_FINFO]
+
+int	access()
+long	finfo()
+errchk	access, finfo
+
+begin
+	# If the quickref file exists, and it is older than the
+	# hlib$extern.pkg file which defines the helpdb rebuild
+	# the quickref file so we're current, otherwise rebuild 
+	# the quickref file anyway.
+	if ((access (QUICKREF(xh), 0, 0) == NO) ||
+	    (access (PKGFILE, 0, 0) == NO)) {
+		call xh_make_quickref (xh, QUICKREF(xh))
+	} else {
+            if (finfo (QUICKREF(xh), fiq) == ERR)
+            	call filerr (QUICKREF(xh), SYS_FOPNNEXFIL)
+            if (finfo (EXTPKG, fie) == ERR)
+            	call filerr (EXTPKG, SYS_FOPNNEXFIL)
+
+	    if (FI_MTIME[fie] > FI_MTIME[fiq]) {
+	        call delete (QUICKREF(xh))
+	        call xh_make_quickref (xh, QUICKREF(xh))
+	    }
+	}
+end
+
+
+# XH_MAKE_QUICKREF -- Make a quick reference file for the help database.
+
+procedure xh_make_quickref (xh, quickref)
+
+pointer	xh					#i task struct pointer
+char	quickref[ARB]				#i quickref filename
+
+pointer	pkglist, pp
+int	fdi, fdo, fd_err
+int	sz_pbuf, npkgs
+char	fname[SZ_FNAME], efname[SZ_FNAME]
+char	pkg[SZ_PKG], lastpkg[SZ_PKG], line[SZ_LINE]
+
+int	open(), access(), getline(), gstrcpy(), strsearch(), strcmp()
+errchk	open
+
+begin
+	# Tell the GUI we're actually busy doing something....
+	call gmsg (XH_GP(xh), "alert", 
+	    "Please wait,\ngenerating QuickRef file...")
+	call gflush (XH_GP(xh))
+
+	# Open the STDERR stream onto a dummy file to catch any
+	# "no help available" messages that may interfere with the
+	# textout GUI parameter.
+	call mktemp ("tmp$xerr", efname, SZ_FNAME)
+	fd_err = open (efname, NEW_FILE, TEXT_FILE)
+	call frediro (STDERR, fd_err)
+
+        # Open a temp file with the raw search information.
+        call mktemp ("tmp$xhelpq", fname, SZ_FNAME)
+        fdi = open (fname, NEW_FILE, TEXT_FILE)
+
+        call xh_get_help (fdi, "[a-z]*.", "", "", HF_HTML, HELPDB(xh), "help",
+	    "references")
+        call close (fdi)
+
+	# Close the error file, swap back the descriptors first and delete it.
+	call fswapfd (STDERR, fd_err)
+	call close (fd_err)
+	call delete (efname)
+
+	# Delete existing files.
+	if (access (quickref, 0, 0) == YES)
+	    call delete (quickref)
+
+	# Open the references file.
+	iferr (fdo = open (quickref, NEW_FILE, TEXT_FILE)) {
+	    call sprintf (line, SZ_LINE, "Cannot create quickref file `%s'.")
+     		call pargstr (quickref)
+	    call error (0, line)
+	}
+        fdi = open (fname, READ_ONLY, TEXT_FILE)
+
+	# Initialize for the package list.
+	pkg[1] = EOS
+	lastpkg[1] = EOS
+	npkgs = 0
+	sz_pbuf = MAX_PKGS * SZ_PKG
+	call malloc (pkglist, sz_pbuf, TY_CHAR)
+	pp = pkglist
+
+	# Loop over the lines in the file, save the ones that are the
+	# descriptions.  Descriptions are assumed to contain a '-' since
+	# this is standard for .men files, to be safe we also required 
+	# the bracketed package name generated with a 'references' query.
+	# The package name is used to create the package list symtab we
+	# use to show the item as a task or package.
+
+	while (getline(fdi,line) != EOF) {
+	    if (strsearch (line, " - ") != 0 && strsearch (line, "[") != 0) {
+		call putline (fdo, line)
+
+		# Extract the package name.  The results aren't sorted at
+		# this point so we see whether this package is the same as
+		# the last before adding it to the package list.  This 
+		# gives us a unique list of unsorted package names.
+
+		call xh_gname (line, pkg)
+		if (lastpkg[1] == EOS)
+		    call strcpy (pkg, lastpkg, SZ_PKG)
+		if (strcmp (pkg, lastpkg) != 0) {
+		    npkgs = npkgs + 1
+
+		    # Protect against overflowing the buffer.
+		    if ((pp - pkglist) > sz_pbuf)
+			call error (1, "Package buffer overflow.")
+
+		    pp = pp + gstrcpy (pkg, Memc[pp], SZ_PKG)
+		    pp = pp + gstrcpy (" ", Memc[pp], SZ_PKG)
+		    call strcpy (pkg, lastpkg, SZ_PKG)
+		}
+	    }
+	}
+	Memc[pp] = EOS
+
+	# Close the new references file.
+	call close (fdo)
+
+	# Close and delete the references temp files.
+	call close (fdi)
+	call delete (fname)
+
+	# Sort the lines of the references file.
+	call xh_file_sort (quickref)
+
+	# Now send the package list off to be stored for the next time.
+	call xh_make_pkglist (Memc[pkglist])
+
+	call mfree (pkglist, TY_CHAR)
+
+	# Tell the GUI we're done.
+	call gmsg (XH_GP(xh), "alert", "dismiss")
+	call gflush (XH_GP(xh))
+end
+
+
+# XH_MAKE_PKGLIST -- Update the package symtab file.  We are only called
+# when updating the quickref database so delete any existing file before
+# writing the new one.
+
+procedure xh_make_pkglist (list)
+
+char	list[ARB]				#i package list
+
+pointer	sp, err, pkg
+pointer	stp, sym
+int	i, ip, fd
+
+pointer	stopen(), stenter()
+int	open(), access(), strlen()
+errchk	open
+
+begin
+	call smark (sp)
+	call salloc (err, SZ_LINE, TY_CHAR)
+	call salloc (pkg, SZ_FNAME, TY_CHAR)
+
+	# Delete existing files.
+	if (access(PKGFILE,0,0) == YES)	
+	    call delete (PKGFILE)
+
+	# Open the package symtab file.
+	iferr (fd = open (PKGFILE, NEW_FILE, BINARY_FILE)) {
+	    call sprintf (Memc[err], SZ_LINE, "Can't create pkg symtab `%s'.")
+     		call pargstr (PKGFILE)
+	    call error (0, Memc[err])
+	}
+
+	# Sort the list before making the symtab.
+	call xh_sort_list (list)
+
+	# Open the symtab.
+	stp = stopen ("package list", LEN_INDEX, LEN_STAB, SZ_SBUF)
+
+	# Enter strings in the symtab.
+	i = 0
+	for (ip=1; list[ip] != EOS; ip=ip+1) {
+	    if (list[ip] == ' ') {
+	    	Memc[pkg+i] = EOS
+		sym = stenter (stp, Memc[pkg], strlen (Memc[pkg]) + 1)
+		i = 0
+		ip = ip + 1
+	    }
+	    Memc[pkg+i] = list[ip]
+	    i = i + 1
+	}
+
+	# Save the symtab file.
+	call stsqueeze (stp)
+	call stsave (stp, fd)
+
+	# Clean up.
+	call stclose (stp)
+	call close (fd)
+	call sfree (sp)
+end
+
+
+# XH_GNAME -- Extract the package name from a 'references' line.
+
+procedure xh_gname (line, pkg)
+
+char	line[ARB]				#i references line
+char	pkg[ARB]				#o package name in line
+
+int	ip, strlen()
+
+begin
+	# Clear trailing whitespace back to the ']' at end of the 
+	# package name.
+	for (ip=strlen(line); line[ip] != ']' || IS_WHITE(line[ip]); )
+	    ip = ip - 1
+	line[ip] = EOS
+
+	# Move back to starting '[' of package name.
+	while (line[ip] != '[' && ip > 0)
+	    ip = ip - 1
+
+	# What's left is the package name, copy it to the output.
+	call strcpy (line[ip+1], pkg, SZ_PKG)
+end
diff --git a/pkg/system/help/xhelp/xhroot.x b/pkg/system/help/xhelp/xhroot.x
new file mode 100644
index 00000000..9559662a
--- /dev/null
+++ b/pkg/system/help/xhelp/xhroot.x
@@ -0,0 +1,73 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	"xhelp.h"
+
+
+# XH_ROOT_PKG -- Make the root package.  Search the help database and
+# create a package list for all modules found.  We add special entries for
+# system modules (imfort/math/os) which are not normally in the help tree
+# but provide documents.
+
+procedure xh_root_pkg (xh)
+
+pointer	xh				#i struct pointer.
+
+pointer sp, fname, buf, ip, op, lp
+int     list
+
+int     xh_pkglist()
+int     gstrcpy(), fntopnb(), fntgfnb()
+bool    strne()
+
+begin
+        call smark (sp)
+        call salloc (fname, SZ_FNAME, TY_CHAR)
+        call salloc (buf, SZ_FNAME, TY_CHAR)
+
+        # Set initial packages and help databases.  This consists of the
+        # system documentation (in sys$sys.hd), the contents of the
+        # clpackage module and each of the external packages.
+
+        lp = XH_LPTR(xh)
+
+        # Create an entry for seldom-read system docs.
+	if (XH_SHOWTYPE(xh) == YES) {
+            lp = lp + gstrcpy ("imfort. ", Memc[lp], ARB)
+            lp = lp + gstrcpy ("math. ", Memc[lp], ARB)
+            lp = lp + gstrcpy ("os. ", Memc[lp], ARB)
+	} else {
+            lp = lp + gstrcpy ("imfort ", Memc[lp], ARB)
+            lp = lp + gstrcpy ("math ", Memc[lp], ARB)
+            lp = lp + gstrcpy ("os ", Memc[lp], ARB)
+	}
+
+        # Add the external packages to the list.
+        list = fntopnb (HELPDB(xh), YES)
+        while (fntgfnb (list, Memc[fname], SZ_FNAME) != EOF) {
+            op = buf
+            ip = fname
+            while (Memc[ip] != '$' && Memc[ip] != EOS && Memc[ip] != ',') {
+                Memc[op] = Memc[ip]
+                ip = ip + 1
+                op = op + 1
+            }
+            Memc[op] = EOS
+            if (strne(Memc[buf],"lib")) {
+                lp = lp + gstrcpy (Memc[buf], Memc[lp], ARB)
+		if (XH_SHOWTYPE(xh) == YES)
+                    lp = lp + gstrcpy (".", Memc[lp], ARB)
+                lp = lp + gstrcpy (" ", Memc[lp], ARB)
+            }
+        }
+
+        # Add the clpackage contents to the list.
+        lp = lp + xh_pkglist (xh, "clpackage", HELPDB(xh), Memc[lp])
+
+	if (lp > (XH_LPTR(xh) + SZ_HELPLIST))
+	    call error (1, "Memory error: LIST pointer overflow.")
+
+        # Sort the list so it's presentable.
+        call xh_sort_list (LIST(xh))
+
+	call sfree (sp)
+end
diff --git a/pkg/system/help/xhelp/xhsave.x b/pkg/system/help/xhelp/xhsave.x
new file mode 100644
index 00000000..f280146d
--- /dev/null
+++ b/pkg/system/help/xhelp/xhsave.x
@@ -0,0 +1,184 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include <ctype.h>
+include "../help.h"
+include "xhelp.h"
+
+define	SZ_EXTN		10
+
+
+# XH_SAVE_FILE --  Save the currently displayed page to the named file in
+# the specified format.
+
+procedure xh_save_file (xh, iname, oname, format, overwrite)
+
+pointer	xh					#i package descriptor
+char	iname[ARB]				#i input file
+char	oname[ARB]				#i output filename
+char	format[ARB]				#i format
+int	overwrite				#i overwrite flag
+
+pointer	sp, buf
+char	extn[SZ_EXTN]
+
+int	access(), fnextn(), strlen(), strcmp()
+
+begin
+	if (access (oname, 0, 0) == YES && overwrite == NO) {
+	    call smark (sp)
+	    call salloc (buf, SZ_LINE, TY_CHAR)
+	    call sprintf (Memc[buf], SZ_LINE,
+		"Operation would overwrite\nexisting file '%s'")
+		    call pargstr (oname)
+	    call gmsg (XH_GP(xh), "alert", Memc[buf])
+	    call sfree (sp)
+	    return
+
+	} else if (access (oname, 0, 0) == YES)
+	    call delete (oname)
+
+	call aclrc (extn, SZ_EXTN)
+
+	# Strip off any braces around the arguments that were put
+	# in by the Tcl script.
+	if (iname[1] == '{') {
+	    call amovc (iname[2], iname, strlen(iname)-2)
+	    iname[strlen(iname)-1] = EOS
+	}
+	if (oname[1] == '{') {
+	    call amovc (oname[2], oname, strlen(oname)-2)
+	    oname[strlen(oname)-1] = EOS
+	}
+
+	switch (format[1]) {
+	case 's':				# source
+	    call fcopy (iname, oname)
+	case 't':				# text
+	    if (fnextn(iname, extn, 3) > 0 && strcmp("hlp", extn) == 0)
+		call xh_save_text (xh, iname, oname)
+	    else
+	        call fcopy (iname, oname)
+	case 'h':				# html
+	    if (fnextn(iname, extn, 3) > 0 && strcmp("hlp", extn) == 0)
+	        call xh_save_html (xh, iname, oname, YES)
+	    else
+	        call xh_save_html (xh, iname, oname, NO)
+	case 'p':				# postscript
+	    if (fnextn(iname, extn, 3) > 0 && strcmp("hlp", extn) == 0)
+	        call xh_save_ps (xh, iname, oname, YES)
+	    else
+	        call xh_save_ps (xh, iname, oname, NO)
+	default:
+	    call gmsg (XH_GP(xh), "alert", "Invalid format specifier")
+	}
+end
+
+
+# XH_SAVE_TEXT -- Save the page as a formatted text file.  
+
+procedure xh_save_text (xh, in, out)
+
+pointer	xh					#i package descriptor
+char	in[ARB], out[ARB]			#i file names
+
+char	err[SZ_LINE]
+int	fdout, open()
+errchk	open
+
+begin
+	# Open the output file.
+	iferr (fdout = open (out, NEW_FILE, TEXT_FILE)) {
+            call sprintf (err, SZ_LINE, "Cannot open output file `%s'.")
+               call pargstr (out)
+	    call gmsg (XH_GP(xh), "alert", err)
+	    return
+	}
+
+	# Format the help as text.
+	call xh_get_help (fdout, "", "", in, HF_TEXT, HELPDB(xh), "all", "help")
+
+	# Close the file.
+	call close (fdout)
+end
+
+
+# XH_SAVE_HTML -- Save the page as an HTML  file.  
+
+procedure xh_save_html (xh, in, out, ishelp)
+
+pointer	xh					#i package descriptor
+char	in[ARB], out[ARB]			#i file names
+int	ishelp					#i is this a help file?
+
+char	err[SZ_LINE]
+int	fdout, fdin, open()
+errchk	open
+
+begin
+	# Open the output file.
+	iferr (fdout = open (out, NEW_FILE, TEXT_FILE)) {
+            call sprintf (err, SZ_LINE, "Cannot open output file `%s'.")
+               call pargstr (out)
+	    call gmsg (XH_GP(xh), "alert", err)
+	    return
+	}
+
+	# Format the help as text.
+	if (ishelp == YES) {
+	    call xh_get_help (fdout, "", "", in, HF_HTML, HELPDB(xh),
+		"all", "help")
+	} else {
+	    # Open the input file.
+	    iferr (fdout = open (out, NEW_FILE, TEXT_FILE)) {
+                call sprintf (err, SZ_LINE, "Cannot open output file `%s'.")
+                   call pargstr (out)
+	        call gmsg (XH_GP(xh), "alert", err)
+	        return
+	    }
+
+	    call fprintf (out, "<HTML><BODY>\n")
+	    call fprintf (out, "<TITLE>%s</TITLE>\n")
+		call pargstr (in)
+	    call fprintf (out, "<PRE>\n")
+	    call fcopyo (in, out)
+	    call fprintf (out, "</PRE>\n")
+	    call fprintf (out, "</BODY></HTML>\n")
+
+	    call close (fdin)
+	}
+
+	# Close the file.
+	call close (fdout)
+end
+
+
+# XH_SAVE_PS -- Save the page as a postscript file.  
+
+procedure xh_save_ps (xh, in, out, ishelp)
+
+pointer	xh					#i package descriptor
+char	in[ARB], out[ARB]			#i file names
+int	ishelp					#i is this a help file?
+
+char	err[SZ_LINE]
+int	fdout, open()
+errchk	open
+
+begin
+	if (ishelp == NO)
+	    return
+
+	# Open the output file.
+	iferr (fdout = open (out, NEW_FILE, TEXT_FILE)) {
+            call sprintf (err, SZ_LINE, "Cannot open output file `%s'.")
+               call pargstr (out)
+	    call gmsg (XH_GP(xh), "alert", err)
+	    return
+	}
+
+	# Format the help as text.
+	call xh_get_help (fdout, "", "", in, HF_PS, HELPDB(xh), "all", "help")
+
+	# Close the file.
+	call close (fdout)
+end
diff --git a/pkg/system/help/xhelp/xhsearch.x b/pkg/system/help/xhelp/xhsearch.x
new file mode 100644
index 00000000..18f1aa28
--- /dev/null
+++ b/pkg/system/help/xhelp/xhsearch.x
@@ -0,0 +1,185 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<fset.h>
+include <ctype.h>
+include	"xhelp.h"
+
+
+# XH_SEARCH -- Given a search pattern generate a list of tasks matching the
+# pattern.
+
+procedure xh_search (xh, exact_match, pattern)
+
+pointer	xh					#i task struct pointer
+int	exact_match				#i require an exact match?
+char	pattern[ARB]				#i search pattern
+
+pointer	sp, lfile, line, helpstr, item
+pointer	tp, lp, pat
+char	task[SZ_FNAME], pkg[SZ_FNAME], desc[SZ_FNAME]
+int	i, ip, fd, fdl
+long	fsize
+
+long	fstatl()
+int	open(), xh_match(), getline(), gstrcpy()
+errchk	open, xh_updt_quickref
+
+begin
+	call smark (sp)
+	call salloc (lfile, SZ_FNAME, TY_CHAR)
+	call salloc (line, SZ_LINE, TY_CHAR)
+	call salloc (item, SZ_LINE, TY_CHAR)
+	call salloc (pat, SZ_FNAME, TY_CHAR)
+
+	# Open the quick reference file.
+	iferr (fd = open (QUICKREF(xh), READ_ONLY, TEXT_FILE))
+	    call error (0, "Cannot open quick-reference file.")
+
+	# Open the temp file for the matched lines.
+        call mktemp ("tmp$xhelpq", Memc[lfile], SZ_FNAME)
+        fdl = open (Memc[lfile], NEW_FILE, TEXT_FILE)
+
+	# Allocate space for the results string.
+	fsize = fstatl (fd, F_FILESIZE)
+
+	# Check pattern for multi-word searches.
+	call aclrc (Memc[pat], SZ_FNAME)
+	for (ip=1; IS_WHITE(pattern[ip]); ip=ip+1)
+	    ;
+	if (pattern[ip] == '{') {
+	    ip = ip + 1
+	    for (i=0; pattern[ip] != '}'; ip=ip+1) {
+		Memc[pat+i] = pattern[ip]
+		i = i + 1
+	    }
+	} else
+	    call strcpy (pattern[ip], Memc[pat], SZ_FNAME)
+
+	# Loop over the lines in the file and match the pattern.
+	while (getline(fd,Memc[line]) != EOF) {
+	    if (xh_match (Memc[line], Memc[pat], exact_match) != 0)
+		call putline (fdl, Memc[line])	# save it to the file
+	}
+	call close (fdl)
+	call close (fd)
+
+	# Now read back the file to see if we matched anything, and so we
+	# can parse the file for task/package names.
+        fdl = open (Memc[lfile], READ_ONLY, TEXT_FILE)
+	fsize = fstatl (fdl, F_FILESIZE)
+
+	# See whether we got anything, if not return.
+	if (fsize == 0) {
+	    call sprintf (Memc[line], SZ_LINE, 
+		"No help available for\npattern `%s'.")
+	            call pargstr (Memc[pat])
+	    call gmsg (XH_GP(xh), "alert", Memc[line])
+
+	    call close (fdl)
+	    call delete (Memc[lfile])
+	    call sfree (sp)
+	    return
+	}
+	call calloc (helpstr, 5*fsize, TY_CHAR)
+
+	# Read back the sorted list, separate the task name, package, and
+	# descriptions and format it for the GUI.
+	tp = helpstr
+	tp = tp + gstrcpy ("<HTML><BODY><PRE>", Memc[tp], ARB)
+	while (getline(fdl, Memc[line]) != EOF) {
+	    Memc[item] = EOS
+	    lp = line
+
+	    # Get the task name.
+	    for (i=1; !IS_WHITE(Memc[lp]); i=i+1) {
+		task[i] = Memc[lp]
+		lp = lp + 1
+		if (Memc[lp] == EOS || i >= SZ_FNAME)
+		    break 
+	    }
+	    task[i] = EOS
+
+	    # Skip delimiter.
+	    while (IS_WHITE(Memc[lp]) || Memc[lp] == '-') 
+		lp = lp + 1
+
+	    # Get the description up to the '[' package name.
+	    for (i=1; Memc[lp] != '['; i=i+1) {
+		desc[i] = Memc[lp]
+		lp = lp + 1
+		if (Memc[lp] == EOS || i >= SZ_FNAME)
+		    break 
+	    }
+	    desc[i] = EOS
+
+	    # Get the package name up to the ']' delimiter.
+	    if (Memc[lp] != EOS) {
+		lp = lp + 1
+	        for (i=1; Memc[lp] != ']'; i=i+1) {
+		    pkg[i] = Memc[lp]
+		    lp = lp + 1
+		    if (Memc[lp] == '\n' || Memc[lp] == EOS || i >= SZ_FNAME)
+		        break 
+	        }
+	    }
+	    pkg[i] = EOS
+
+	    call sprintf(Memc[item], SZ_LINE, "<A HREF=%s.%s>%10.10s</A> ")
+		call pargstr (pkg)
+		call pargstr (task)
+		call pargstr (task)
+	    tp = tp + gstrcpy (Memc[item], Memc[tp], ARB)
+	    call sprintf(Memc[item], SZ_LINE, "<A HREF=%s.%s>%10.10s</A> ")
+		call pargstr (pkg)
+		call pargstr (pkg)
+		call pargstr (pkg)
+	    tp = tp + gstrcpy (Memc[item], Memc[tp], ARB)
+	    call sprintf(Memc[item], SZ_LINE, "  %s\n")
+		call pargstr (desc)
+	    tp = tp + gstrcpy (Memc[item], Memc[tp], ARB)
+	}
+	call strcat ("</PRE></BODY></HTML>\n", Memc[tp], SZ_LINE)
+	Memc[tp] = EOS
+	call close (fdl)
+
+	# Send it to the GUI.
+	call gmsg (XH_GP(xh), "apropos", Memc[helpstr])
+
+	# Clean up.
+	call delete (Memc[lfile])
+	call mfree (helpstr, TY_CHAR)
+	call sfree (sp)
+end
+
+
+# XH_MATCH -- Match any or all words in a pattern against the given line.
+# We can either look for an exact match or just the occurence of one word
+# in the pattern.
+
+int procedure xh_match (line, pattern, exact_match)
+
+char	line[ARB]				#i line to be matched
+char	pattern[ARB]				#i pattern words
+int	exact_match
+
+char	word[SZ_FNAME]
+int	i, j
+int	strsearch()
+
+begin
+	if (exact_match == YES) {
+	    return (strsearch(line, pattern))
+	} else {
+	    # See if any word in the pattern matches in the line.
+	    for (i=1; pattern[i] != EOS; i=i+1) {
+		for (j=1; pattern[i] != EOS && pattern[i] != ' '; j=j+1) {
+		    word[j] = pattern[i]
+		    i = i + 1
+		}
+		word[j] = EOS
+	        if (strsearch(line, word) != 0)
+		    return (YES)
+	    }
+	    return (NO)
+	}
+end
diff --git a/pkg/system/help/xhelp/xhsort.x b/pkg/system/help/xhelp/xhsort.x
new file mode 100644
index 00000000..b5ccc081
--- /dev/null
+++ b/pkg/system/help/xhelp/xhsort.x
@@ -0,0 +1,223 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include <ctype.h>
+include "xhelp.h"
+
+define  MAXPTR          20000
+define  SZ_LINBUF       300000
+
+
+# XH_SORT_LIST -- Take a list of words (as with a package list) and sort them.
+
+procedure xh_sort_list (list)
+
+char	list[ARB]					#u list to be sorted
+
+pointer	sp, index, buf, ip
+int	i, j, count, len
+
+int	strlen()
+
+begin
+	len = strlen (list)
+
+	call smark (sp)
+	call salloc (index, SZ_HELPLIST, TY_INT)
+	call salloc (buf, len+2, TY_CHAR)
+
+	# Build up the index array.
+	count = 1
+	Memi[index] = 1
+	for (i=2; i<len; i=i+1) {
+	    if (list[i] == ' ') {
+		list[i] = EOS
+		Memi[index+count] = i + 1
+		count = count + 1
+	    }
+	}
+
+	# Sort the list.
+	call strsrt (Memi[index], list, count)
+
+	# Restore the list.
+	ip = buf
+	do i = 1, count {
+	    for (j=0; list[Memi[index+i-1]+j] != EOS; j=j+1) {
+	        Memc[ip] = list[Memi[index+i-1]+j]
+	        ip = ip + 1
+	    }
+	    Memc[ip] = ' '
+	    ip = ip + 1
+	}
+	Memc[ip-1] = EOS
+	call strcpy (Memc[buf], list, strlen(Memc[buf]))
+	call sfree (sp)
+end
+
+
+# XH_FILE_SORT -- Sort the lines in the named file.
+
+procedure xh_file_sort (fname)
+
+char	fname[SZ_FNAME]				#i file to be sorted
+
+pointer	linbuf, linptr
+int	nlines, fd
+int	open()
+
+begin
+        call calloc (linptr, MAXPTR, TY_INT)
+        call calloc (linbuf, SZ_LINBUF, TY_CHAR)
+
+	# Sort the file then write it back out.
+	fd = open (fname, READ_ONLY, TEXT_FILE)
+	call xh_gtext (fd,  Memi[linptr], nlines, Memc[linbuf])
+	call close (fd)
+	call delete (fname)
+
+        call xh_quick (Memi[linptr], Memc[linbuf], nlines)
+
+	fd = open (fname, NEW_FILE, TEXT_FILE)
+        call xh_ptext (fd, Memi[linptr], nlines, Memc[linbuf])
+
+	call mfree (linbuf, TY_CHAR)
+	call mfree (linptr, TY_INT)
+	call close (fd)
+end
+
+
+# XH_GTEXT -- Get text lines into linbuf.
+
+procedure xh_gtext (infile, linptr, nlines, linbuf)
+
+int	infile, linptr[ARB], nlines
+char	linbuf[ARB]
+
+int	lbp, len, getline()
+errchk	getline
+
+begin
+	nlines = 0
+	lbp = 1
+
+	repeat {
+	    len = getline (infile, linbuf[lbp])
+	    if (len == EOF)
+		break
+	    else if (len == 1)
+		# ignore blank lines
+	    else {
+		nlines = nlines + 1
+		linptr[nlines] = lbp
+		lbp = lbp + len + 1		# '1' = room for EOS
+	    }
+	} until (lbp >= SZ_LINBUF - SZ_LINE || nlines >= MAXPTR)
+end
+
+
+# XH_PTEXT -- Output text lines from linbuf to outfile.
+
+procedure xh_ptext (outfil, linptr, nlines, linbuf)
+
+int	outfil, linptr[ARB], nlines
+char	linbuf[ARB]
+int	i, j
+errchk	putline
+
+begin
+	for (i=1;  i <= nlines;  i=i+1) {
+	    j = linptr[i]
+	    call putline (outfil, linbuf[j])
+	}
+end
+
+
+# XH_QUICK -- Quicksort for text data.  NOTE -- This algorithm is quadratic in
+# the worst case, i.e., when the data is already sorted.  A random method of
+# selecting the pivot should be used to improve the behaviour on sorted arrays.
+
+procedure xh_quick (linptr, linbuf, nlines)
+
+int	linptr[ARB]		# indices of strings in buffer
+char	linbuf[ARB]		# string buffer
+int	nlines			# number of strings
+
+define  LOGPTR          32
+define  swap            {temp=$1;$1=$2;$2=temp}
+
+int	i, j, k, temp, lv[LOGPTR], p, pivlin, uv[LOGPTR]
+int	xh_compare()
+
+begin
+	lv[1] = 1
+	uv[1] = nlines
+	p = 1
+
+	while (p > 0) {
+	    if (lv[p] >= uv[p])		# only one elem in this subset
+		p = p - 1		# pop stack
+	    else {
+		# Dummy loop to trigger optimizer.
+		do p = p, ARB {
+		    i = lv[p] - 1
+		    j = uv[p]
+
+		    # Select pivot element at midpoint of interval to avoid
+		    # quadratic behavior on a sorted list.
+
+		    k = (lv[p] + uv[p]) / 2
+		    swap (linptr[j], linptr[k])
+		    pivlin = linptr[j]
+
+		    while (i < j) {
+			for (i=i+1; xh_compare (linptr[i], pivlin, linbuf) < 0;
+			    i=i+1)
+			    ;
+			for (j=j-1;  j > i;  j=j-1)
+			    if (xh_compare (linptr[j], pivlin, linbuf) <= 0)
+				break
+			if (i < j)		# out of order pair
+			    swap (linptr[i], linptr[j])
+		    }
+
+		    j = uv[p]			# move pivot to position i
+		    swap (linptr[i], linptr[j])
+
+		    if (i-lv[p] < uv[p] - i) {	# stack so shorter done first
+			lv[p+1] = lv[p]
+			uv[p+1] = i - 1
+			lv[p] = i + 1
+		    } else {
+			lv[p+1] = i + 1
+			uv[p+1] = uv[p]
+			uv[p] = i - 1
+		    }
+
+		    break
+		}
+
+		p = p + 1			# push onto stack
+	    }
+	}
+end
+
+
+# XH_COMPARE -- Compare two strings.  Return -1 if str1<str2, 1 if str1>str2,
+# or 0 if the two strings are equal.
+
+int procedure xh_compare (lp1, lp2, linbuf)
+
+int	lp1, lp2		# pointers to substrings in linbuf
+char	linbuf[ARB]		# text buffer
+
+int	ip1, ip2, answer
+int	strcmp()
+
+begin
+	for (ip1=lp1;  IS_WHITE(linbuf[ip1]);  ip1=ip1+1)
+		;
+	for (ip2=lp2;  IS_WHITE(linbuf[ip2]);  ip2=ip2+1)
+		;
+	answer = strcmp (linbuf[ip1], linbuf[ip2])
+	return (answer)
+end
diff --git a/pkg/system/help/xhelp/zzdebug.x b/pkg/system/help/xhelp/zzdebug.x
new file mode 100644
index 00000000..70a95d9b
--- /dev/null
+++ b/pkg/system/help/xhelp/zzdebug.x
@@ -0,0 +1,59 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+
+# ZZDEBUG.X -- Debug routines for the help formatting system.
+
+task	lroff2html = t_lroff2html,
+	lroff2ps   = t_lroff2ps
+
+
+# LROFF2HTML -- Test program to convert an Lroff source file to HTML.
+
+procedure t_lroff2html ()
+
+int	fdi, fdo
+char	iname[SZ_FNAME], oname[SZ_FNAME]
+
+int	open()
+errchk	open
+
+begin
+	call clgstr ("input",  iname, SZ_FNAME) 	# get parameters
+	call clgstr ("output", oname, SZ_FNAME)
+
+	fdi = open (iname, READ_ONLY, TEXT_FILE) 	# open the file
+	fdo = open (oname, NEW_FILE, TEXT_FILE)
+
+	# Process it.
+	call lroff2html (fdi, fdo, iname, "", "", "", "")
+
+	call close (fdi)
+	call close (fdo)
+end
+
+
+# LROFF2PS -- Test program to convert an Lroff source file to Postscript.
+
+procedure t_lroff2ps ()
+
+int	fdi, fdo
+char	iname[SZ_FNAME], oname[SZ_FNAME]
+
+pointer	ps
+int	open()
+errchk	open
+
+begin
+	call clgstr ("input",  iname, SZ_FNAME) 	# get parameters
+	call clgstr ("output", oname, SZ_FNAME)
+
+	fdi = open (iname, READ_ONLY, TEXT_FILE) 	# open the files
+	fdo = open (oname, READ_ONLY, TEXT_FILE)
+
+	# Process it.
+	ps = NULL
+	call lroff2ps (fdi, fdo, ps, "", "")
+
+	call close (fdi)
+	call close (fdo)
+end
diff --git a/pkg/system/lprint.par b/pkg/system/lprint.par
new file mode 100644
index 00000000..268ccac1
--- /dev/null
+++ b/pkg/system/lprint.par
@@ -0,0 +1,5 @@
+files,s,a,,,,"list of files to be printed"
+device,s,h,"printer",,,"name of output device"
+map_cc,b,h,yes,,,"make unknown control chars printable"
+paginate,s,h,"auto",,,"break pages (auto, yes, no)"
+label,s,h,"STDIN",,,"file name label if input redirected"
diff --git a/pkg/system/lprint.x b/pkg/system/lprint.x
new file mode 100644
index 00000000..3b61328c
--- /dev/null
+++ b/pkg/system/lprint.x
@@ -0,0 +1,213 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<syserr.h>
+include	<error.h>
+include	<finfo.h>
+include	<time.h>
+include	<ttyset.h>
+
+define	WIDE_PAGE	100
+define	SZ_DDSTR	256
+
+
+# LPRINT -- Print a file or files on the lineprinter.  A one line header is
+# printed at the top of each page, unless we are reading from the standard
+# input.  If printing multiple files, each file begins on a new page.
+# TTYPUTLINE is used to output lines to the printer file, so that formfeeds,
+# standout mode, etc., may be properly translated for the indicated device.
+
+procedure t_lprint()
+
+int	out, list, map_cc
+bool	input_redirected, one_file
+bool	print_heading, auto_header
+pointer	sp, fname, device, label, lp, ddstr
+
+pointer	ttyodes()
+bool	streq(), clgetb()
+int	clpopni(), clplen(), envgets(), lpopen(), clgfil(), ttygets(), btoi()
+string	printer "printer"
+errchk	clgfil, lp_print_file
+
+begin
+	call smark (sp)
+	call salloc (fname,  SZ_FNAME, TY_CHAR)
+	call salloc (label,  SZ_FNAME, TY_CHAR)
+	call salloc (device, SZ_FNAME, TY_CHAR)
+	call salloc (ddstr,  SZ_DDSTR, TY_CHAR)
+
+	# Open list of files to be printed.
+	list = clpopni ("files")
+	one_file = (clplen (list) == 1)
+
+	# Get device name.  Default is "printer", which means that the actual
+	# device name is given by the environment variable "printer".
+
+	call clgstr ("device", Memc[device], SZ_FNAME)
+	if (streq (Memc[device], printer))
+	    if (envgets (printer, Memc[device], SZ_FNAME) <= 0) {
+		call clpcls (list)
+		call syserrs (SYS_ENVNF, printer)
+	    }
+
+	# Map unknown control chars into printable sequences?
+	map_cc = btoi (clgetb ("map_cc"))
+
+	# If automatic pagination is selected, lprint will break pages and
+	# print headers only if the standard input is not redirected.
+	# Otherwise, pagination is disabled unless paginate=yes.  The file
+	# name buffer is used as a temporary to hold the value string.
+
+	call clgstr ("paginate", Memc[fname], SZ_FNAME)
+	auto_header = streq (Memc[fname], "auto")
+	if (!auto_header)
+	    print_heading = streq (Memc[fname], "yes")
+
+	# Open TTY descriptor and printer file.  We deal only with character
+	# data, so open printer as a text file.  Output the files to the line
+	# printer device.
+
+	lp = ttyodes (Memc[device])
+	if (ttygets (lp, "DD", Memc[ddstr], SZ_DDSTR) <= 0)
+	    call error (1, "missing 'DD' parameter in termcap entry")
+
+	out = lpopen (Memc[ddstr], APPEND, TEXT_FILE)
+
+	while (clgfil (list, Memc[fname], SZ_FNAME) != EOF) {
+	    # If only one file and it is the standard input (i.e., reading
+	    # from a pipe), and if header printing is not explicitly enabled
+	    # or disabled, do not break pages and print headers.
+
+	    input_redirected = (one_file && streq (Memc[fname], "STDIN"))
+	    if (auto_header)
+		print_heading = !input_redirected
+
+	    # Get page header label string.  File name is used unless reading
+	    # from the standard input.
+
+	    if (input_redirected)
+		call clgstr ("label", Memc[label], SZ_FNAME)
+	    else
+		call strcpy (Memc[fname], Memc[label], SZ_FNAME)
+
+	    # Output file, followed by a form feed.
+	    iferr {
+		call lp_print_file (out, lp, Memc[fname], Memc[label], map_cc,
+		    print_heading)
+		call flush (out)
+	    } then
+		call erract (EA_WARN)
+	}
+
+	call close (out)
+	call clpcls (list)
+	call ttycdes (lp)
+	call sfree (sp)
+end
+
+
+# LP_PRINT_FILE -- Open and print the named text file on the output file,
+# under control of the tty device descriptor LP.  Print a header on each
+# page if enabled, and formfeed at the end of the file.  The number of lines
+# per page is given by the tty descriptor.
+
+procedure lp_print_file (out, lp, fname, label, map_cc, print_heading)
+
+int	out
+pointer	lp
+char	fname[ARB]
+char	label[ARB]
+int	map_cc
+bool	print_heading
+
+bool	one_tab_in, streq()
+int	pageno, lineno, maxlines, totlines, status, in
+long	fi[LEN_FINFO], time, clktime()
+pointer	sp, ip, lbuf, timebuf
+int	open(), getline(), finfo(), ttystati()
+errchk	salloc, finfo, cnvtime, open, ttystati, getline, ttyputline
+
+begin
+	call smark (sp)
+	call salloc (lbuf, SZ_LINE+1, TY_CHAR)
+	call salloc (timebuf, SZ_TIME, TY_CHAR)
+
+	# Get time and date file was last modified.
+	if (print_heading) {
+	    if (streq (fname, "STDIN"))
+		time = clktime (long(0))
+	    else {
+		if (finfo (fname, fi) == ERR) {
+		    call sprintf (Memc[lbuf], SZ_LINE, "%s '%s'\n")
+			call pargstr ("Cannot get info on file")
+			call pargstr (fname)
+		    call error (1, Memc[lbuf])
+		} else
+		    time = FI_MTIME(fi)
+	    }
+	    call cnvtime (time, Memc[timebuf], SZ_TIME)
+	}
+
+	in = open (fname, READ_ONLY, TEXT_FILE)
+	maxlines = ttystati (lp, TTY_NLINES)
+
+	# If printer page is very wide, set page in one tabstop from left
+	# margin.
+
+	one_tab_in = (ttystati (lp, TTY_NCOLS) > WIDE_PAGE)
+	if (one_tab_in) {
+	    Memc[lbuf] = '\t'
+	    ip = lbuf + 1
+	} else
+	    ip = lbuf
+
+	# If printing header and breaking pages, allow for the 3 line header
+	# and a three line border at the bottom of the page.
+
+	if (print_heading)
+	    maxlines = maxlines - 6
+
+	totlines = 0
+	status = OK
+
+	# Format and write each page of output.  If headers are enabled,
+	# output formfeed and header between pages.
+
+	for (pageno=1;  status != EOF;  pageno=pageno+1) {
+	    # Print header, if enabled.  We assume that we are already
+	    # positioned to the top of a page.
+
+	    if (print_heading) {
+		call sprintf (Memc[ip], SZ_LINE, "%s  %s  Page %d\n\n\n")
+		    call pargstr (Memc[timebuf])
+		    call pargstr (label)
+		    call pargi (pageno)
+		call ttyputline (out, lp, Memc[lbuf], map_cc)
+		totlines = totlines + 3
+	    }
+
+	    # Output one page of text.  Each output line is processed by
+	    # ttyputline to expand tabs, underline, etc.
+
+	    for (lineno=1;  lineno <= maxlines;  lineno=lineno+1) {
+		status = getline (in, Memc[ip])
+		if (status == EOF)
+		    break
+		if (Memc[ip] == '\f') {
+		    call ttyputline (out, lp, "\f", map_cc)
+		    call strcpy (Memc[ip+1], Memc[ip], SZ_LINE)
+		}
+		call ttyputline (out, lp, Memc[lbuf], map_cc)
+		totlines = totlines + 1
+	    }
+
+	    # Output formfeed, leaving printer positioned to top of page.
+	    # Do not break pages if headers are disabled.
+
+	    if (print_heading && totlines > 0)
+		call ttyputline (out, lp, "\f", map_cc)
+	}
+
+	call close (in)
+	call sfree (sp)
+end
diff --git a/pkg/system/lroff.par b/pkg/system/lroff.par
new file mode 100644
index 00000000..512769eb
--- /dev/null
+++ b/pkg/system/lroff.par
@@ -0,0 +1,4 @@
+input_file,f,a,,,,file containing help text
+lmargin,i,h,1,1,160,left margin
+rmargin,i,h,72,1,160,right margin
+format,s,h,"text","|text|html|ps|postscript|",,format for output
diff --git a/pkg/system/match.par b/pkg/system/match.par
new file mode 100644
index 00000000..fa2670b3
--- /dev/null
+++ b/pkg/system/match.par
@@ -0,0 +1,5 @@
+pattern,s,a,,,,text pattern to be matched
+files,s,a,,,,files to be searched for the pattern
+stop,b,h,no,,,"stop, rather than pass, matched lines?"
+print_file_names,b,h,yes,,,print file names if multiple files?
+metacharacters,b,h,yes,,,recognize pattern matching metacharacters?
diff --git a/pkg/system/match.x b/pkg/system/match.x
new file mode 100644
index 00000000..b9f8feb2
--- /dev/null
+++ b/pkg/system/match.x
@@ -0,0 +1,96 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<pattern.h>
+include	<ctype.h>
+
+# MATCH -- Search for the indicated pattern in each line of a file or files,
+# passing or stopping only those lines which contain the pattern.
+
+procedure t_match()
+
+bool	stop_matched_lines		# pass or stop matched lines?
+char	chset[2]
+int	fd, list
+pointer	sp, ip, line, fname, patbuf, pattern
+bool	line_matches, encode_pattern, print_file_names, match_pattern
+bool	clgetb()
+int	strmatch(), strsearch(), patmatch(), patmake(), stridxs()
+int	open(), getline(), clpopni(), clplen(), clgfil()
+
+begin
+	call smark (sp)
+	call salloc (line, SZ_LINE, TY_CHAR)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+	call salloc (patbuf, SZ_LINE, TY_CHAR)
+	call salloc (pattern, SZ_FNAME, TY_CHAR)
+
+	# Get pattern.  If it contains either * or [, general pattern
+	# matching with an encoded pattern must be used.  If no metacharacters
+	# at all are used, or if metacharacters are disabled, use STRSEARCH
+	# for maximum efficiency.
+
+	call clgstr ("pattern", Memc[pattern], SZ_FNAME)
+	if (clgetb ("metacharacters")) {
+
+	    # Check for the tough pattern matching metachars.
+	    chset[1] = CH_CLOSURE
+	    chset[2] = CH_CCL
+	    chset[3] = EOS
+	    encode_pattern = (stridxs (chset, Memc[pattern]) > 0)
+
+	    # If don't have to use patmatch, see if we have a simple alphanum
+	    # string so that we can use strsearch.
+
+	    if (!encode_pattern) {
+		for (ip=pattern;  IS_ALNUM(Memc[ip]);  ip=ip+1)
+		    ;
+		match_pattern = (Memc[ip] != EOS)
+	    }
+
+	} else {
+	    encode_pattern = false
+	    match_pattern = false
+	}
+
+	stop_matched_lines = clgetb ("stop")
+	list = clpopni ("files")
+	print_file_names = (clplen(list) > 1) && (clgetb("print_file_names"))
+
+	if (encode_pattern)
+	    if (patmake (Memc[pattern], Memc[patbuf], SZ_LINE) == ERR)
+		call error (1, "Pattern is too complex")
+
+	# Search each file in the list, passing a line on to the output
+	# only if the line matches and stop is false, or if the line does
+	# not match and stop is true.
+
+	while (clgfil (list, Memc[fname], SZ_FNAME) != EOF) {
+	    fd = open (Memc[fname], READ_ONLY, TEXT_FILE)
+
+	    while (getline (fd, Memc[line]) != EOF) {
+		if (encode_pattern)
+		    line_matches = (patmatch (Memc[line], Memc[patbuf]) > 0)
+		else if (match_pattern)
+		    line_matches = (strmatch (Memc[line], Memc[pattern]) > 0)
+		else
+		    line_matches = (strsearch (Memc[line], Memc[pattern]) > 0)
+
+		if ((line_matches && !stop_matched_lines) ||
+		    (!line_matches && stop_matched_lines)) {
+
+		    if (print_file_names) {
+			call printf ("%s:")
+			    call pargstr (Memc[fname])
+		    }
+
+		    call putline (STDOUT, Memc[line])
+		    call flush (STDOUT)
+		}
+	    }
+
+	    call close (fd)
+	}
+
+	call sfree (sp)
+	call clpcls (list)
+end
diff --git a/pkg/system/mkdir.par b/pkg/system/mkdir.par
new file mode 100644
index 00000000..b2cbe135
--- /dev/null
+++ b/pkg/system/mkdir.par
@@ -0,0 +1 @@
+newdir,s,a,,,,name of the new directory
diff --git a/pkg/system/mkdir.x b/pkg/system/mkdir.x
new file mode 100644
index 00000000..a1711d51
--- /dev/null
+++ b/pkg/system/mkdir.x
@@ -0,0 +1,12 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# MKDIR -- Make a directory. (May be an IRAF VFN or an OS directory spec)
+
+procedure t_mkdir ()
+
+char	newdir[SZ_FNAME]
+
+begin
+	call clgstr ("newdir", newdir, SZ_FNAME)
+	call fmkdir (newdir)
+end
diff --git a/pkg/system/mkhelpdb.par b/pkg/system/mkhelpdb.par
new file mode 100644
index 00000000..0c1139f5
--- /dev/null
+++ b/pkg/system/mkhelpdb.par
@@ -0,0 +1,3 @@
+helpdir,s,a,"lib$root.hd",,,"root help directory to be compiled"
+helpdb,s,a,"lib$helpdb.mip",,,"name of help database to be created"
+verbose,b,h,no,,,"detail the structure of the package hierarchy"
diff --git a/pkg/system/mkpkg b/pkg/system/mkpkg
new file mode 100644
index 00000000..1392b3c6
--- /dev/null
+++ b/pkg/system/mkpkg
@@ -0,0 +1,53 @@
+# Make the SYSTEM system utilities package (including HELP)
+
+$call	relink
+$exit
+
+update:
+	$call	relink
+	$call	install
+	;
+
+relink:
+	$update	libpkg.a
+	$omake	x_system.x
+	$link	x_system.o libpkg.a -o xx_system.e
+	;
+
+install:
+	$move	xx_system.e bin$x_system.e
+	;
+
+libpkg.a:
+	@help
+
+	cmdstr.x	<ctype.h>
+	chkupdate.x	<fio.h> <fset.h> <finfo.h> <syserr.h>
+	concatenate.x	
+	copy.x		<error.h>
+	count.x		<chars.h> <ctype.h> <error.h>
+	delete.x	<error.h>
+	directory.x	<chars.h> <ctype.h> <diropen.h> <finfo.h>\
+			<fset.h> <protect.h> <time.h>
+	files.x	
+	head.x		<error.h>
+	lprint.x	<error.h> <finfo.h> <time.h> <ttyset.h>
+	match.x		<ctype.h> <pattern.h>
+	mkdir.x	
+	movefiles.x	<error.h>
+	mtclean.x	
+	netstatus.x	
+	page.x		<fset.h>
+	pathnames.x	
+	protect.x	<error.h> <protect.h>
+	rename.x	<error.h>
+	rewind.x	
+	sort.x		sort.com <ctype.h>
+	tail.x		<error.h>
+	tee.x		<fset.h>
+	touch.x 	<error.h> <finfo.h> <time.h>
+	type.x		<error.h>
+	unprotect.x	<error.h> <protect.h>
+        t_fcache.x      <error.h>
+	t_urlget.x	<error.h> <imhdr.h> <imset.h> <mach.h>
+	;
diff --git a/pkg/system/mkscript.cl b/pkg/system/mkscript.cl
new file mode 100644
index 00000000..984cdf27
--- /dev/null
+++ b/pkg/system/mkscript.cl
@@ -0,0 +1,79 @@
+#{ MKBATCH -- Make a batch script.
+
+{
+	# Make a copy of the script file query parameter and delete the
+	# file if it exists and not appending.
+
+	scrpt = script
+	if (access (scrpt) && !append)
+	     delete (scrpt, verify=verify)
+
+	# Outer loop provides return when the user wants to start over.
+
+	while (YES) {
+
+	    # Add commands until the user is done.
+
+	    while (YES) {
+
+		# Determine the task name and check that it is loaded.
+
+		ltask = task
+		if (!deftask (ltask)) {
+		    print ("Task ", ltask, " not loaded.")
+		    next
+		}
+
+		# Edit the task parameters.
+
+	        eparam (ltask)
+		print ("")
+
+		# If the user is verifying then display the command.
+
+	        if (verify) {
+	            lparam (ltask) | cmdstr (ltask, hidden=hidden)
+	            cok = YES
+		    if (cok) {
+		        lparam (ltask) | cmdstr (ltask, hidden=hidden, >> scrpt)
+			print ("", >> scrpt)
+		    }
+	        } else {
+		    lparam (ltask) | cmdstr (ltask, hidden=hidden, >> scrpt)
+		    print ("", >> scrpt)
+		}
+
+		# Ask if more commands are to be added to the script.
+
+		more = YES
+	        if (!more)
+		    break
+	    }
+
+	    # If verifying page the script and let the user decide to start
+	    # over.
+
+	    if (verify) {
+		page (scrpt)
+		sok = YES
+		if (sok)
+		    break
+		else
+		    delete (scrpt, verify=verify)
+	    } else
+		break
+	}
+
+	# If the user does not want to submit the script then quit.
+
+	if (!submit)
+    	    bye
+}
+
+# Submit the script as a background process.
+
+print ("Script ", scrpt, " submitted at:")
+time
+print ("\nScript ", scrpt, " submitted at:", >> logfile)
+time (>> logfile)
+cl (< scrpt, >>& logfile) &
diff --git a/pkg/system/mkscript.par b/pkg/system/mkscript.par
new file mode 100644
index 00000000..46afb1b6
--- /dev/null
+++ b/pkg/system/mkscript.par
@@ -0,0 +1,17 @@
+script,s,a,script.cl,,,Script file name
+task,s,a,,,,Task name of command to be added to script
+submit,b,a,yes,,,Submit the script as a background job?
+
+append,b,h,no,,,Append new to existing script?
+hidden,b,h,yes,,,Include hidden parameters in command?
+verify,b,h,yes,,,Verify each command?
+logfile,f,h,script.log,,,Script log file name
+
+# Script query parameters and variables.
+
+cok,b,a,yes,,,Is the command ok?
+sok,b,a,yes,,,Is the script ok?
+more,b,a,yes,,,Add another command?
+
+scrpt,s,h
+ltask,s,h
diff --git a/pkg/system/movefiles.par b/pkg/system/movefiles.par
new file mode 100644
index 00000000..1aa0cc27
--- /dev/null
+++ b/pkg/system/movefiles.par
@@ -0,0 +1,3 @@
+files,s,a,,,,list of files to be moved
+newdir,f,a,,,,name of destination directory
+verbose,b,h,no,,,print names of files as they are moved
diff --git a/pkg/system/movefiles.x b/pkg/system/movefiles.x
new file mode 100644
index 00000000..32842f61
--- /dev/null
+++ b/pkg/system/movefiles.x
@@ -0,0 +1,52 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+
+# MOVEFILES -- Move a set of files to another directory.  If the destination
+# is not a directory it is a fatal error.  Ambiguous directory specifications
+# are resolved in favor of subdirectories, rather than logical directories.
+
+procedure t_movefiles()
+
+bool	verbose
+int	list, root_len
+pointer	sp, fname, newdir, pathname, newname, junkstr
+bool	clgetb()
+int	clpopni(), clgfil(), fnldir(), isdirectory()
+
+begin
+	call smark (sp)
+	call salloc (fname, SZ_FNAME, TY_CHAR)
+	call salloc (newdir, SZ_FNAME, TY_CHAR)
+	call salloc (pathname, SZ_PATHNAME, TY_CHAR)
+	call salloc (newname, SZ_PATHNAME, TY_CHAR)
+	call salloc (junkstr, SZ_FNAME, TY_CHAR)
+
+	list = clpopni ("files")
+	call clgstr ("newdir", Memc[newdir], SZ_FNAME)
+	verbose = clgetb ("verbose")
+
+	if (isdirectory (Memc[newdir], Memc[pathname], SZ_PATHNAME) == 0) {
+	    call strcat ("$", Memc[newdir], SZ_FNAME)
+	    if (isdirectory (Memc[newdir], Memc[pathname], SZ_PATHNAME) == 0)
+		call error (1, "destination is not a directory")
+	}
+
+	while (clgfil (list, Memc[fname], SZ_FNAME) != EOF) {
+	    call strcpy (Memc[pathname], Memc[newname], SZ_PATHNAME)
+	    root_len = fnldir (Memc[fname], Memc[junkstr], SZ_FNAME)
+	    call strcat (Memc[fname+root_len], Memc[newname], SZ_PATHNAME)
+
+	    if (verbose) {
+		call eprintf ("%s -> %s\n")
+		    call pargstr (Memc[fname])
+		    call pargstr (Memc[newname])
+	    }
+
+	    iferr (call rename (Memc[fname], Memc[newname]))
+		call erract (EA_WARN)
+	}
+
+	call clpcls (list)
+	call sfree (sp)
+end
diff --git a/pkg/system/mtclean.par b/pkg/system/mtclean.par
new file mode 100644
index 00000000..d30e2b82
--- /dev/null
+++ b/pkg/system/mtclean.par
@@ -0,0 +1,3 @@
+all,b,h,no,,,delete all lok files unconditionally
+stale,i,h,3600,,,threshold (seconds) after which old lok files are deleted
+verbose,b,h,no,,,print names of lok files as they are deleted
diff --git a/pkg/system/mtclean.x b/pkg/system/mtclean.x
new file mode 100644
index 00000000..d79359c5
--- /dev/null
+++ b/pkg/system/mtclean.x
@@ -0,0 +1,25 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# MTCLEAN -- Clean the magtape .lok file area, i.e., delete any old .lok files.
+# These files no longer serve any device locking purpose, rather, they are
+# used by the magtape system to store knowledge of the tape position when
+# the drive is not being accessed.  The mtclean task is called during CL
+# startup to delete any old .lok files to keep these from being erroneously
+# used to indicate the tape position.  In normal use the .lok files are
+# created when a tape is allocated and deleted when the tape is deallocated,
+# but the files can be left behind if the CL is killed without doing a
+# deallocate.
+
+procedure t_mtclean()
+
+int	out
+bool	clgetb()
+int	clgeti(), btoi()
+
+begin
+	out = NULL
+	if (clgetb ("verbose"))
+	    out = STDOUT
+
+	call mtclean (btoi(clgetb("all")), clgeti("stale"), out)
+end
diff --git a/pkg/system/netstatus.x b/pkg/system/netstatus.x
new file mode 100644
index 00000000..b7ecb225
--- /dev/null
+++ b/pkg/system/netstatus.x
@@ -0,0 +1,9 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# NETSTATUS -- Print the network status.
+
+procedure t_netstatus()
+
+begin
+	call ki_shownet (STDOUT)
+end
diff --git a/pkg/system/news.cl b/pkg/system/news.cl
new file mode 100644
index 00000000..ee3e88ec
--- /dev/null
+++ b/pkg/system/news.cl
@@ -0,0 +1,5 @@
+#{ NEWS -- Page the system news file.
+
+{
+	page "doc$newsfile"
+}
diff --git a/pkg/system/page.par b/pkg/system/page.par
new file mode 100644
index 00000000..2419bf06
--- /dev/null
+++ b/pkg/system/page.par
@@ -0,0 +1,6 @@
+files,s,a,,,,Files to be paged
+map_cc,b,h,yes,,,Map control characters into printable sequences?
+clear_screen,b,h,yes,,,Clear screen between pages?
+first_page,i,h,1,1,,First page of file to be displayed
+prompt,s,h,"",,,Prompt string if not filename
+device,s,h,"terminal",,,Output device
diff --git a/pkg/system/page.x b/pkg/system/page.x
new file mode 100644
index 00000000..208bb56b
--- /dev/null
+++ b/pkg/system/page.x
@@ -0,0 +1,41 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<fset.h>
+
+# PAGE -- Display a text file or files on the standard output (the user
+# terminal) one screen at a time, pausing after each screen has been filled.
+# The program is keystroke driven in raw mode, and currently recognizes the
+# keystrokes defined above.
+
+procedure t_page()
+
+bool	redirin
+pointer	sp, device, prompt, files
+int	map_cc, clear_screen, first_page
+
+bool	clgetb()
+int	fstati(), clgeti(), btoi()
+
+begin
+	call smark (sp)
+	call salloc (device, SZ_FNAME, TY_CHAR)
+	call salloc (prompt, SZ_FNAME, TY_CHAR)
+	call salloc (files, SZ_LINE, TY_CHAR)
+
+	redirin = (fstati (STDIN, F_REDIR) == YES)
+	if (redirin)
+	    call strcpy ("STDIN", Memc[files], SZ_LINE)
+	else
+	    call clgstr ("files", Memc[files], SZ_LINE)
+
+	map_cc = btoi (clgetb ("map_cc"))
+	clear_screen = btoi (clgetb ("clear_screen"))
+	first_page = clgeti ("first_page")
+	call clgstr ("prompt", Memc[prompt], SZ_FNAME)
+	call clgstr ("device", Memc[device], SZ_FNAME)
+
+	call xpagefiles (Memc[files], Memc[device], Memc[prompt],
+	    first_page, clear_screen, map_cc)
+
+	call sfree (sp)
+end
diff --git a/pkg/system/pathnames.par b/pkg/system/pathnames.par
new file mode 100644
index 00000000..20e6a230
--- /dev/null
+++ b/pkg/system/pathnames.par
@@ -0,0 +1,2 @@
+template,s,a,,,,file template
+sort,b,h,yes,,,sort file list
diff --git a/pkg/system/pathnames.x b/pkg/system/pathnames.x
new file mode 100644
index 00000000..86ee74f4
--- /dev/null
+++ b/pkg/system/pathnames.x
@@ -0,0 +1,55 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# PATHNAMES -- Convert one or more filenames into OS pathnames.  If the input
+# is redirected, filenames are read from the standard input.  Otherwise,
+# a file name template is expanded and used as the list of filenames to be
+# transformed.  If called without any arguments, return the pathname of the
+# current working directory.
+
+procedure t_pathnames()
+
+int	list
+pointer	sp, fname, osfname
+bool	streq(), clgetb()
+int	clgeti(), clpopni(), clpopnu(), clgfil(), fscan()
+
+begin
+	call smark (sp)
+	call salloc (fname, SZ_LINE, TY_CHAR)
+	call salloc (osfname, SZ_LINE, TY_CHAR)
+
+	# If no arguments, return the pathame of the current
+	# directory: do not prompt for the template.
+
+	if (clgeti ("$nargs") == 0) {
+	    call fpathname ("", Memc[osfname], SZ_LINE)
+	    call printf ("%s\n")
+		call pargstr (Memc[osfname])
+	    call sfree (sp)
+	    return
+	}
+
+	# Expand template, output pathname of each file therein.
+	if (clgetb ("sort"))
+	    list = clpopni ("template")
+	else
+	    list = clpopnu ("template")
+
+	while (clgfil (list, Memc[fname], SZ_FNAME) != EOF) {
+	    if (streq (Memc[fname], "STDIN")) {
+		while (fscan (STDIN) != EOF) {
+		    call gargstr (Memc[fname], SZ_LINE)
+		    call fpathname (Memc[fname], Memc[osfname], SZ_LINE)
+		    call printf ("%s\n")
+			call pargstr (Memc[osfname])
+		}
+	    } else {
+		call fpathname (Memc[fname], Memc[osfname], SZ_LINE)
+		call printf ("%s\n")
+		    call pargstr (Memc[osfname])
+	    }
+	}
+
+	call clpcls (list)
+	call sfree (sp)
+end
diff --git a/pkg/system/phelp.cl b/pkg/system/phelp.cl
new file mode 100644
index 00000000..4de3eab2
--- /dev/null
+++ b/pkg/system/phelp.cl
@@ -0,0 +1,41 @@
+# PHELP -- Paged help.
+
+procedure phelp (template)
+
+string	template		{prompt="module name template"}
+
+bool	file_template=no	{prompt="print help file"}
+bool	all=yes			{prompt="find all modules matching template"}
+string	parameter="all"		{prompt="parameter for which help is desired"}
+string	section="all"		{prompt="section for which help is desired"}
+string	option="help"		{prompt="type of help desired"}
+int	lmargin=1		{min=1, prompt="left margin"}
+int	rmargin=72		{min=2, prompt="right margin"}
+string	helpdb="helpdb"		{prompt="help database to be used"}
+
+begin
+	file	helptext
+	string	s_template
+
+	# Get a temp file to hold help text.
+	helptext = mktemp ("tmp$htx")
+	s_template = template
+
+	# Run HELP, redirecting the output to the temp file.
+	help (s_template, > helptext, page=no,
+	    all = all,
+	    file_template = file_template,
+	    parameter = parameter,
+	    section = section,
+	    option = option,
+	    lmargin = lmargin,
+	    rmargin = rmargin,
+	    device = "terminal",
+	    helpdb = helpdb)
+
+	# Page saved text output.
+	page (helptext, prompt = s_template)
+
+	# Delete temp file.
+	delete (helptext, verify-)
+end
diff --git a/pkg/system/protect.par b/pkg/system/protect.par
new file mode 100644
index 00000000..cb9721ac
--- /dev/null
+++ b/pkg/system/protect.par
@@ -0,0 +1 @@
+files,s,a,,,,list of files to be protected
diff --git a/pkg/system/protect.x b/pkg/system/protect.x
new file mode 100644
index 00000000..4ed12db9
--- /dev/null
+++ b/pkg/system/protect.x
@@ -0,0 +1,23 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<protect.h>
+include	<error.h>
+
+# PROTECT -- Protect a list of files.
+
+procedure t_protect()
+
+char	fname[SZ_FNAME]
+int	list, status
+
+int	clpopns(), clgfil(), protect()
+
+begin
+	list = clpopns ("files")
+
+	while (clgfil (list, fname, SZ_FNAME) != EOF)
+	    iferr (status = protect (fname, SET_PROTECTION))
+		call erract (EA_WARN)
+
+	call clpcls (list)
+end
diff --git a/pkg/system/references.cl b/pkg/system/references.cl
new file mode 100644
index 00000000..ca93e6a6
--- /dev/null
+++ b/pkg/system/references.cl
@@ -0,0 +1,50 @@
+# REFERENCES -- Search the help database for help on a given subject.
+# By default a runtime search is performed of all the package menus in the
+# help database, but the user may prepare a "quick-reference" file to speed
+# further searches if desired.  This cannot easily be done automatically
+# since any set of packages may form the help database.
+#
+# NOTE -- Uses LISTS.
+
+procedure references (topic)
+
+string	topic			     { prompt = "topic to find help for" }
+file	quickref = "uparm$quick.ref" { prompt = "quick-reference file" }
+bool	updquick = no		     { prompt = "update quick-reference file" }
+bool	usequick = no		     { prompt = "use quick-reference file" }
+
+begin
+	string	pattern
+	file	fname
+
+	# Make a quick-search file if so requested.
+	if (updquick) {
+	    fname = quickref
+	    print ("generating new quick-reference file " // fname // "...")
+	    if (access (fname))
+		delete (fname, verify-)
+	    help ("[a-z]*.", option="ref", curpack="AsckCL", device="terminal",
+		helpdb="helpdb") |& match ("-", metacharacters=yes) |
+		sort(ignore+) | unique ( > fname)
+	    references.quickref = fname
+	    references.usequick = yes
+
+	} else {
+	    # Ignore case.
+	    pattern = ("{" // topic // "}")
+
+	    # If the user has prepared a quick-search file (e.g., by running
+	    # this task with topic=* and saving the output to make the quick
+	    # file), search that if it exists, otherwise perform a runtime
+	    # search of the package menu files for the entire help database.
+
+	    if (usequick && access (quickref))
+		match (pattern, quickref, metacharacters=yes)
+	    else {
+		print ("searching the help database...")
+		help ("[a-z]*.", section="all", option="ref", curpack="AsckCL",
+		    device="terminal", helpdb="helpdb") |& sort(ignore+) | 
+		    unique | match (pattern, metacharacters=yes)
+	    }
+	}
+end
diff --git a/pkg/system/rename.par b/pkg/system/rename.par
new file mode 100644
index 00000000..b7e19293
--- /dev/null
+++ b/pkg/system/rename.par
@@ -0,0 +1,3 @@
+files,s,a,,,,list of files to be renamed
+newname,s,a,,,,new file name or field name
+field,s,h,all,"|all|ldir|root|extn|",,field to be modified (all|ldir|root|extn)
diff --git a/pkg/system/rename.x b/pkg/system/rename.x
new file mode 100644
index 00000000..95cf395c
--- /dev/null
+++ b/pkg/system/rename.x
@@ -0,0 +1,176 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+
+define	F_ALL		0
+define	F_LDIR		1
+define	F_ROOT		2
+define	F_EXTN		3
+
+
+# RENAME -- Rename a file, or rename a list of files.  In the latter case
+# the ldir, root, or extn field in each of the input files is changed to
+# the value of the "output file", which in this case is the value of the
+# specified field of an output file name.  Thus,
+#
+# 	cl> rename task.* newtask
+#
+# would rename task.x as newtask.x, task.cl as newtask.cl, and so on.
+# If 'newtask' is a logical or host directory the input files are moved to
+# this directory with the same name.
+
+procedure t_rename()
+
+bool	isdir
+pointer	sp, oldfile, newname, field
+pointer	o_ldir, o_root, o_extn, n_ldir, n_root, n_extn
+pointer	pathname, newfile, junkstr
+int	list, len, modfield
+
+bool	streq()
+int	clpopni(), clgfil(), clplen(), stridxs()
+int	access(), fnldir(), fnroot(), fnextn(), isdirectory()
+string	s_clobber "Warning: %s would overwrite existing file %s - skipping\n"
+
+begin
+	call smark (sp)
+	call salloc (oldfile, SZ_FNAME, TY_CHAR)
+	call salloc (newname, SZ_PATHNAME, TY_CHAR)
+	call salloc (pathname, SZ_PATHNAME, TY_CHAR)
+	call salloc (newfile, SZ_PATHNAME, TY_CHAR)
+	call salloc (junkstr, SZ_FNAME, TY_CHAR)
+	call salloc (field, SZ_FNAME, TY_CHAR)
+	call salloc (o_ldir, SZ_FNAME, TY_CHAR)
+	call salloc (n_ldir, SZ_FNAME, TY_CHAR)
+	call salloc (o_root, SZ_FNAME, TY_CHAR)
+	call salloc (n_root, SZ_FNAME, TY_CHAR)
+	call salloc (o_extn, SZ_FNAME, TY_CHAR)
+	call salloc (n_extn, SZ_FNAME, TY_CHAR)
+
+	# Open the list of files to be renamed.  This is done first so that
+	# the old file name is queried for before the new file name.
+
+	list = clpopni ("files")
+
+	# Get the output file name, or the new name for the field to be
+	# changed.
+
+	call clgstr ("newname", Memc[newname], SZ_PATHNAME)
+	call clgstr ("field", Memc[field], SZ_FNAME)
+
+	if (streq (Memc[field], "ldir"))
+	    modfield = F_LDIR
+	else if (streq (Memc[field], "root"))
+	    modfield = F_ROOT
+	else if (streq (Memc[field], "extn"))
+	    modfield = F_EXTN
+	else if (streq (Memc[field], "all"))
+	    modfield = F_ALL
+	else {
+	    call clpcls (list)
+	    call error (1, "rename: unrecognized filename field code")
+	}
+
+	# See if we're modifying the logical directory, If so, move the input
+	# files to this directory, otherwise do the renaming as we've always
+	# done.
+
+	Memc[pathname] = EOS
+	isdir = (isdirectory (Memc[newname], Memc[pathname], SZ_PATHNAME) > 0)
+
+	if (isdir || modfield == F_LDIR) {
+	    # Move each of the files in the list to the destination dir.
+	    call fdirname (Memc[newname], Memc[n_ldir], SZ_FNAME)
+
+	    while (clgfil (list, Memc[oldfile], SZ_FNAME) != EOF) {
+		if (Memc[pathname] != EOS)
+		    call strcpy (Memc[pathname], Memc[newfile], SZ_PATHNAME)
+		else
+		    call strcpy (Memc[n_ldir], Memc[newfile], SZ_PATHNAME)
+		len = fnldir (Memc[oldfile], Memc[junkstr], SZ_FNAME)
+		call strcat (Memc[oldfile+len], Memc[newfile], SZ_PATHNAME)
+
+		iferr (call rename (Memc[oldfile], Memc[newfile]))
+		    call erract (EA_WARN)
+	    }
+
+	} else if (modfield == F_ALL) {
+	    # "newname" is the new filename.  This makes sense only if there
+	    # is a single input filename.  Note that the new name may contain
+	    # a new logical directory, renaming both the file and moving it		    # to a new directory.
+
+	    if (clplen (list) > 1)
+		call error (2, "rename: `newname' must be a directory")
+
+	    # Rename the file.
+	    if (clgfil (list, Memc[oldfile], SZ_FNAME) != EOF)
+		iferr (call rename (Memc[oldfile], Memc[newname]))
+		    call erract (EA_WARN)
+
+	} else {
+	    # We're either modifying the root or the extension.  Break out
+	    # the ldir, root and extn for the input and output file names
+	    # then construct the new name from these components.
+
+	    Memc[n_root] = EOS
+	    Memc[n_extn] = EOS
+
+	    if (modfield == F_ROOT) {
+	        call strcpy (Memc[newname], Memc[n_root], SZ_FNAME)
+		if (stridxs ("$/", Memc[n_root]) > 0) {
+		    call clpcls (list)
+		    call error (3, "rename: bad replacement root field")
+		}
+	    }
+	    if (modfield == F_EXTN) {
+	        call strcpy (Memc[newname], Memc[n_extn], SZ_FNAME)
+		if (stridxs ("$/", Memc[n_extn]) > 0) {
+		    call clpcls (list)
+		    call error (4, "rename: bad replacement extn field")
+		}
+	    }
+
+	    # Process the files.
+	    while (clgfil (list, Memc[oldfile], SZ_FNAME) != EOF) {
+
+	        # Get the ldir, root and extension names of the old filename.
+	        len = fnroot (Memc[oldfile], Memc[o_root], SZ_FNAME)
+	        len = fnextn (Memc[oldfile], Memc[o_extn], SZ_FNAME)
+	        len = fnldir (Memc[oldfile], Memc[o_ldir], SZ_PATHNAME)
+
+		# Start by copying the ldir to the new name.
+		call aclrc (Memc[newname], SZ_PATHNAME)
+		call strcpy (Memc[o_ldir], Memc[newname], SZ_PATHNAME)
+
+		# Build up the new file name.
+	        if (modfield == F_ROOT) {
+		    call strcat (Memc[n_root], Memc[newname], SZ_FNAME)
+                    if (Memc[o_extn] != EOS) {
+		        call strcat (".", Memc[newname], SZ_PATHNAME)
+		        call strcat (Memc[o_extn], Memc[newname], SZ_FNAME)
+		    }
+	        } else if (modfield == F_EXTN) {
+		    call strcat (Memc[o_root], Memc[newname], SZ_FNAME)
+		    call strcat (".", Memc[newname], SZ_PATHNAME)
+		    call strcat (Memc[n_extn], Memc[newname], SZ_FNAME)
+	        }
+
+		# Check to see if we're going to clobber a file.
+		if (clplen (list) > 1) {
+		    if (access (Memc[newname], 0, 0) == YES) {
+			call eprintf (s_clobber)
+			    call pargstr (Memc[oldfile])
+			    call pargstr (Memc[newname])
+			next
+		    }
+		}
+
+		# Rename the file.
+	        iferr (call rename (Memc[oldfile], Memc[newname]))
+		    call erract (EA_WARN)
+	    }
+	}
+
+	call clpcls (list)
+	call sfree (sp)
+end
diff --git a/pkg/system/rewind.par b/pkg/system/rewind.par
new file mode 100644
index 00000000..714a56bf
--- /dev/null
+++ b/pkg/system/rewind.par
@@ -0,0 +1,2 @@
+device,f,a,mta,,,device to be rewound
+initcache,b,h,yes,,,discard cached information on tape contents
diff --git a/pkg/system/rewind.x b/pkg/system/rewind.x
new file mode 100644
index 00000000..9316b212
--- /dev/null
+++ b/pkg/system/rewind.x
@@ -0,0 +1,14 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# REWIND -- Rewind a (previously allocated) magtape device.
+
+procedure t_rewind()
+
+char	device[SZ_FNAME]
+bool	clgetb()
+int	btoi()
+
+begin
+	call clgstr ("device", device, SZ_FNAME)
+	call mtrewind (device, btoi(clgetb("initcache")))
+end
diff --git a/pkg/system/sort.com b/pkg/system/sort.com
new file mode 100644
index 00000000..4b66ac5f
--- /dev/null
+++ b/pkg/system/sort.com
@@ -0,0 +1,10 @@
+# Common for the SORT program.
+
+int	column			# column to be sorted (if nonzero)
+int	ignore_whitespace	# ignore leading whitespace
+int	numeric_sort		# sort numerically rather than alphabetically
+int	reverse_sort		# reverse the sense of the sort
+int	use_strsrt		# ok to use standard STRSRT routine?
+
+common	/srtcom/ column, ignore_whitespace, numeric_sort, reverse_sort,
+	use_strsrt
diff --git a/pkg/system/sort.par b/pkg/system/sort.par
new file mode 100644
index 00000000..84d721a3
--- /dev/null
+++ b/pkg/system/sort.par
@@ -0,0 +1,5 @@
+input_file,s,a,,,,file to be sorted
+column,i,h,0,,,column to be sorted (if nonzero)
+ignore_whitespace,b,h,no,,,ignore leading whitespace
+numeric_sort,b,h,no,,,sort numerically rather than alphabetically
+reverse_sort,b,h,no,,,reverse the sense of the sort
diff --git a/pkg/system/sort.x b/pkg/system/sort.x
new file mode 100644
index 00000000..e53b5fd8
--- /dev/null
+++ b/pkg/system/sort.x
@@ -0,0 +1,434 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<ctype.h>
+
+# SORT -- Sort a text file, alphabetically or numerically, in forward or
+# reverse order, by line or by the contents of any column.  Derived from
+# Chap. 4 of Software Tools.  This is old code, not very pretty.
+
+define	MERGEORDER	8
+define	MAXPTR		20000
+define	SZ_LINBUF	200000
+define	LOGPTR		32
+define	swap		{temp=$1;$1=$2;$2=temp}
+
+
+# SORT -- The main CL callable routine.
+
+procedure t_sort()
+
+pointer	linbuf, linptr
+int	infil[MERGEORDER], nlines
+char	name[SZ_FNAME], source_file[SZ_FNAME]
+int	high, lim, low, fd, outfil, t, list, junk
+
+bool	clgetb()
+int	ss_gtext(), ss_mkfile()
+int	open(), clpopni(), clplen(), clgfil(), clgeti(), btoi()
+include	"sort.com"
+
+begin
+	list = clpopni ("input_file")
+	if (clplen (list) > 1) {
+	    call clpcls (list)
+	    call error (1, "Cannot yet sort more than one file at a time")
+	}
+
+	# Determine type of sort to be applied.
+	column = clgeti ("column")
+	numeric_sort = btoi (clgetb ("numeric_sort"))
+	reverse_sort = btoi (clgetb ("reverse_sort"))
+
+	ignore_whitespace = NO
+	if (column == 0 && numeric_sort == NO)
+	    if (clgetb ("ignore_whitespace"))
+		ignore_whitespace = YES
+
+	# The standard STRSRT is fastest if we can use it.
+	use_strsrt = NO
+	if (ignore_whitespace == NO && numeric_sort == NO &&
+	    reverse_sort == NO && column == 0)
+	    use_strsrt = YES
+
+	# Allocate buffer space.
+	call malloc (linbuf, SZ_LINBUF, TY_CHAR)
+	call malloc (linptr, MAXPTR, TY_INT)
+
+	# Perform the sort. 
+	junk = clgfil (list, source_file, SZ_FNAME)
+	fd = open (source_file, READ_ONLY, TEXT_FILE)
+
+	# Initial formation of runs.
+	high = 0
+	repeat {
+	    t = ss_gtext (fd, Memi[linptr], nlines, Memc[linbuf])
+	    if (use_strsrt == YES)
+		call strsrt (Memi[linptr], Memc[linbuf], nlines)
+	    else
+		call ss_quick (Memi[linptr], Memc[linbuf], nlines)
+
+	    high = high + 1
+	    outfil = ss_mkfile (high)
+	    call ss_ptext (outfil, Memi[linptr], nlines, Memc[linbuf])
+	    call close (outfil)
+
+	} until (t == EOF)
+
+	# Merge sorted chunks.
+	for (low = 1;  low < high;  low = low + MERGEORDER) {
+	    lim = min (low + MERGEORDER - 1, high)
+	    call ss_gopen (infil, low, lim)
+	    high = high + 1
+	    outfil = ss_mkfile (high)
+	    call ss_merge (infil, outfil, lim - low + 1)
+	    call close (outfil)
+	    call ss_gremov (infil, low, lim)
+	}
+
+	call mfree (linbuf, TY_CHAR)
+	call mfree (linptr, TY_INT)
+
+	call ss_gname (high, name)		# final cleanup
+	outfil = open (name, READ_ONLY, TEXT_FILE)
+	call fcopyo (outfil, STDOUT)
+	call close (outfil)
+
+	call delete (name)
+	call close (fd)
+end
+
+
+# SS_MKFILE -- Open a temporary file.
+
+int procedure ss_mkfile (n)
+
+int	n
+char	name[SZ_FNAME]
+int	open()
+errchk	open, ss_gname
+
+begin
+	call ss_gname (n, name)
+	return (open (name, NEW_FILE, TEXT_FILE))
+end
+
+
+# SS_GNAME -- Quickie routine to generate a temporary file name in tmp$ (the
+# original tools version would put the temp files in the cwd).  Make a name
+# of the form "tmp$srtPID_XX", where XX is the file number argument N,
+# and PID is the process pid.  The tools sort code requires that the file
+# name be easily reproduced given only N, so we cannot use mktemp.
+
+procedure ss_gname (n, name)
+
+int	n
+char	name[ARB]
+int	pid
+
+begin
+	call zgtpid (pid)
+	call sprintf (name, SZ_FNAME, "tmp$srt%d_%d")
+	    call pargi (pid)
+	    call pargi (n)
+end
+
+
+# SS_GOPEN -- Open a range of files.
+
+procedure ss_gopen (infil, low, lim)
+
+int	infil[ARB], lim, low
+char	name[SZ_FNAME]
+int	i, open()
+errchk	open, ss_gname
+
+begin
+	for (i=1;  i <= lim-low+1;  i=i+1) {
+	    call ss_gname (low+i-1, name)
+	    infil[i] = open (name, READ_ONLY, TEXT_FILE)
+	}
+end
+
+
+# SS_GREMOV -- Remove a range of files.
+
+procedure ss_gremov (infil, low, lim)
+
+int	infil[ARB], low, lim
+
+int	i
+char	name[SZ_FNAME]
+errchk	ss_gname, delete, close
+
+begin
+	for (i=1;  i <= lim-low+1;  i=i+1) {
+	    call close (infil[i])
+	    call ss_gname (low+i-1, name)
+	    call delete (name)
+	}
+end
+
+
+# SS_MERGE -- Merge file onto outfil.
+
+procedure ss_merge (infil, outfil, nfiles)
+
+int	infil[ARB]		# input file numbers
+int	outfil			# output file number
+int	nfiles			# number of files to be merged
+
+pointer	sp, linbuf
+int	linptr[MERGEORDER]
+int	i, inf, lbp, lp1, nf
+int	getline()
+errchk	getline, putline, ss_quick
+include	"sort.com"
+
+begin
+	call smark (sp)
+	call salloc (linbuf, MERGEORDER * SZ_LINE, TY_CHAR)
+
+	lbp = 1
+	nf = 0
+	
+	# Get one line from each file.
+	for (i=1;  i <= nfiles;  i=i+1) {
+	    if (getline (infil[i], Memc[linbuf+lbp-1]) != EOF) {
+		nf = nf + 1
+		linptr[nf] = lbp
+		lbp = lbp + SZ_LINE
+	    }
+	}
+
+	# Make initial heap.
+	if (use_strsrt == YES)
+	    call strsrt (linptr, Memc[linbuf], nf)
+	else
+	    call ss_quick (linptr, Memc[linbuf], nf)
+
+	while (nf > 0) {
+	    lp1 = linptr[1]
+	    call putline (outfil, Memc[linbuf+lp1-1])
+	    inf = lp1 / SZ_LINE + 1		# compute file index
+	    if (getline (infil[inf], Memc[linbuf+lp1-1]) == EOF) {
+		linptr[1] = linptr[nf]
+		nf = nf - 1
+	    }
+	    call ss_reheap (linptr, Memc[linbuf], nf)
+	}
+
+	call sfree (sp)
+end
+
+
+# SS_REHEAP -- propagate linbuf[linptr[1]] to proper place in heap.
+
+procedure ss_reheap (linptr, linbuf, nf)
+
+int	linptr[ARB]
+char	linbuf[ARB]
+int	nf
+
+int	i, j, temp
+int	ss_compare()
+
+begin
+	for (i=1;  2 * i <= nf;  i=j) {
+	    j = 2 * i
+	    if (j < nf)			# find smaller child
+		if (ss_compare (linptr[j], linptr[j+1], linbuf) > 0)
+		    j = j + 1
+	    if (ss_compare (linptr[i], linptr[j], linbuf) <= 0)
+		break			# proper position found
+
+	    swap (linptr[i], linptr[j])
+	}
+end
+
+
+# SS_QUICK -- Quicksort for text data.  NOTE -- This algorithm is quadratic in
+# the worst case, i.e., when the data is already sorted.  A random method of
+# selecting the pivot should be used to improve the behaviour on sorted arrays.
+
+procedure ss_quick (linptr, linbuf, nlines)
+
+int	linptr[ARB]		# indices of strings in buffer
+char	linbuf[ARB]		# string buffer
+int	nlines			# number of strings
+
+int	i, j, k, temp, lv[LOGPTR], p, pivlin, uv[LOGPTR]
+int	ss_compare()
+
+begin
+	lv[1] = 1
+	uv[1] = nlines
+	p = 1
+
+	while (p > 0) {
+	    if (lv[p] >= uv[p])		# only one elem in this subset
+		p = p - 1		# pop stack
+	    else {
+		# Dummy loop to trigger optimizer.
+		do p = p, ARB {
+		    i = lv[p] - 1
+		    j = uv[p]
+
+		    # Select pivot element at midpoint of interval to avoid
+		    # quadratic behavior on a sorted list.
+
+		    k = (lv[p] + uv[p]) / 2
+		    swap (linptr[j], linptr[k])
+		    pivlin = linptr[j]
+
+		    while (i < j) {
+			for (i=i+1; ss_compare (linptr[i], pivlin, linbuf) < 0;
+			    i=i+1)
+			    ;
+			for (j=j-1;  j > i;  j=j-1)
+			    if (ss_compare (linptr[j], pivlin, linbuf) <= 0)
+				break
+			if (i < j)		# out of order pair
+			    swap (linptr[i], linptr[j])
+		    }
+
+		    j = uv[p]			# move pivot to position i
+		    swap (linptr[i], linptr[j])
+
+		    if (i-lv[p] < uv[p] - i) {	# stack so shorter done first
+			lv[p+1] = lv[p]
+			uv[p+1] = i - 1
+			lv[p] = i + 1
+		    } else {
+			lv[p+1] = i + 1
+			uv[p+1] = uv[p]
+			uv[p] = i - 1
+		    }
+
+		    break
+		}
+
+		p = p + 1			# push onto stack
+	    }
+	}
+end
+
+
+# SS_COMPARE -- Compare two strings.  Return -1 if str1<str2, 1 if str1>str2,
+# or 0 if the two strings are equal.
+
+int procedure ss_compare (lp1, lp2, linbuf)
+
+int	lp1, lp2		# pointers to substrings in linbuf
+char	linbuf[ARB]		# text buffer
+
+double	num1, num2
+int	ip1, ip2, answer, len1, len2
+int	strcmp(), ss_findcolumn(), gctod()
+include	"sort.com"
+
+begin
+	if (column != 0) {
+	    ip1 = ss_findcolumn (linbuf, lp1, column)
+	    ip2 = ss_findcolumn (linbuf, lp2, column)
+	} else if (ignore_whitespace == YES) {
+	    for (ip1=lp1;  IS_WHITE(linbuf[ip1]);  ip1=ip1+1)
+		;
+	    for (ip2=lp2;  IS_WHITE(linbuf[ip2]);  ip2=ip2+1)
+		;
+	} else {
+	    ip1 = lp1
+	    ip2 = lp2
+	}
+
+	if (numeric_sort == YES) {
+	    len1 = gctod (linbuf, ip1, num1)
+	    len2 = gctod (linbuf, ip2, num2)
+
+	    # If fields are nonnumeric, compare as strings.
+	    if (len1 == 0 || len2 == 0)
+		answer = strcmp (linbuf[ip1-len1], linbuf[ip2-len2])
+	    else if (num1 < num2)
+		answer = -1
+	    else if (num1 > num2)
+		answer = 1
+	    else
+		answer = 0
+	} else
+	    answer = strcmp (linbuf[ip1], linbuf[ip2])
+
+	if (reverse_sort == YES)
+	    return (-answer)
+	else
+	    return (answer)
+end
+
+
+# SS_FINDCOLUMN -- Determine the offset of the first character in the 
+# desired column.
+
+int procedure ss_findcolumn (buf, start, column)
+
+char	buf[ARB]
+int	start, column
+int	ip, col
+
+begin
+	for (ip=start;  IS_WHITE(buf[ip]);  ip=ip+1)
+	    ;
+	for (col=1;  col < column;  col=col+1) {
+	    for (;  !IS_WHITE(buf[ip]) && buf[ip] != EOS;  ip=ip+1)
+		;
+	    for (;  IS_WHITE(buf[ip]);  ip=ip+1)
+		;
+	}
+
+	return (ip)
+end
+
+
+# SS_GTEXT -- Get text lines into linbuf.
+
+int procedure ss_gtext (infile, linptr, nlines, linbuf)
+
+int	infile, linptr[ARB], nlines
+char	linbuf[ARB]
+
+int	lbp, len, getline()
+errchk	getline
+
+begin
+	nlines = 0
+	lbp = 1
+
+	repeat {
+	    len = getline (infile, linbuf[lbp])
+	    if (len == EOF)
+		break
+	    else if (len == 1)
+		# ignore blank lines
+	    else {
+		nlines = nlines + 1
+		linptr[nlines] = lbp
+		lbp = lbp + len + 1		# '1' = room for EOS
+	    }
+	} until (lbp >= SZ_LINBUF - SZ_LINE || nlines >= MAXPTR)
+
+	return (len)
+end
+
+
+# SS_PTEXT -- Output text lines from linbuf to outfile.
+
+procedure ss_ptext (outfil, linptr, nlines, linbuf)
+
+int	outfil, linptr[ARB], nlines
+char	linbuf[ARB]
+int	i, j
+errchk	putline
+
+begin
+	for (i=1;  i <= nlines;  i=i+1) {
+	    j = linptr[i]
+	    call putline (outfil, linbuf[j])
+	}
+end
diff --git a/pkg/system/system.cl b/pkg/system/system.cl
new file mode 100644
index 00000000..1e114d72
--- /dev/null
+++ b/pkg/system/system.cl
@@ -0,0 +1,55 @@
+lists
+
+#{ SYSTEM.CL -- Package script task for the SYSTEM package.  This package is
+# loaded by the CL upon startup, and is always in the search path.
+
+package	system
+
+task	cmdstr,
+	chkupdate,
+	concatenate,
+	copy,
+	count,
+	delete,
+	directory,
+	files,
+	head,
+	lprint,
+	match,
+	mkdir,
+	movefiles,
+	mtclean,
+	$netstatus,
+	page,
+	pathnames,
+	protect,
+	rename,
+	sort,
+	tail,
+	tee,
+	touch,
+	type,
+	rewind,
+	unprotect,
+	fcache,
+	urlget,
+	help		= "system$x_system.e"
+
+task	mkscript	= "system$mkscript.cl"; hidetask cmdstr
+task	$news		= "system$news.cl"
+task	$bench		= "system$bench.cl"
+
+task	allocate	= "hlib$allocate.cl"
+task	gripes		= "hlib$gripes.cl"
+task	deallocate	= "hlib$deallocate.cl"
+task	devstatus	= "hlib$devstatus.cl"
+task	$diskspace	= "hlib$diskspace.cl"
+task	$spy		= "hlib$spy.cl"
+
+task	$devices	= "system$devices.cl"
+task	references	= "system$references.cl"
+task	phelp		= "system$phelp.cl"
+
+hidetask mtclean
+
+keep
diff --git a/pkg/system/system.hd b/pkg/system/system.hd
new file mode 100644
index 00000000..be4b1bbe
--- /dev/null
+++ b/pkg/system/system.hd
@@ -0,0 +1,47 @@
+# Help directory for the SYSTEM package.
+
+$defdir	= "system$"
+$doc	= "system$doc/"
+$help	= "system$help/"
+
+allocate	hlp=doc$allocate.hlp,	 src=hlib$allocate.cl
+concatenate	hlp=doc$concatenate.hlp, src=concatenate.x
+chkupdate	hlp=doc$chkupdate.hlp, 	 src=chkupdate.x
+copy		hlp=doc$copy.hlp,	 src=copy.x
+count		hlp=doc$count.hlp,	 src=count.x
+deallocate	hlp=doc$deallocate.hlp,	 src=hlib$deallocate.cl
+delete		hlp=doc$delete.hlp,	 src=delete.x
+devices		hlp=dev$devices.hlp,	 src=devices.cl
+devstatus	hlp=doc$devstatus.hlp,	 src=hlib$devstatus.cl
+directory	hlp=doc$directory.hlp,	 src=directory.x
+diskspace	hlp=doc$diskspace.hlp,	 src=hlib$diskspace.cl
+fcache		hlp=doc$fcache.hlp,	 src=t_fcache.x
+files		hlp=doc$files.hlp,	 src=files.x
+gripes		hlp=doc$gripes.hlp,	 src=hlib$gripes.cl
+head		hlp=doc$head.hlp,	 src=head.x
+help		hlp=doc$help.hlp,	 src=help$t_help.x
+lprint		hlp=doc$lprint.hlp,	 src=lprint.x
+match		hlp=doc$match.hlp,	 src=match.x
+mkdir		hlp=doc$mkdir.hlp,	 src=mkdir.x
+mkscript	hlp=doc$mkscript.hlp,	 src=mkscript.cl
+movefiles	hlp=doc$movefiles.hlp,	 src=movefiles.x
+news		hlp=doc$news.hlp,	 src=news.cl
+netstatus	hlp=doc$netstatus.hlp,	 src=netstatus.x
+page		hlp=doc$page.hlp,	 src=page.x
+pathnames	hlp=doc$pathnames.hlp,	 src=pathnames.x
+phelp		hlp=doc$phelp.hlp,	 src=phelp.cl
+protect		hlp=doc$protect.hlp,	 src=protect.x
+references	hlp=doc$references.hlp,	 src=references.cl
+rename		hlp=doc$rename.hlp,	 src=rename.x
+revisions	hlp=doc$revisions.hlp,	 src=revisions.cl
+rewind		hlp=doc$rewind.hlp,	 src=rewind.x
+sleep		hlp=doc$sleep.hlp,	 src=sleep.x
+sort		hlp=doc$sort.hlp,	 src=sort.x
+spy		hlp=doc$spy.hlp,	 src=hlib$spy.cl
+system		hlp=doc$system.hlp,	 src=system.cl
+tail		hlp=doc$tail.hlp,	 src=tail.x
+tee		hlp=doc$tee.hlp,	 src=tee.x
+touch		hlp=doc$touch.hlp,	 src=touch.x
+type		hlp=doc$type.hlp,	 src=type.x
+urlget		hlp=doc$urlget.hlp,	 src=urlget.x
+unprotect	hlp=doc$unprotect.hlp,	 src=unprotect.x
diff --git a/pkg/system/system.men b/pkg/system/system.men
new file mode 100644
index 00000000..b79ae550
--- /dev/null
+++ b/pkg/system/system.men
@@ -0,0 +1,39 @@
+       allocate - Allocate a device, i.e., magtape drive mta, mtb, ...
+    concatenate - Concatenate a list of files
+      chkupdate - Check for an available IRAF update
+	   copy - Copy a file or files (use IMCOPY for imagefiles)
+	  count - Count the number of lines, words, characters in a text file
+     deallocate - Deallocate a previously allocated device
+         delete - Delete a file or files (use IMDELETE to delete imagefiles)
+	devices - Print information on the locally available devices
+      devstatus - Print the status of a device (mta, mtb, ...)
+      directory - List the files in a directory
+      diskspace - Show how much diskspace is available
+	 fcache - List, clean or manipulate the file cache
+	  files - Expand a file template into a list of files
+         gripes - Send suggestions, complaints, etc. to the system
+	   head - Print the first few lines of a text file
+           help - Print online documentation
+         lprint - Print a file on the line printer device
+	  match - Print all lines in a file that match a pattern
+	  mkdir - Create a new directory
+       mkscript - Make a command script
+      movefiles - Move files to a directory
+      netstatus - Print the status of the local network
+	   news - Page through the system news file
+           page - Page through a file
+      pathnames - Expand a file template into a list of OS pathnames
+	  phelp - Paged HELP: collects and pages the output of HELP
+        protect - Protect a file from deletion
+     references - Find all help database references for a given topic
+         rename - Rename a file
+         rewind - Rewind a device (magtape)
+	   sort - Sort a text file
+	    spy - Show processor status
+	 system - Simple benchmarking tool
+	   tail - Print the last few lines of a file
+	    tee - Tee the standard output into a file
+          touch - Change file access and modification times
+	   type - Type a text file on the standard output
+         urlget - Get a (http) URL to a named file
+      unprotect - Remove delete protection from a file
diff --git a/pkg/system/system.par b/pkg/system/system.par
new file mode 100644
index 00000000..238ccfcd
--- /dev/null
+++ b/pkg/system/system.par
@@ -0,0 +1,5 @@
+# Package parameter file for SYSTEM.
+
+version,s,h,"12-Nov-83"
+release,s,h,)cl.release,,,IRAF release version
+mode,s,h,ql
diff --git a/pkg/system/t_fcache.x b/pkg/system/t_fcache.x
new file mode 100644
index 00000000..236c33c4
--- /dev/null
+++ b/pkg/system/t_fcache.x
@@ -0,0 +1,118 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+
+define	FC_CMDS	"|init|purge|destroy|list|lookup|access|add|delete|wait|help|"
+
+define	FC_INIT		1
+define	FC_PURGE	2
+define	FC_DESTROY	3
+define	FC_LIST		4
+define	FC_LOOKUP	5
+define	FC_ACCESS	6
+define	FC_ADD		7
+define	FC_DELETE	8
+define	FC_WAIT		9
+define	FC_HELP		10
+
+
+# FCACHE -- Task interface to the file cache.
+
+procedure t_fcache ()
+
+char	cmd[SZ_FNAME], cache[SZ_FNAME], fname[SZ_FNAME]
+char	pattern[SZ_FNAME], src[SZ_FNAME], cname[SZ_FNAME], extn[SZ_FNAME]
+bool	verbose, in_src, exists
+int	age
+
+int	strdic(), clgeti(), envgeti()
+bool	clgetb(), fcaccess()
+
+begin
+	# Get the common parameters.
+	call clgstr ("cmd", cmd, SZ_FNAME)
+	call clgstr ("cache", cache, SZ_FNAME)
+	verbose = clgetb ("verbose")
+
+	# Process the requested command.
+	switch (strdic (cmd, cmd, SZ_FNAME, FC_CMDS)) {
+	case FC_INIT:
+	    call clgstr ("pattern", pattern, SZ_FNAME)
+	    call fcinit (cache, pattern)
+
+	case FC_PURGE:
+	    age = clgeti ("age")
+	    if (age < 0)
+		age = envgeti ("cache_age")
+	    call fcpurge (cache, verbose, age)
+
+	case FC_DESTROY:
+	    call fcdestroy (cache, verbose)
+
+	case FC_LIST:
+	    call fclist (cache, verbose, STDOUT)
+
+	case FC_LOOKUP:
+	    call clgstr ("src", src, SZ_FNAME)
+	    if (src[1] != EOS) {
+	        call fclookup (cache, src, fname, extn, SZ_FNAME)
+	        if (verbose) {
+		    call printf ("%s\n")
+		        call pargstr (fname)
+	        }
+	        call clpstr ("fname", fname)
+	        call clpstr ("extn", extn)
+	    } else {
+	        call clgstr ("fname", fname, SZ_FNAME)
+	        call clgstr ("extn", extn, SZ_FNAME)
+	        call fclookup (cache, src, fname, extn, SZ_FNAME)
+		if (verbose) {
+		    call printf ("%s\n")
+		        call pargstr (src)
+		}
+	        call clpstr ("src", src)
+	    }
+
+	case FC_ACCESS:
+	    call clgstr ("src", src, SZ_FNAME)
+	    call clgstr ("extn", extn, SZ_FNAME)
+
+	    exists =  fcaccess (cache, src, extn)
+	    call printf ("%b\n")
+	        call pargb (exists)
+
+	case FC_ADD:
+	    call clgstr ("src", src, SZ_FNAME)
+	    call clgstr ("extn", extn, SZ_FNAME)
+	    call fcadd (cache, src, extn, cname, SZ_FNAME)
+	    if (verbose) {
+		call eprintf ("%s\n")
+		    call pargstr (cname)
+	    }
+	    call clpstr ("fname", cname)
+	    if (clgetb ("wait"))
+	        call fcwait (cache, cname)
+
+	case FC_DELETE:
+	    call clgstr ("src", src, SZ_FNAME)
+	    if (src[1] != EOS) {
+		# Delete by src string.
+	        call fclookup (cache, src, fname, extn, SZ_FNAME)
+	        call fcdelete (cache, fname)
+	    } else {
+		# Delete by cached filename.
+	        call clgstr ("fname", fname, SZ_FNAME)
+	        call fcdelete (cache, fname)
+	    }
+
+	case FC_WAIT:
+	    call clgstr ("src", src, SZ_FNAME)
+	    call fcwait (cache, src)
+
+	case FC_HELP:
+
+	default:
+	    call eprintf ("Unknown command '%s'\n")
+		call pargstr (cmd)
+	}
+end
diff --git a/pkg/system/t_urlget.x b/pkg/system/t_urlget.x
new file mode 100644
index 00000000..d1c234df
--- /dev/null
+++ b/pkg/system/t_urlget.x
@@ -0,0 +1,96 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+include	<syserr.h>
+include	<imhdr.h>
+include	<imset.h>
+include <mach.h>
+
+
+
+# URLGET -- Do an HTTP GET of a URL to the named file.
+
+procedure t_urlget ()
+
+pointer	reply
+char	url[SZ_PATHNAME], fname[SZ_PATHNAME], extn[SZ_PATHNAME]
+char	cache[SZ_PATHNAME], lfname[SZ_PATHNAME]
+int	nread
+bool	use_cache, verbose
+
+int	nargs, clgeti(), url_get()
+bool	clgetb(), fcaccess()
+
+begin
+	# Get the parameters
+	call clgstr ("url", url, SZ_PATHNAME)
+	nargs = clgeti ("$nargs")
+	if (nargs < 2) {
+	    # No output name specified, so create one from URL
+	    call url_to_name (url, fname, SZ_PATHNAME)
+	} else 
+	    call clgstr ("fname", fname, SZ_PATHNAME)
+	call clgstr ("extn", extn, SZ_PATHNAME)
+	call clgstr ("cache", cache, SZ_PATHNAME)
+	verbose = clgetb ("verbose")
+	use_cache = clgetb ("use_cache")
+
+
+	# Tell them what we're doing.
+	if (verbose) {
+	    call printf ("%s -> %s\n")
+		call pargstr (url)
+		call pargstr (fname)
+	    call flush (STDOUT)
+	}
+
+	# Retrieve the URL.
+	if (use_cache) {
+	    call aclrc (lfname, SZ_FNAME);
+
+	    if (fcaccess (cache, url, "fits")) {
+	        call fcname (cache, url, "f", lfname, SZ_PATHNAME)
+		if (extn[1] != EOS) {
+		    # Add an extension to the cached file.
+		    call strcat (".", lfname, SZ_PATHNAME)
+		    call strcat (extn, lfname, SZ_PATHNAME)
+		}
+	    } else {
+	        # Add it to the cache, also handles the download.
+	        call fcadd (cache, url, extn, lfname, SZ_PATHNAME)
+	    }
+	    call fcopy (lfname, fname)
+
+	} else {
+	    # Not in cache, or not using the cache, so force the download.
+	    call calloc (reply, SZ_LINE, TY_CHAR)
+	    nread = url_get (url, fname, reply)
+	    call mfree (reply, TY_CHAR)
+	}
+end
+
+
+# URL_TO_NAME -- Generate a filename from a URL.
+
+procedure url_to_name (url, name, maxch)
+
+char	url[ARB]				#i URL being accessed
+char	name[ARB]				#o output name
+int	maxch					#i max size of output name
+
+int	ip, strlen()
+char	ch
+
+begin
+	ip = strlen (url)
+	while (ip > 1) {
+	    ch = url[ip]
+	    if (ch == '/' || ch == '?' || ch == '&' || ch == ';' || ch == '=') {
+		call strcpy (url[ip+1], name, maxch)
+	        return
+	    }
+	    ip = ip - 1
+	}
+
+	call strcpy (url[ip], name, maxch)
+end
diff --git a/pkg/system/tail.par b/pkg/system/tail.par
new file mode 100644
index 00000000..649e66ad
--- /dev/null
+++ b/pkg/system/tail.par
@@ -0,0 +1,2 @@
+input_files,s,a,,,,list of files to be printed
+nlines,i,h,12,,,number of lines to print from each file
diff --git a/pkg/system/tail.x b/pkg/system/tail.x
new file mode 100644
index 00000000..ad4ebae0
--- /dev/null
+++ b/pkg/system/tail.x
@@ -0,0 +1,107 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+
+# TAIL -- Print the tail (last few lines) of each of the named text files on the
+# standard output.  If more than one file is to be printed, a brief header is
+# printed for each file.
+# 
+# Params:
+# 	input_files		File matching template
+# 	nlines [h,12]		Number of lines at tail of file to be printed
+#
+# If NLINES is negative, its absolute value is the number of lines to skip at
+# head of file.
+
+procedure t_tail()
+
+char	fname[SZ_FNAME]
+bool	multiple_files
+int	nlines, list
+int	clpopni(), clplen(), clgfil(), clgeti()
+
+begin
+	list = clpopni ("input_files")
+	nlines = clgeti ("nlines")
+
+	multiple_files = (clplen (list) > 1)
+	
+	while (clgfil (list, fname, SZ_FNAME) != EOF)
+	    iferr (call print_tail (STDOUT, fname, nlines, multiple_files))
+		call erract (EA_WARN)
+
+	call clpcls (list)
+end
+
+
+# PRINT_TAIL -- Print the last few lines of a file on the output stream
+# given as the first argument, optionally plus a header.
+
+procedure print_tail (out, fname, nlines, print_file_name)
+
+char	fname[ARB]
+int	out, nlines
+bool	print_file_name
+
+pointer	sp, offsets, line
+int	in, linenum, index, noffsets, nlines_in_file
+int	open(), getline()
+long	note()
+errchk	open, getline, putline, note
+
+begin
+	if (nlines == 0)
+	    return
+	else
+	    noffsets = abs (nlines) + 1
+
+	call smark (sp)
+	call salloc (line, SZ_LINE, TY_CHAR)
+	call salloc (offsets, noffsets, TY_LONG)
+
+	# Open the file.
+	in = open (fname, READ_ONLY, TEXT_FILE)
+
+	if (print_file_name) {
+	    call fprintf (out, "\n\n===> %s <===\n")
+		call pargstr (fname)
+	}
+
+	# If nlines is negative, skip nlines lines at head of file and
+	# print rest of file.  If nlines is positive, print nlines lines
+	# at tail of file.
+
+	linenum = 0
+	repeat {
+	    linenum = linenum + 1
+	    if (nlines < 0 && linenum > -nlines)
+		break
+	    else if (nlines > 0) {
+		index = mod (linenum - 1, noffsets)
+		Meml[offsets+index] = note (in)
+	    }
+	} until (getline (in, Memc[line]) == EOF)
+
+	nlines_in_file = linenum - 1
+
+	# Seek back to offset of desired line, if not skipping head of file.
+	if (nlines > 0)
+	    if (nlines_in_file <= nlines)
+		call seek (in, BOFL)
+	    else {
+		index = mod (nlines_in_file - nlines, noffsets)
+		call seek (in, Meml[offsets+index])
+	    }
+
+	# If nlines is positive, print nlines lines, otherwise print rest
+	# of file.
+	linenum = 1
+	while ((linenum <= nlines || nlines < 0) &&
+	    (getline (in, Memc[line]) != EOF)) {
+		call putline (out, Memc[line])
+		linenum = linenum + 1
+	}
+
+	call sfree (sp)
+	call close (in)
+end
diff --git a/pkg/system/tee.par b/pkg/system/tee.par
new file mode 100644
index 00000000..9e833007
--- /dev/null
+++ b/pkg/system/tee.par
@@ -0,0 +1,3 @@
+tee_file,s,a,,,,name of file into which STDOUT is to be copied
+out_type,s,h,"text",,,type of output file (text|binary)
+append,b,h,no,,,append to output file?
diff --git a/pkg/system/tee.x b/pkg/system/tee.x
new file mode 100644
index 00000000..b22abd32
--- /dev/null
+++ b/pkg/system/tee.x
@@ -0,0 +1,58 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<fset.h>
+
+define	SZ_OTSTR	10
+
+# TEE -- Tee the standard output.  The type of the output file defaults to
+# text file unless the caller specifies otherwise.  Note that this tee
+# writes the entire output all at once.
+
+procedure t_tee()
+
+char	tee_file[SZ_FNAME], out_type_string[SZ_OTSTR]
+bool	clgetb()
+pointer	sp, buf
+int	out, sz_buf, out_type, nchars
+int	open(), read(), strmatch(), getline(), fstati()
+
+begin
+	# Get params and open the output file.
+	call clgstr ("tee_file", tee_file, SZ_FNAME)
+	call clgstr ("out_type", out_type_string, SZ_OTSTR)
+
+	if (strmatch (out_type_string, "^#{b}") > 0)
+	    out_type = BINARY_FILE
+	else
+	    out_type = TEXT_FILE
+
+	if (clgetb ("append"))
+	    out = open (tee_file, APPEND, out_type)
+	else
+	    out = open (tee_file, NEW_FILE, out_type)
+
+
+	# Make a buffer and perform the copy operation.
+	call smark (sp)
+
+	if (out_type == TEXT_FILE) {
+	    call salloc (buf, SZ_LINE, TY_CHAR)
+	    while (getline (STDIN, Memc[buf]) != EOF) {
+		call putline (STDOUT, Memc[buf])
+		call putline (out, Memc[buf])
+	    }
+
+	} else {
+	    sz_buf = fstati (STDIN, F_BUFSIZE)
+	    call salloc (buf, sz_buf, TY_CHAR)
+
+	    while (read (STDIN, Memc[buf], sz_buf) != EOF) {
+		nchars = fstati (STDIN, F_NCHARS)
+		call write (STDOUT, Memc[buf], nchars)
+		call write (out, Memc[buf], nchars)
+	    }
+	}
+
+	call sfree (sp)
+	call close (out)
+end
diff --git a/pkg/system/touch.par b/pkg/system/touch.par
new file mode 100644
index 00000000..4adb14b3
--- /dev/null
+++ b/pkg/system/touch.par
@@ -0,0 +1,7 @@
+files,s,a,,,,list of files to be created or touched
+create,b,h,yes,,,create file if it doesn't exist?
+atime,b,h,yes,,,update file access time?
+mtime,b,h,yes,,,update file modification time?
+time,s,h,"",,,time to set (empty for current time)
+ref_file,s,h,"",,,reference file for times
+verbose,b,h,no,,,verbose output flag
diff --git a/pkg/system/touch.x b/pkg/system/touch.x
new file mode 100644
index 00000000..5635dbec
--- /dev/null
+++ b/pkg/system/touch.x
@@ -0,0 +1,193 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<finfo.h>
+include	<error.h>
+include	<time.h>
+
+
+# TOUCH -- Change file access and modification times, creating an empty
+# file if necessary.  File times may come from the current system time,
+# a user-specified string, or a reference file.
+
+procedure t_touch()
+
+pointer	list
+char	fname[SZ_FNAME], time[SZ_FNAME]
+char	ref_file[SZ_PATHNAME]
+char	at_str[SZ_TIME], mt_str[SZ_TIME]
+bool	create, atime, mtime, verbose
+
+long	ref[LEN_FINFO]
+long	t_atime, t_mtime, ltime
+
+long	clktime()
+int	clpopni(), clgfil()
+int	finfo(), dtm_ltime()
+bool	clgetb()
+
+errchk	touch
+
+begin
+	# Initialize.
+	call aclrc (time, SZ_FNAME)
+	call aclrc (ref_file, SZ_PATHNAME)
+
+	# Process the parameters.
+	list = clpopni ("files")
+
+	create  = clgetb ("create")
+	atime   = clgetb ("atime")
+	mtime   = clgetb ("mtime")
+	verbose = clgetb ("verbose")
+	call clgstr ("time", time, SZ_FNAME)
+
+	# Check for error conditions.
+	if (!atime && !mtime) {
+	    call eprintf (
+		"ERROR: Must specify at least one of 'atime' or 'mtime'.\n")
+	    call clpcls (list)
+	    return
+	} 
+
+	# Get the time to be set.
+	if (time[1] == EOS) {
+	    # No 'time' param, look for a reference file to use.
+	    call clgstr ("ref_file", ref_file, SZ_PATHNAME)
+	    if (ref_file[1] == EOS) {
+		# No 'ref_file' param, use the current system time.
+		t_atime = clktime (long(0))
+		t_mtime = t_atime
+
+		if (verbose) {
+	    	    call cnvtime (t_atime, at_str, SZ_TIME)
+	    	    call printf ("Modifying to system time: %s\n\n")
+			call pargstr (at_str)
+		}
+
+	    } else {
+		# Open the reference file and use those times.
+		if (finfo (ref_file, ref) == ERR) {
+		    call eprintf ("Error opening reference file: '%s'\n")
+		        call pargstr (ref_file)
+		    return
+		}
+		t_atime = FI_ATIME(ref)
+		t_mtime = FI_MTIME(ref)
+
+		if (verbose) {
+	    	    call cnvtime (t_atime, at_str, SZ_TIME)
+	    	    call cnvtime (t_mtime, mt_str, SZ_TIME)
+	    	    call printf ("Modifying to reference time: ")
+		    call printf ("%s (atime)\n%30t%s (mtime)\n\n")
+			call pargstr (at_str)
+			call pargstr (mt_str)
+		}
+	    }
+
+	} else {
+	    # Parse the time parameter to get the modification time.
+	    if (dtm_ltime (time, ltime) == ERR) {
+		if (ltime < 0) 
+		    call eprintf ("Invalid time string: '%s'\n")
+		else
+		    call eprintf ("Error parsing time string: '%s'\n")
+		        call pargstr (time)
+		return
+	    }
+	    t_atime = ltime
+	    t_mtime = ltime
+
+	    if (verbose) {
+	        call cnvtime (t_atime, at_str, SZ_TIME)
+	    	call printf ("Modifying to user-specified time: %s\n\n")
+		    call pargstr (at_str)
+	     }
+	}
+
+	# Now apply the atime/mtime params to update only what's needed.
+	if (!atime) t_atime = NULL
+	if (!mtime) t_mtime = NULL
+
+
+	# Process the list of input files.
+	while (clgfil (list, fname, SZ_FNAME) != EOF) {
+	    iferr (call touch (fname, create, t_atime, t_mtime, verbose))
+		;
+	}
+
+	# Clean up and close the list.
+	call clpcls (list)
+end
+
+
+# TOUCH -- Touch a file to modify the times.
+
+procedure touch (fname, create, atime, mtime, verbose)
+
+char	fname[ARB]			#i file name to touch
+bool	create				#i create file if necessary?
+long	atime, mtime			#i access and modify time
+bool	verbose				#i verbose output?
+
+char 	dir[SZ_PATHNAME], ip
+char 	vfn[SZ_PATHNAME]
+int	fd
+int	access(), open(), futime()
+
+begin
+	if (verbose) {
+	    call printf ("%s: ")
+		call pargstr (fname)
+	}
+
+	# Check first it the file exists.
+	if (access (fname, 0, 0) == NO) {
+	    if (create) {
+
+		call fnldir (fname, vfn, SZ_PATHNAME)
+		call fpathname (vfn, dir, SZ_PATHNAME)
+		if (access (dir, READ_WRITE, 0) == NO) {
+		    for (ip=1; dir[ip] != '!'; ip=ip+1)
+			;
+		    call eprintf ("Error: Cannot open directory '%s'\n")
+		        call pargstr (dir[ip+1])
+		    call erract (EA_ERROR)
+		    return;
+		}
+
+		# Create a new empty file.
+	        iferr (fd = open (fname, NEW_FILE, TEXT_FILE)) {
+		    call eprintf ("Error: Cannot touch file '%s'\n")
+		        call pargstr (fname)
+		    call erract (EA_ERROR)
+		    return;
+		}
+	        call close (fd)
+		if (verbose) call printf ("(created) ")
+
+	    } else {
+		if (verbose) call printf ("(not created)\n")
+	        return
+	    }
+	}
+
+	# Update the times.
+	if (futime (fname, atime, mtime) == ERR)
+	    call eprintf ("Error processing file\n")
+
+	else if (verbose) {
+	    call printf ("(updated %s%s%s)\n")
+		if (atime > 0)
+		    call pargstr ("atime") 
+		else
+		    call pargstr ("")
+		if (atime > 0 && mtime > 0)
+		    call pargstr ("/")     
+		else
+		    call pargstr ("")
+		if (mtime > 0)
+		    call pargstr ("mtime") 
+		else
+		    call pargstr ("")
+	}
+end
diff --git a/pkg/system/type.par b/pkg/system/type.par
new file mode 100644
index 00000000..62ce1380
--- /dev/null
+++ b/pkg/system/type.par
@@ -0,0 +1,3 @@
+input_files,s,a,,,,list of files to be typed
+map_cc,b,h,yes,,,make unknown control chars printable
+device,s,h,"terminal",,,output device
diff --git a/pkg/system/type.x b/pkg/system/type.x
new file mode 100644
index 00000000..156cb5c5
--- /dev/null
+++ b/pkg/system/type.x
@@ -0,0 +1,68 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<error.h>
+
+# TYPE -- Type the named files on the standard output.  If more than one
+# file, print a brief header identifying each file.
+
+procedure t_type()
+
+char	fname[SZ_FNAME], device[SZ_FNAME]
+bool	multiple_files
+int	map_cc, list
+pointer	tty, ttyodes()
+bool	clgetb()
+int	clpopni(), clplen(), clgfil(), btoi()
+
+begin
+	list = clpopni ("input_files")
+	multiple_files = (clplen (list) > 1)
+	map_cc = btoi (clgetb ("map_cc"))
+	call clgstr ("device", device, SZ_FNAME)
+	tty = ttyodes (device)
+
+	while (clgfil (list, fname, SZ_FNAME) != EOF)
+	    iferr (call type_file (STDOUT, tty, fname, multiple_files, map_cc))
+		call erract (EA_WARN)
+
+	call ttycdes (tty)
+	call clpcls (list)
+end
+
+
+# TYPE_FILE -- Print the named file on the output stream given as the first
+# argument, optionally with a leading header.
+
+procedure type_file (out, tty, fname, print_file_name, map_cc)
+
+int	out
+pointer	tty
+char	fname[ARB]
+bool	print_file_name
+int	map_cc
+
+int	in
+pointer	sp, buf
+int	open(), getline()
+errchk	salloc, open, getline, ttyputline
+
+begin
+	call smark (sp)
+	call salloc (buf, SZ_LINE, TY_CHAR)
+
+	in = open (fname, READ_ONLY, TEXT_FILE)
+
+	if (print_file_name) {
+	    call fprintf (out, "===> %s <===\n")
+		call pargstr (fname)
+	}
+
+	while (getline (in, Memc[buf]) != EOF)
+	    call ttyputline (out, tty, Memc[buf], map_cc)
+
+	if (print_file_name)
+	    call fprintf (out, "\n\n")
+
+	call close (in)
+	call sfree (sp)
+end
diff --git a/pkg/system/unprotect.par b/pkg/system/unprotect.par
new file mode 100644
index 00000000..f57772b3
--- /dev/null
+++ b/pkg/system/unprotect.par
@@ -0,0 +1 @@
+files,s,a,,,,list of files from which protection is to be removed
diff --git a/pkg/system/unprotect.x b/pkg/system/unprotect.x
new file mode 100644
index 00000000..c265bbf8
--- /dev/null
+++ b/pkg/system/unprotect.x
@@ -0,0 +1,23 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+include	<protect.h>
+include	<error.h>
+
+# UNPROTECT -- Remove protection from a list of files.
+
+procedure t_unprotect()
+
+char	fname[SZ_FNAME]
+int	list, status
+
+int	clpopns(), clgfil(), protect()
+
+begin
+	list = clpopns ("files")
+
+	while (clgfil (list, fname, SZ_FNAME) != EOF)
+	    iferr (status = protect (fname, REMOVE_PROTECTION))
+		call erract (EA_WARN)
+
+	call clpcls (list)
+end
diff --git a/pkg/system/urlget.par b/pkg/system/urlget.par
new file mode 100644
index 00000000..64126be8
--- /dev/null
+++ b/pkg/system/urlget.par
@@ -0,0 +1,6 @@
+url,s,q,"",,,"URL (http) to retrieve"
+fname,s,q,"",,,"Output file name"
+use_cache,b,h,yes,,,"Use previously cached file?"
+extn,s,h,"",,,"Filename extn for cached file"
+verbose,b,h,no,,,"Use verbose output?"
+cache,s,h,"cache$",,,"Cache directory"
diff --git a/pkg/system/x_system.x b/pkg/system/x_system.x
new file mode 100644
index 00000000..0695b0d3
--- /dev/null
+++ b/pkg/system/x_system.x
@@ -0,0 +1,36 @@
+# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
+
+# Process x_system (the executable part of the SYSTEM package).
+
+task	chkupdate	= t_chkupdate,
+	concatenate	= t_concatenate,
+	copy		= t_copy,
+	count		= t_count,
+	delete		= t_delete,
+	directory	= t_directory,
+	fcache		= t_fcache,
+	files		= t_files,
+	head		= t_head,
+	match		= t_match,
+	mkdir		= t_mkdir,
+	movefiles	= t_movefiles,
+	mtclean		= t_mtclean,
+	netstatus	= t_netstatus,
+	page		= t_page,
+	pathnames	= t_pathnames,
+	protect		= t_protect,
+	rename		= t_rename,
+	sort		= t_sort,
+	tail		= t_tail,
+	tee		= t_tee,
+	type		= t_type,
+	unprotect	= t_unprotect,
+	rewind 		= t_rewind,
+	lprint		= t_lprint,
+	help		= t_help,
+	lroff		= t_lroff,
+	mkhelpdb	= t_mkhelpdb,
+	hdbexamine	= t_hdbexamine,
+	cmdstr		= t_cmdstr,
+	touch		= t_touch,
+	urlget		= t_urlget