Accept logging level names (#5765) #5772
Conversation
Signed-off-by: Uilian Ries <uilianries@gmail.com>
Signed-off-by: Uilian Ries <uilianries@gmail.com>
|
Seems like the parallel test has some fragility. I can't reproduce the same behavior when running virtual env on my machine. |
Signed-off-by: Uilian Ries <uilianries@gmail.com>
Signed-off-by: Uilian Ries <uilianries@gmail.com>
|
The functional tests was contaminated by env vars. |
| self.assertEqual(logging.DEBUG, config.logging_level) | ||
|
|
||
| def test_log_level_numbers_invalid(self): | ||
| save(os.path.join(self.tmp_dir, CONAN_CONF), default_client_conf_log.format("level = wakawaka")) |
There was a problem hiding this comment.
from early condition, any error is treated as default logging level (CRITICAL). Should we change it?
There was a problem hiding this comment.
The condition is here: https://github.com/conan-io/conan/blob/1.18.5/conans/client/conf/__init__.py#L511
| @@ -506,7 +507,8 @@ def logging_level(self): | |||
| if level is None: | |||
| level = self.get_item("log.level") | |||
| try: | |||
| level = int(level) | |||
| parsed_level = get_log_level_by_name(level) | |||
| level = parsed_level if parsed_level is not None else int(level) | |||
There was a problem hiding this comment.
int("wakawaka") doesn't crash?
Signed-off-by: Uilian Ries <uilianries@gmail.com>
Let's go ahead with this little change, I always have to think about the pairing between numbers and log-levels.
|
@jgsogo all yours. |
Please, @uilianries, consider these changes to keep the environment modification local to the test-case (uilianries#15) and I'll merge
Keep environment modification local to the testcase
|
@jgsogo Done, it has been merged! |
Perfect. Now move to the docs, we have the opportunity to improve them a little bit
Now any user can set Conan logging level using common names as debug, info, critical ...
Related Issue: #5765
Changelog: Feature: Accept logging level as logging names
Docs: conan-io/docs#1419
developbranch, documenting this one.Note: By default this PR will skip the slower tests and will use a limited set of python versions. Check here how to increase the testing level by writing some tags in the current PR body text.