1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
|
include <mach.h>
include <imhdr.h>
include <error.h>
include "cyber.h"
# UNPK_12 -- Unpack 12-bit unsigned integers from an array containing
# one sixty bit cyber word in each pair of array elements.
# Each output word contains successive 12-bit pixels from the
# input array.
# Pixels are unpacked starting with element "first_word" of the input array.
# Each cyber 60-bit word contains 5 packed 12-bit pixels, the first pixel
# in the highest 12 bits. The input array contains one cyber word per two
# array elements; the cyber word occupies the lower 60 bits of each pair
# of array values.
procedure unpk_12 (input, first_word, output, npix_unpk)
int input[ARB] #
int first_word #
int output[npix_unpk] #
int npix_unpk #
int n, nn, i, offset[5], off
int npix_word, ncyber_words, index
data (offset[i], i=1, 5) /15, 27, 39, 51, 63/
int bitupk()
begin
npix_word = 5
if (mod (npix_unpk, npix_word) == 0)
ncyber_words = (npix_unpk) / npix_word
else
call error (0, "Incorrect number of pixels to be unpacked")
index = 1
i = 1
for (n = first_word; i <= npix_unpk; n = n + 2) {
do nn = 1, npix_word {
off = (n + 1) * NBITS_INT - offset[nn]
output[i] = bitupk (input, off, 12)
if (output[i] == 7777B)
output[i] = BLANK
i = i + 1
}
}
end
# UNPK-20 -- Unpack 20-bit signed integers from an array containing
# one 60-bit Cyber word per pair of array elements. Each output
# word contains successive 20-bit pixels from the input array. Pixels
# are unpacked starting with array element "first_word". Conversion
# from one's complement to two's complement is performed. Each Cyber
# word contains 3 packed 20-bit pixels, the first pixel in the highest
# 20 bits.
procedure unpk_20 (input, first_word, output, npix_unpk)
int input[ARB]
int output[npix_unpk], npix_unpk, first_word
int n, i, nn, off
int npix_word, ncyber_words, pix_val, offset[3]
data (offset[i], i=1, 3) /23, 43, 63/
int bitupk()
begin
npix_word = 3
if (mod (npix_unpk, npix_word) == 0)
ncyber_words = npix_unpk / npix_word
else
call error (0, "Incorrect number of pixels to be unpacked")
i = 1
for (n = first_word; i <= npix_unpk; n = n + 2) {
do nn = 1, npix_word {
off = (n + 1) * NBITS_INT - offset[nn]
pix_val = bitupk (input, off, 20)
if (pix_val == 3777777B)
pix_val = BLANK
else if (and (pix_val, 2000000B) != 0)
# negative pixel
pix_val = -and (3777777B, not(pix_val))
output[i] = pix_val
i = i + 1
}
}
end
# UNPK_60R -- Unpack Cyber 60-bit floating point numbers from an array
# containing one 60-bit word per pair of array elements.
# The 30 most significant bits from each 60-bit word are
# unpacked and then reconstructed as a floating point number with
# with REPACK_FP. The extracted bits include an 18-bit mantissa,
# 11-bit exponent and a sign bit. This routine is used for getting
# the min and max values from the header; no 60-bit IPPS pixels are
# expected.
procedure unpk_60r (input, first_word, fp_value, nwords)
int input[ARB]
real fp_value[nwords]
int first_word, nwords, n, i
pointer int_buf, sp
int bitupk()
begin
# Allocate space on stack
call smark (sp)
call salloc (int_buf, nwords, TY_INT)
i = 1
for (n = first_word; i <= nwords; n = n + 2) {
Memi[int_buf + i - 1] = bitupk (input[n], 31, 30)
i = i + 1
}
call repack_fp (Memi[int_buf], fp_value, nwords)
call sfree (sp)
end
# UNPK_60I -- Unpack 60-bit integers from an array containing one Cyber
# word per each NINT_CYBER_WRD elements. Each word
# of output contains only the lower 32 bits of each input word, as this
# procedure is called only for getting the reduction flags from the
# IDSFILE header.
procedure unpk_60i (input, initial_bit_offset, output, nwords)
char input[ARB]
int output[nwords]
int initial_bit_offset, nwords, bit_offset, n
int bitupk()
errchk bitupk
begin
bit_offset = initial_bit_offset
do n = 1, nwords {
output[n] = bitupk (input, bit_offset, NBITS_INT)
if (and (output[n], 2000000000B) != 0)
# negative value
output[n] = -not(output[n])
bit_offset = bit_offset + (NINT_CYBER_WRD * NBITS_INT)
}
end
# CONVERT_60BIT_FP -- returns a floating point number equivalent to the Cyber
# 60-bit number input. The full 48-bit mantissa and 11-bit exponent
# is used in reconstructing the floating point value.
double procedure convert_60bit_fp (cyber_word, bit_offset)
int cyber_word[ARB], bit_offset
int temp1, temp2
double float_value
int exp, lower_mantissa, upper_mantissa, i
int bitupk(), and(), not()
double tbl[255]
include "powd.inc"
errchk bitupk
begin
# Extract cyber word in question into temp array
temp1 = bitupk (cyber_word, bit_offset, 30)
temp2 = bitupk (cyber_word, bit_offset + 30, 30)
# Check "bit59" and complement all bits if it is set
if (bitupk (temp2, 30, 1) != 0) {
temp1 = not (temp1)
temp2 = not (temp2)
lower_mantissa = -and (temp1, 7777777777B)
upper_mantissa = -and (temp2, 777777B)
} else {
lower_mantissa = temp1
upper_mantissa = and (temp2, 777777B)
}
# Extract and interpret exponent; remove Cyber bias of 2000B and
# convert to two's complement if negative number
exp = bitupk (temp2, 19, 11)
if (exp > 1777B)
# "bit58" is set, positive exponent
exp = exp - 2000B
else
# negative exponent
exp = exp - 1777B
# Reconstruct the floating point number. 30 is added to the
# exponent for the upper mantissa. The 129 is to register the data
# array matrix: tbl[1] = 2 ** -128 ==> 2 ** n = tbl[n + 129]
# float_value = mantissa * 2 ** (exp + 129)
#
# float_value = (lower_mantissa) * 2 ** (exp + 129) +
# (upper_mantissa) * 2 ** (exp + 30 + 129)
float_value = double (lower_mantissa) * tbl[exp + 129] +
double (upper_mantissa) * tbl[exp + 30 + 129]
return (float_value)
end
# UNPK_30 -- unpack Cyber 30-bit floating point numbers from an array
# containing one 60-bit Cyber word in each pair of array elements. Each
# 30-bit pixel is unpacked from this array; procedure REPACK_FP is called
# to reconstruct the floating point number. Pixels are unpacked starting
# with array element "first_word". Each Cyber word contains 2 30-bit
# pixels, the first pixel in the higher 30 bits.
procedure unpk_30 (input, first_word, fp_value, npix)
int input[ARB]
int first_word, npix
real fp_value[npix]
pointer int_buf, sp
int n, i, off, offset[2]
int bitupk()
data (offset[i], i = 1, 2) /33, 63/
errchk bitupk
begin
# Allocate buffer space, allowing for maximum of 1 extraneous pixel
call smark (sp)
call salloc (int_buf, npix + 1, TY_INT)
i = 1
for (n = first_word; i <= npix; n = n + 2) {
off = (n + 1) * NBITS_INT - offset[1]
Memi[int_buf + i - 1] = bitupk (input, off, 30)
off = (n + 1) * NBITS_INT - offset[2]
Memi[int_buf + i] = bitupk (input, off, 30)
i = i + 2
}
call repack_fp (Memi[int_buf], fp_value, npix)
call sfree (sp)
end
# UNPK_ID -- Unpacks ID string from input array, which contains one Cyber
# word per two array elements. The word_offset equals the number of Cyber
# words to skip before beginning to unpack. If the character string
# begins in word one of "input", word_offset = 0. The IPPS ID string
# is written in 7-bit ASCII, with eight characters per Cyber word. The lowest
# 4 bits of each 60-bit word is unused. The highest 7 bits of the first Cyber
# word contain the character count.
procedure unpk_id (input, word_offset, output)
int input[ARB]
int word_offset
char output[SZ_IPPS_ID]
int nbits, nchar_offset, id_offset, nchars, n
int nchars_word, ncyber_words, nn, index
int bitupk()
begin
nbits = 7
nchar_offset = (word_offset * NBITS_INT * NINT_CYBER_WRD) +
NBITS_CYBER_WORD - 6
nchars = bitupk (input, nchar_offset, nbits)
ncyber_words = (nchars + 6) / 7
index = 1
for (n = 1; n <= ncyber_words; n = n + 1) {
if (n == 1) {
nchars_word = 7
id_offset = nchar_offset - 7
} else {
nchars_word = 8
id_offset = nchar_offset + ((n-1) * NBITS_INT * NINT_CYBER_WRD)
}
do nn = 1, nchars_word {
output[index] = bitupk (input, id_offset, nbits)
index = index + 1
id_offset = id_offset - 7
}
}
output[nchars+1] = EOS
end
# REPACK_FP -- Reconstruct a floating point number. The input to REPACK_FP
# is an array of integers containing Cyber 30-bit floating point numbers
# in the least significant bits of each array element. The exponent, mantissa
# and two bits indicating the sign are extracted and used to reassemble
# the floating point value. Cyber blanks and overflows are returned as BLANK.
procedure repack_fp (int_value, float_value, nvalues)
int int_value[ARB], nvalues
real float_value[nvalues]
int i, pixel
int exp, mantissa
real tbl[255]
int bitupk(), and(), not()
include "pow.inc"
begin
do i=1, nvalues {
pixel = int_value[i]
# Check for blanks
if (pixel == 1777000000B) {
float_value[i] = BLANK
next
}
# Check "bit59" and complement all bits if it is set
if (and (pixel, 4000000000B) != 0) {
pixel = not (pixel)
mantissa = -and (pixel, 777777B)
} else
mantissa = and (pixel, 777777B)
# Extract and interpret exponent: remove Cyber bias of 2000B
# and convert to two's complement if negative number
exp = bitupk (pixel, 19, 11)
if (exp > 1777B)
# "bit58" is set, positive exponent
exp = exp - 2000B
else
# negative exponent
exp = exp - 1777B
# Reconstruct the floating point value: 30 is added to the
# exponent because only the top 18 bits of the 48-bit mantissa
# were extracted; the 129 is to register the data array index.
# float_value[i] = real(mantissa) * 2 ** (exp + 30)
# (tbl[1] = 2 ** -128) ==> (2 ** n = tbl[n + 129]).
exp = exp + 30 + 129
if (exp <= 0) {
#call eprintf (
#"RDUMPF_REPACK_FP: pixel with exponent underflow seen\n")
float_value[i] = 0.0
} else if (exp > 255) {
#call eprintf (
#"RDUMPF_REPACK_FP: pixel with exponent overflow seen\n")
float_value[i] = MAX_REAL
} else if (exp > 0 && exp <= 255)
float_value[i] = real(mantissa) * tbl[exp]
}
end
# DISPLAY_CODE -- returns the ascii character equivalent to the display
# code character input. The Cyber uses the 63-character display code
# set internally, although the 64-character set is used for output.
procedure display_code (in_char, out_char)
char in_char, out_char
char dc[64]
int i
data (dc[i], i=1,8) /072B, 101B, 102B, 103B, 104B, 105B, 106B, 107B/
data (dc[i], i=9,16) /110B, 111B, 112B, 113B, 114B, 115B, 116B, 117B/
data (dc[i], i=17,24) /120B, 121B, 122B, 123B, 124B, 125B, 126B, 127B/
data (dc[i], i=25,32) /130B, 131B, 132B, 060B, 061B, 062B, 063B, 064B/
data (dc[i], i=33,40) /065B, 066B, 067B, 070B, 071B, 053B, 055B, 052B/
data (dc[i], i=41,48) /057B, 050B, 051B, 044B, 075B, 040B, 054B, 056B/
data (dc[i], i=49,56) /043B, 133B, 135B, 045B, 042B, 137B, 041B, 046B/
data (dc[i], i=57,64) /047B, 077B, 074B, 076B, 100B, 134B, 136B, 073B/
begin
out_char = dc[in_char + 1]
end
|