aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Remove unused codeJoseph Hunkeler2024-11-021-2/+0
|
* Squelch shell's pushd/popd commandsJoseph Hunkeler2024-11-021-0/+8
|
* Add missing doc parameterJoseph Hunkeler2024-11-021-0/+1
|
* Combine initializer with declarationJoseph Hunkeler2024-11-022-8/+5
|
* Return shell() directlyJoseph Hunkeler2024-11-021-3/+1
| | | | * The status variable is redundant
* Bug fixes:Joseph Hunkeler2024-11-024-9/+38
| | | | | | | | * Remove PATH modification from check_pathvar * Rename check_pathvar to check_system_path * Add conda_prepend_bin function to handle PATH modification during activation * Add logic to handle first-run activation, and subsequent calls to change conda environments (otherwise conda clobbers its own shell functions) * Use $CONDA_PYTHON_EXE and $CONDA_EXE to ensure conda comes from the just-installed tree
* Initialize structs to {0} and combine declaration and assignment where possibleJoseph Hunkeler2024-11-0130-220/+141
|
* Add missing parameters to docstringsJoseph Hunkeler2024-11-011-0/+3
|
* Support automatic workspace setup/teardownJoseph Hunkeler2024-11-011-7/+87
| | | | * inline the code (its a header library)
* Change mock delivery name to be more uniqueJoseph Hunkeler2024-11-011-1/+1
|
* Change default RT workspace directory pathJoseph Hunkeler2024-11-011-1/+1
|
* Tests: Force the removal of podman containersJoseph Hunkeler2024-11-011-1/+11
|
* Fix SYSERROR macroJoseph Hunkeler2024-11-011-1/+1
| | | | * "info" was always printed instead of the errno string. Now it prints the error and returns "info" on "Success"
* Replace initialize character arrays using {0} instead of memsetJoseph Hunkeler2024-11-0115-78/+41
|
* Merge pull request #66 from jhunkeler/order-mattersJoseph Hunkeler2024-10-312-4/+2
|\ | | | | BUG: Order matters
| * Remote PATH= definition in commandorder-mattersJoseph Hunkeler2024-10-311-3/+1
| | | | | | | | * check_pathvar was the culprit
| * Move check_pathvarJoseph Hunkeler2024-10-311-1/+1
|/ | | | * This needed to be one line up, otherwise the next function would not be able to import conda
* Merge pull request #65 from jhunkeler/more-rtJoseph Hunkeler2024-10-3013-115/+514
|\ | | | | More RT
| * Update _test_boilerplate.cmore-rtJoseph Hunkeler2024-10-301-2/+10
| |
| * Add tests/README.mdJoseph Hunkeler2024-10-301-0/+48
| |
| * Add RT boilerplateJoseph Hunkeler2024-10-301-0/+27
| |
| * Implement missing SKIP conditionsJoseph Hunkeler2024-10-301-7/+17
| | | | | | | | | | * The not (bang) overrides the return code. Split the command from the if-statement to obtain the real code, not just true/false * Print the last n lines of the log only when there's something to print
| * Fixes download testJoseph Hunkeler2024-10-301-4/+17
| | | | | | | | | | * NGINX updated and changed the default error page source code * Look for "404" in the page contents
| * Update location of test dataJoseph Hunkeler2024-10-301-2/+2
| |
| * Update RT tests to use run_command functionJoseph Hunkeler2024-10-302-16/+19
| |
| * Implement a runner (run_command) and generate a summary (run_summary) when ↵Joseph Hunkeler2024-10-301-2/+72
| | | | | | | | the script ends
| * Exposes TOPDIR and TEST_DATA variables to test script(s)Joseph Hunkeler2024-10-301-1/+2
| |
| * generic_based_on.ini using TEST_DATA variable from the environment to find ↵Joseph Hunkeler2024-10-301-1/+1
| | | | | | | | yaml config
| * Copy test data in bulkJoseph Hunkeler2024-10-301-6/+11
| | | | | | | | | | * Detect RT scripts * Generalize file extension pattern use
| * Re-enable regression testsJoseph Hunkeler2024-10-281-0/+1
| |
| * Add rt_generic_based_on.shJoseph Hunkeler2024-10-284-0/+106
| | | | | | | | * Tests the "based_on" key
| * Workaround for latest conda regressionJoseph Hunkeler2024-10-281-1/+3
| | | | | | | | | | * profile.d/{conda,mamba}.sh don't set PATH so "import conda" and "import mamba" fail when conda is invoked. Fantastic. * Put conda's python interpreter on the PATH before using conda for the first time
| * rt_generic: Use setup scriptJoseph Hunkeler2024-10-281-108/+27
| |
| * Add environment setup script to test directoryJoseph Hunkeler2024-10-281-0/+186
|/
* Merge pull request #64 from jhunkeler/combine-providesJoseph Hunkeler2024-10-258-108/+180
|\ | | | | Combine functions: pip_index_provides and conda_index_provides
| * Strip the version spec off of the package name if presentcombine-providesJoseph Hunkeler2024-10-241-3/+15
| |
| * Fix regression:Joseph Hunkeler2024-10-241-34/+45
| | | | | | | | | | * Wheels are only built now if they appear in the pip_packages_deferred list. * Strip the version spec off of the package name if present
| * Change signature of check_pathvar()Joseph Hunkeler2024-10-242-2/+2
| | | | | | | | * From (char *) to (void)
| * Move fd initializer downJoseph Hunkeler2024-10-241-2/+2
| |
| * Generalize *_index_provides interfaceJoseph Hunkeler2024-10-244-67/+116
|/ | | | | * Replaces conda_* and pip_* with pkg_index_provides * Because this function can fail in so many ways I've added pkg_index_provides_strerror() and a detection macro PKG_INDEX_PROVIDES_FAILED() to make things easier
* Merge pull request #63 from jhunkeler/update-testsJoseph Hunkeler2024-10-2212-36/+298
|\ | | | | Update tests / Bug fixes
| * Update test_mp_fail_fastupdate-testsJoseph Hunkeler2024-10-221-2/+1
| | | | | | | | * macOS Actions runners have 3 vCPUs. No wonder this wasn't working. Reduce the success count to 1.
| * Update test_mp_fail_fastJoseph Hunkeler2024-10-221-5/+11
| | | | | | | | * Brute force the conditions required to make this work. Issue 128 jobs.
| * Update test_mp_fail_fastJoseph Hunkeler2024-10-221-7/+13
| | | | | | | | | | * Record status(s) and just verify they're non-zero. Checking for exact values is difficult when you don't know the hardware ahead of time. * Apply HOLD state
| * mp_pool_show_summary now prints "HOLD" when queued tasks haven't been ↵Joseph Hunkeler2024-10-221-1/+7
| | | | | | | | executed and the pool is killed
| * Add MP_POOL_TASK_STATUS_INITIALJoseph Hunkeler2024-10-222-2/+5
| | | | | | | | * Better than putting -1 everywhere
| * The CI internal clocks are off? I don't see why these failed.Joseph Hunkeler2024-10-211-7/+6
| | | | | | | | * Increasing sleep times might help.
| * Add tests:Joseph Hunkeler2024-10-211-0/+72
| | | | | | | | | | * test_envctl_init * test_envctl_register
| * Add tests:Joseph Hunkeler2024-10-211-0/+82
| | | | | | | | | | * test_mp_fail_fast * test_mp_stop_continue
| * Add tests:Joseph Hunkeler2024-10-211-2/+52
| | | | | | | | | | | | | | | | | | | | | | | | | | * test_pip_index_provides * test_conda_provides * test_conda_get_active_environment * test_delivery_gather_tool_versions Fix tests: * global.sysconfdir duplicates the environment string * Use delivery_free() * Use globals_free()
generated by cgit (git 2.34.1) at 2025-05-12 15:36:00 -0400