Small docs update #1
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The original names conflict with the auth_* mechanism handler methods.
Forgot an `await` keyboard resulting in trying to split a coroutine. In the process of fixing that, I went a bit wild with type hinting to prevent similar mistakes. And while doing that, discovered potential problem if `login` is `None`
* RFC4954 specified "Enhanced Status Codes" for AUTH-related replies * Also some status codes are semantically incorrect
* Correct pragma * Proper venv exclusion
Asserts that keep happening with the exact same params are now collected in `testing/helpers.py` This will ease modification because these exact same params are now in one place instead of distributed all over the tests.
On Windows, intializing the server takes quite some time, the original wait was too short. We extend the delay + parameterize it so it's tunable from one location.
We got 100% coverage with this one as we delve deep into some internal logic.
Because `atpublic=2.0` (latest release as of this commit) _requires_ Python>=3.6
Only now do I understand what ${PLATFORM} means.
I have no other idea on how to do this. I mean, *many* test cases already implicitly 'requires' the code to go through the branches. But coverage.py doesn't seem to realize this. I was left with no other option but to do "pragma: no branch" Also, removed a no-longer-needed "pragma: nocover"
E133 is actually contrary to PEP8, which DOES allow closing brace/bracket/ parenthesis right under the first character of the line that starts the multiline construct.[1] W503 and W504 are conflicting with each other, and also prevent multiline expressions within parentheses. The acrobatics required to not violate both W503 and W504 will have seriously detrimental effects to the readability of the code, so I disabled them. Personal note: I really *really* REALLY want to change max-line-length to 90 as proposed by Raymond Hettinger in "Beyond PEP8"[2], but thankfully it's not really necessary for aiosmtpd. I'll live. Fuming, but alive. [1] https://www.python.org/dev/peps/pep-0008/#indentation [2] https://www.youtube.com/watch?v=wf-BqAjZb8M
I figure I'll do this first to reduce technical debt. Improvements & docs are to follow, Real Soon.
Documentation update included.
Previously, one will need to remember to change the indices. Adding (in- sertion) and deleting (removal) affects the line and all lines following that one. Result: Big `git diff`, too many chaff that hide what actually happened. With this change, you can easily add/remove/move around the expected lines.
So I changed how I enumerate the handlers, wrote some additional tests, and finally ensure that all tox scenarios pass. Documentation has been updated to reflect this.
`login_data` is definitely more generic than `login_id`
... and somehow achieved an unbelievable **order-of-magnitude increase** in performance! WHOA. Behavioral changes: * List of valid (non-excluded) AUTH hooks now created exactly once during class instantiation * That means, you can no change the AUTH hooks "on the fly" * Not that you can do that easily in previous versions anyways... ALSO: This passes all tests (including new ones I whup up to test AUTH hook exclusion) and the all-important "qa" tox test.
This still passes all tox tests including qa
It seems the last supported version for pypy36 is 0.5.0?
Co-authored-by: LGTM Migrator <lgtm-migrator@users.noreply.github.com>
* Add note on asyncore+asynchat deprecation * Remove LGTM badge, Add Python Implementation badge * Add downloads badges to PyPI description * Some comment fixing in DESCRIPTION.rst
* Rename method to no longer shadow superclass attrib Closes aio-libs#343 and aio-libs#344 * Remove double assignment Already handled in the if..else Closes aio-libs#345 * Remove double imports Closes aio-libs#346 * Do something on handling of KeyboardInterrupt Closes aio-libs#349 * Tighten up Protocol definition Closes aio-libs#350 * Move ... to its own line flake8 hates the tighter format * Remove totally unused var & enumerate() Closes aio-libs#351 * Add CodeQL badge to README and DESCRIPTION I spent effort, I'm allowed to boast, dammit :D
Protect 1.4.3 from tainting if we accidentally released something from `master` 😁
* Support explicit specification of PyPy version I can't stress how important envdir is. tox managed to somehow b0rked my WHOLE aiosmtpd repo when I forgot to explicitly specify the envdirs for pypy37 and pypy38 * Latest tox forces recreate if env changes So our previous 'optimizations' to prevent reinstallation of packages are rendered moot. Let's optimize it the other way around then: Install only what's necessary. Also since tox will recreate environments anyways, let's no longer use per-python-version envdir; rather specify the full env name instead. * Force using latest coverage.py To ensure we get the PyPy v7.3.10 decorator-tracing fix.
* Drop testing for PyPy3.6, add PyPy3.9 Also force coverage version, and add comment about why coverage for PyPy3.8 is not done. * Add PyPy versions to listed envnames * Change example to use py37
* Workaround for asyncio.get_event_loop() deprecation * Remove unused import in main.py That import was used only to get the current event loop. Now we handle that in `__init__.py` * BUGFIX: if loop is not None then use that Also simplify the line because we have a helper func now * Add pragma: no cover Because at the point the expected behavior is not yet happening. * Update NEWS + bump to 1.4.4a1 * Don't ignore the warning, rather promote it into an error * Some test cases to ensure the get event func works as intended * Skip coverage only on Python<3.10 * Create new event loop only on failure Because apparently in Python<3.12, even though get_event_loop() raises a warning, it still returns existing event loop. The behavior is expected to change on Python>=3.12, so let's cut off at that point. * More stringent testing for _get_or_new_eventloop * FIX: Type hint for fixture should be Generator[]
Dang.
Committed the wrong file, dang.
After careful consideration, it seems that pytype was actually complaining of the possible "implicit return None" if a suitable value is not found in the `for a` loop. Lacking a better solution, I decided to temporarily disable this kind of error just for this function.
* Stop release process if repo is not clean * Update test_release_date Now can match versions with alphanumeric suffixes * Give last chance to cancel upload after build * Bump version to 1.4.4.post2
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
There was a problem hiding this comment.
CodeQL found more than 10 potential problems in the proposed changes. Check the Files changed tab for more details.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I did a quick gander at the docs and didn't see any particular links against any specific Python version.
However, I'm pretty sure that some of the module
:ref:links will fial against >=3.12, since smtpd is dropped there.I also didn't confirm that any of the example docs match the (updated?) API/actual code.