- Date: 2024-02-03
- Version: >=3.8 and <=3.12
- Note : this version break major versioning, because old acceptation test fail for this reason
- Status: Finish
- Note: Branch containing the base of the pytest fixture 'lock', it must be able to easily integrate new functions, CLI arguments, etc…
- Modified
--lock
argument, now target only test withlock
fixture- The tests with the
lock
fixture hadskipped
status, now they havepassed
status - The tests without the
lock
fixture hadpassed
status, now they haveskipped
status
- The tests with the
- Status: Finish
- Note: This branch requires that the branch "feature/fixture-lock-method" be finalized.
- Add
pickle
extension forlock.lock
to support more types of data - Now
pickle
is the default extension forlock.lock
if no extension is specified
- Status: Start
- Note: This branch requires that the branch "feature/fixture-lock-method" be finalized.
- If test use
--lock
and--clean
argument, then clean all unused cache files - If test use
--lock
and--clean
argument and--only-skip
argument, then do anything, it's certainly a mistake (why clean only test with existing lock ?) - If test use
--lock
and--clean
argument and--simulate
argument, list all unused cache files who will be removed without remove them. - If test use
--lock
and--clean
argument and--lock-date
argument, thrown exception (can't lock a remove cache file)
- Date: 2024-01-27
- Version: >=3.8 and <=3.12
- Status: Finish
- Note: Branch containing the base of the pytest fixture 'lock', it must be able to easily integrate new functions, CLI arguments, etc…
- Add fixtures
lock
- Status: Finish
- Note: This branch requires that the branch "feature/lock-fixture" be finalized.
-
Add fixtures method
lock.lock
to lock the result of a test to a cache file-
If test use
lock.lock
and result was not locked, exception is thrown -
If test use
lock.lock
result is in the cache file and is the same as the result of the test, test is valid -
If test use
lock.lock
result is in the cache file and is not the same as the result of the test, test is invalid (failed) -
If test use
lock.lock
and--lock
in cli argument, then start test and lock the result in cache file. -
If test use
lock.lock
and--simulate
in cli argument, then simulate the result of the test, not write to the cache file. -
If test use
lock.lock
and--only-skip
in cli argument, then don't update lock if the result was not locked. -
If test use
lock.lock
and--lock-date
in cli argument, then lock the result of the test to a cache file with the date of the lock, if date was expired, then the test is failed -
If test use
--simulate
argument and not--lock
argument, then it's invalid, throw exception -
If test use
--lock-date
argument and not--lock
argument, then it's invalid, throw exception -
If test use
--only-skip
argument and not--lock
argument, then it's invalid, throw exception
-
- Status: Finish
- Note: This branch requires that the branch "feature/fixture-lock-method" be finalized.
- Add support for
pytest --lock --lock-date 13/12/2023
, if test haslock
fixture, then lock the result of the test to a cache file with the date of the lock, if date was expired, then the test is skipped
- Date: 2024-01-xx
- Version: xxxxxx
The package upload tests on Pypi were awkwardly carried out with version v0.1.0
and v0.1.1
, the package being defective, these were removed but Pypi refuses to change the package contents even after use, even if no download has been made. The first version is therefore v0.1.2
.