aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Fix zstd file magicJoseph Hunkeler40 hours1-1/+1
|
* Add GBO test using a modern conda installerJoseph Hunkeler2 days2-0/+96
|
* Update stasis.iniJoseph Hunkeler3 days1-2/+2
| | | | | * Install boa, conda-build, and rattler-build * If the version of conda or mamba is too new, boa is automatically removed from the list at runtime (its deprecated and causes the solver to fail)
* Increase verbosity of the string testsJoseph Hunkeler3 days1-9/+46
|
* Remove unused argumentJoseph Hunkeler3 days1-1/+1
|
* pkg_index_providesJoseph Hunkeler3 days1-29/+74
| | | | | | * Modern mamba returns zero on error, and prints its error messages to stdout. FANTASTIC. * The error detection seems to still handle the previous behavior as well with older versions. * And now we print the contents of stdout and stderr on a non-zero exit from the package manager
* Update test_wheel.c to use conda_capababilities()Joseph Hunkeler3 days1-2/+4
|
* Update test_conda.c to use conda_capababilities()Joseph Hunkeler3 days1-30/+22
| | | | * Remove inner workspace direction creation (left over from before the test framework handled it)
* Rewrite debug_hexdumpJoseph Hunkeler3 days2-47/+33
|
* Replace __FUNCTION__ with __func__Joseph Hunkeler3 days2-2/+2
|
* tests: Do not print the reason message for tests that passed when a failure ↵Joseph Hunkeler3 days1-0/+1
| | | | occurs
* Passing the conda installation directory to delivery_conda_enable isn't ↵Joseph Hunkeler3 days2-9/+9
| | | | | | necessary * The delivery context already provides this path
* Ignore stderr from installation dry-runJoseph Hunkeler3 days1-0/+1
|
* Ensure .condrc exists and conda knows where to find itJoseph Hunkeler3 days1-0/+13
|
* Store conda installation prefix in capabilities structureJoseph Hunkeler3 days3-3/+7
|
* Update test_recipe.cJoseph Hunkeler3 days1-6/+6
|
* Remove redundant code left behind from developmentJoseph Hunkeler3 days1-11/+0
|
* Add python_importlib_metadata_version functionJoseph Hunkeler3 days2-18/+46
| | | | * Use it for conda and libmamapy version detection
* Add support for modern versions of conda and mambaJoseph Hunkeler3 days1-22/+54
| | | | | | * Boa is dead, but we can still use it if conda is old enough * Also because boa is dead we purge it from the list if the conda version is too new * Initial environment activation changed as well, but still compatible with older versions of conda
* Separate the recipe style and build system.Joseph Hunkeler3 days1-22/+45
| | | | * Add "support" for rattler. Very loosely.
* Set mamba-related environment variablesJoseph Hunkeler3 days1-0/+2
|
* Add STASIS_ALWAYS_BUILD_FOR_HOST environment variable triggerJoseph Hunkeler3 days2-0/+5
|
* Disallow extraneous conda and pip version specsJoseph Hunkeler3 days1-0/+48
| | | | | | * STASIS wants equals-to, or nothing (latest). * The use of URLs with '@' is what we need to avoid * Tell the user the expected format on error
* Add recipe_get_build_systemJoseph Hunkeler3 days2-17/+43
| | | | | * Rename RECIPE_TYPE_* to RECIPE_STYLE_* * Add defines for RECIPE_BUILD_* that denote which build system should be used to build the recipe
* delivery_conda_enable: always pin conda's version to the same version ↵Joseph Hunkeler3 days1-0/+24
| | | | | | provided by the installer * Some commands upgrade conda whether it is configured to allow the behavior or not
* shell: throw warning if temporary script's permissions cannot be modifiedJoseph Hunkeler3 days1-1/+3
|
* delivery_conda_enable populates CondaCapabilities membersJoseph Hunkeler3 days1-1/+11
|
* Add CondaCapabilities as argument to conda_setup_headlessJoseph Hunkeler3 days2-2/+2
|
* Hook CondaCapabilities up to the delivery contextJoseph Hunkeler3 days2-0/+4
|
* String handlingJoseph Hunkeler3 days1-19/+18
|
* Add conda_capable, conda_capable functionsJoseph Hunkeler3 days2-0/+108
| | | | * Add struct CondaCapabilities
* General fixes (#149)HEAD1.7.1masterJoseph Hunkeler3 days7-24/+32
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Fix missing LF * Return when item is NULL * Explicitly state error condition * strlist_contains: Allow index result argument to be NULL * BUGFIX: strlist_contains now uses strstr instead of strcmp * The match was never intended to be exact * BUGFIX: log_msgv now prints a LF to the same stream as the message * BUGFIX: env_manipulate_pathstr PM_ONCE logic did opposite of what was intended * When strstr finds the path in the system path, it's supposed to free the new system path and exit. In this case it was freeing the path when the path wasn't found that lead to several instances of PATH not being updated correctly * Free result when shell_output fails * join: Fix segfault on memory error * join_ex: Fix leak on memory error * Initialize installer_url and python_override_version to zero * Remove redundant calls to memset
* Always display information about docker even if its not available (#144)Joseph Hunkeler3 days2-29/+44
| | | * Update/add error messages indicating why docker checks failed
* Micromamba improvement (#148)1.7.0Joseph Hunkeler7 days14-31/+227
| | | | | | | | | | | | | | | | | | | | | * Update micromamba installation logic * Split installation from micromamba() into micromamba_install() * Return -1 when micromamba cannot be installed * Add stasis.ini option "indexer.micromamba_download_url" * Add global variable micromamba_download_url * The installation function attempts two known-good URLs by default * * Add indexer arugment '--micromamba-download-url' * Add is_file_compressed() function to utils * Call micromamba_install() from tests * Add space in usage statement * Fix usage output when option array contains arguments without short options * Add --micromamba-download-url to README.md
* Safe strings, finally (#145)Joseph Hunkeler2026-06-0234-331/+236
| | | | | | | | * Add string copy and catonate replacements * safe_strncpy * safe_strncat * Replace string functions * gbo.ini: Update tweakwcs to 0.9.0 * generic.ini: Update tweakwcs to 0.9.0
* Merge pull request #139 from jhunkeler/bughunt-10021.6.1Joseph Hunkeler2026-05-1234-298/+688
|\ | | | | Bughunt 0x1002
| * replace for-loop with while-loop since 'i' was unusedbughunt-1002Joseph Hunkeler2026-05-121-1/+1
| |
| * Replace escape sequence '\e' with '\x1b' (portable)Joseph Hunkeler2026-05-121-6/+6
| |
| * Replace __FUNCTION__ with __func__ (portable)Joseph Hunkeler2026-05-123-6/+6
| |
| * Replace void pointer with pointer to tplfunc_frame structureJoseph Hunkeler2026-05-125-14/+17
| |
| * Remove redundant __FUNCTION__Joseph Hunkeler2026-05-122-4/+4
| |
| * Placate FORTIFY_SOURCE warningJoseph Hunkeler2026-05-121-1/+3
| | | | | | | | * The log_root in almost every circumstance will be able to handle the extra slash chracter. It complains because log_file and log_path are both sizeof PATH_MAX
| * Reverse logic on status checkJoseph Hunkeler2026-05-121-14/+14
| |
| * Usable lines denote the number of entriesJoseph Hunkeler2026-05-121-3/+6
| | | | | | | | | | * reduce complexity * 'x' should not overflow
| * One argument for error message stringJoseph Hunkeler2026-05-121-2/+2
| |
| * Remove linefeed from debug messageJoseph Hunkeler2026-05-121-1/+1
| |
| * Missing temp file is a warning, not an errorJoseph Hunkeler2026-05-121-1/+1
| |
| * exit on errorJoseph Hunkeler2026-05-121-14/+77
| |
| * free result on errorJoseph Hunkeler2026-05-121-0/+1
| |
| * Use log_print_error instead of SYSERRORJoseph Hunkeler2026-05-121-1/+1
| | | | | | | | * Add comment so I don't accidentally do it again
generated by cgit (git 2.34.1) at 2026-06-25 02:11:34 -0400