It is easy to forget backticks in change note bylines. It's happened
in #12531 already, requiring a hotfix in #12560.
The pre-commit based check idea is coming from the Tox project and
have been battle-tested in aiohttp, CherryPy, and other ecosystems.
The following snippet would have resulted in crash on multiple places since
`_get_verbose_word` expects only string, not a tuple.
```python
@pytest.hookimpl(tryfirst=True)
def pytest_report_teststatus(report: pytest.CollectReport | pytest.TestReport, config: pytest.Config):
if report.when == "call":
return ("error", "A", ("AVC", {"bold": True, "red": True}))
return None
```
```
Traceback (most recent call last):
File "/home/pbrezina/workspace/sssd/.venv/bin/pytest", line 8, in <module>
sys.exit(console_main())
^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/config/__init__.py", line 207, in console_main
code = main()
^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/config/__init__.py", line 179, in main
ret: Union[ExitCode, int] = config.hook.pytest_cmdline_main(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_hooks.py", line 513, in __call__
return self._hookexec(self.name, self._hookimpls.copy(), kwargs, firstresult)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_manager.py", line 120, in _hookexec
return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_callers.py", line 139, in _multicall
raise exception.with_traceback(exception.__traceback__)
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_callers.py", line 103, in _multicall
res = hook_impl.function(*args)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/main.py", line 333, in pytest_cmdline_main
return wrap_session(config, _main)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/main.py", line 321, in wrap_session
config.hook.pytest_sessionfinish(
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_hooks.py", line 513, in __call__
return self._hookexec(self.name, self._hookimpls.copy(), kwargs, firstresult)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_manager.py", line 120, in _hookexec
return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_callers.py", line 139, in _multicall
raise exception.with_traceback(exception.__traceback__)
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_callers.py", line 122, in _multicall
teardown.throw(exception) # type: ignore[union-attr]
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/logging.py", line 872, in pytest_sessionfinish
return (yield)
^^^^^
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_callers.py", line 124, in _multicall
teardown.send(result) # type: ignore[union-attr]
^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/terminal.py", line 899, in pytest_sessionfinish
self.config.hook.pytest_terminal_summary(
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_hooks.py", line 513, in __call__
return self._hookexec(self.name, self._hookimpls.copy(), kwargs, firstresult)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_manager.py", line 120, in _hookexec
return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_callers.py", line 139, in _multicall
raise exception.with_traceback(exception.__traceback__)
File "/home/pbrezina/workspace/sssd/.venv/lib64/python3.11/site-packages/pluggy/_callers.py", line 124, in _multicall
teardown.send(result) # type: ignore[union-attr]
^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/terminal.py", line 923, in pytest_terminal_summary
self.short_test_summary()
File "/home/pbrezina/workspace/pytest/src/_pytest/terminal.py", line 1272, in short_test_summary
action(lines)
File "/home/pbrezina/workspace/pytest/src/_pytest/terminal.py", line 1205, in show_simple
line = _get_line_with_reprcrash_message(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/terminal.py", line 1429, in _get_line_with_reprcrash_message
word = tw.markup(verbose_word, **word_markup)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/pbrezina/workspace/pytest/src/_pytest/_io/terminalwriter.py", line 114, in markup
text = "".join(f"\x1b[{cod}m" for cod in esc) + text + "\x1b[0m"
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~
TypeError: can only concatenate str (not "tuple") to str
```
Signed-off-by: Pavel Březina <pbrezina@redhat.com>
BPO is in read-only mode and all issues have been migrated to GitHub.
This patch replaces the use of that role with `:issue:`, recently
integrated via #12522. It also disables the built-in `extlinks` Sphinx
extension, as it's no longer in use.
These tests are known to only be executed partially or not at all. So
we always get incomplete, missing, and sometimes flaky, coverage in
the test functions that are expected to fail.
This change updates the ``coverage.py`` config to prevent said tests
from influencing the coverage level measurement.
This extension implements more generic roles that can also be used
more flexibly. Relying on an external extension allows to stop
maintaining an in-repo copy of the commonly used behavior.
It's necessary since it seems that the currently used Codecov uploader
doesn't read the token from config sometimes.
This is a follow-up for #12508 which wasn't enough.
The #11921 update broke uploading coverage of the `main` branch (or
any in-repo pushes for that matter) to Codecov 4 months ago.
Version 4 requires an upload token to be provided and since there was
no configuration for it, the upload was failing. But the step itself
was showing up as successful due to `fail_ci_if_error: true` being
set. The error is visible in the console output, though.
This patch flips the setting to `fail_ci_if_error: false` and sets the
Codecov upload token in the config in clear text. The non-secret part
allows the PRs uploads to be more stable.
Co-authored-by: Ronny Pfannschmidt <opensource@ronnypfannschmidt.de>
Previously, a part of the environments weren't documented in the
config, making it difficult for the newbies to figure out what
their purposes are. This patch sets the descriptions for all the
envs listed with the `tox -av` command, leveraging the dynamic
factor-dependent explanation fragments.
```console
default environments:
linting -> run pre-commit-defined linters under `python3`
py38 -> run the tests under `py38`
py39 -> run the tests under `py39`
py310 -> run the tests under `py310`
py311 -> run the tests under `py311`
py312 -> run the tests under `py312`
py313 -> run the tests under `py313`
pypy3 -> run the tests under `pypy3`
py38-pexpect -> run the tests against `pexpect` under `py38`
py38-xdist -> run the tests with pytest in parallel mode under `py38`
py38-unittestextras -> run the tests against the unit test extras under `py38`
py38-numpy -> run the tests against `numpy` under `py38`
py38-pluggymain -> run the tests against the bleeding edge `pluggy` from Git under `py38`
py38-pylib -> run the tests against `py` lib under `py38`
doctesting -> run the tests under `~/.pyenv/versions/3.12.3/envs/pytest-pyenv-py3.12.3/bin/python` including doctests
doctesting-coverage -> run the tests collecting coverage under `~/.pyenv/versions/3.12.3/envs/pytest-pyenv-py3.12.3/bin/python` including doctests
plugins -> run reverse dependency testing against pytest plugins under `~/.pyenv/versions/3.12.3/envs/pytest-pyenv-py3.12.3/bin/python`
py38-freeze -> test pytest frozen with `pyinstaller` under `py38`
docs -> build the documentation site under `~/src/github/pytest-dev/pytest/doc/en/_build/html` with `python3`
docs-checklinks -> check the links in the documentation with `python3`
py311-exceptiongroup -> run the tests against `exceptiongroup` under `py311`
additional environments:
regen -> regenerate documentation examples under `python3`
release -> do a release, required posarg of the version number
prepare-release-pr -> prepare a release PR from a manual trigger in GitHub actions
generate-gh-release-notes -> generate release notes that can be published as GitHub Release
nobyte -> run the tests in no-bytecode mode under `~/.pyenv/versions/3.12.3/envs/pytest-pyenv-py3.12.3/bin/python`
lsof -> run the tests with `--lsof` pytest CLI option under `~/.pyenv/versions/3.12.3/envs/pytest-pyenv-py3.12.3/bin/python`
```
The new change note types are `packaging`, `contrib` and `misc`.
`packaging` is intended for the audience of downstream redistributors.
The `contrib` notes are meant to be documenting news affecting the
project contributors, their development, and processes.
Finally, `misc` is for things that don't fit anywhere but are still
desired to be documented for some reason.
It is uncapped in RTD. Apparently, `tox -e docs` is not invoked in CI,
neither is it called in RTD, resulting in the regression having been
caught only in local development environments.
This is a follow-up for #12493.
It is called when building the docs. Apparently, `tox -e docs` is not
invoked in CI, neither is it called in RTD, resulting in the
regression having been caught only in local development environments.
This is a follow-up for #12493.
The earlier implementation was generating a temporary file, when
the docs site was being built with `tox`. However, this was not
enabled in RTD and is hackish.
This patch integrates the `sphinxcontrib-towncrier` extension to
make it work in any environment where Sphinx docs are being built.
Previously, it was possible to use several different config file name
variants. However, there is one canonical name now and the others are
being deprecated [[1]].
[1]: https://docs.rtfd.io/en/stable/config-file/index.html
Previously, this workflow was being used to cherry-pick PRs made
against `main` into older stable branches. Now that #12475 integrated
the Patchback GitHub App, it's no longer needed as it's making
duplicate pull requests which don't even trigger CI runs automatically
due to #10354. So this patch removes said workflow to address the
problem.
There was a regression caused by #12431 that excluded the patch on
broken CPython patch version ranges 3.11.0–3.11.8 and 3.12.0–3.12.2.
This change fixes that by adjusting the patching conditional clause
to include said versions.
Closes#12430.
Co-authored-by: Florian Bruhin <me@the-compiler.org>
* 🚑 Explicitly set encoding @ `subprocess.run()`
This invocation is present in the `test_issue_9765` regression test
that is always skipped unconditionally, resulting in the
`EncodingWarning` never manifesting itself in CI or development
environments of the contributors.
* Change `setup.py` call w/ `pip install` @ tests
Using this CLI interface has been deprecated in `setuptools` [[1]].
[1]: https://blog.ganssle.io/articles/2021/10/setup-py-deprecated.html
* Change Contribution doc title to match sidebar
* Rearrange sentence for clarity
* Update backwards-compatibility.rst
some minor grammar changes
* Update pythonpath.rst
fixed some gramatical errors
* Update AUTHORS
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Update doc/en/explanation/pythonpath.rst
From a quick overview it looks like lowercase is more consistent, although some pages do use `pytest` in code blocks
Co-authored-by: Bruno Oliveira <bruno@soliv.dev>
---------
Co-authored-by: Mackerello <82668740+Mackerello@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Bruno Oliveira <bruno@soliv.dev>
This fixes unittest test reruns when using plugins like
pytest-rerunfailures.
The `instance` property uses AttributeError to check if the instance
needs to be initialized, so `del` is the correct way to clear it, not
setting to `None`.
Regressed in 8.2.2.
* Only rely on _find_lineno on Python 3.10 and earlier.
* Update docstring to reflect current expectation.
* Only rely on _find on Python 3.9 and earlier.
* Mark line as uncovered.
* Remove empty else block (implicit is better than explicit).
Closes#12430Closes#12432
Launchable with ``pre-commit run --hook-stage manual pyupgrade -a``
---------
Signed-off-by: Tomasz Kłoczko <kloczek@github.com>
Co-authored-by: Pierre Sassoulas <pierre.sassoulas@gmail.com>
Fix#12355.
In the issue, it was reported that the `reorder_items` has quadratic (or
worse...) behavior with certain simple parametrizations. After some
debugging I found that the problem happens because the "Fix
items_by_argkey order" loop keeps adding the same item to the deque,
and it reaches epic sizes which causes the slowdown.
I don't claim to understand how the `reorder_items` algorithm works, but
if as far as I understand, if an item already exists in the deque, the
correct thing to do is to move it to the front. Since a deque doesn't
have such an (efficient) operation, this switches to `OrderedDict` which
can efficiently append from both sides, deduplicate and move to front.
Fix#12381
Test plan:
It's possible to write a deterministic test case for this, but somewhat
of a hassle so I tested it manually. I reproduced by removing existing
`.pytest_cache`, adding a sleep before the rename and running two
pytests. I verified that it doesn't reproduce after the fix.
* [pylint 'use-maxsplit-arg'] Do not split more than necessary when using the first element
* [pylint 'consider-using-enumerate'] Use zip when iterating on two iterators
* [pylint] 'cell-var-from-loop' and 'disallowed-name' permanent disable
* [pylint] Disable 'possibly-used-before-assignment' following 3.2.0 release
This is used to set the <link rel="canonical" href="X"> tag that points to the canonical version of the webpage. Including this indicates to search engines which version to include in their indexes, and should prevent older versions showing up.
Fixes#12363
* furo theme for docs site
* removing duplicate tocs from deprecations and reference pages
* removing pallets references in code and config
* reverting trainings to sidebar
* removed sphinx style and unpinned packaging version
* updated styles
The `.parent` was incorrectly removed in
a21fb87a90, but it was actually correct
(well, it assumes that Module.path.parent == Package.path, which I'm not
sure is always correct, but that's a different matter). Restore it.
Fix#12328.
Closes#11706.
Originally fixed in #11721, but then reverted in #12022 due to a regression in pytest-xdist.
The regression was fixed on the pytest-xdist side in pytest-dev/pytest-xdist#1026.
In these two cases which re-raise an immediately-caught exception, do
the re-raising with `raise` instead of `raise exc`, because the `raise
exc` adds an uninteresting entry to the exception's traceback (that of
itself), while `raise` doesn't.
* added test case in testing/python/approx.py based on test case provided by reporter in issue #12114
* test cases pass for pytest testing/python/approx.py
* expanded the type annotation to include objects which may cast to a array and renamed other_side to other_side_as_array and asserted that it is not none
Now `importlib` mode will correctly set the imported modules as an attribute of their parent modules.
As helpfully posted on #12194, that's how the Python import module works so we should follow suit.
In addition, we also try to import the parent modules as part of the process of importing a child module, again mirroring how Python importing works.
Fix#12194
Previously we used a hand crafted approach to detect namespace packages, however we should rely on ``importlib`` to detect them for us.
Fix#12112
---------
Co-authored-by: Ran Benita <ran@unusedvar.com>
Creating and initializing the cache directory is interruptible; this
avoids a pathological case where interrupting a cache write can cause
the cache directory to never be properly initialized with its supporting
files.
Unify `Cache.mkdir` with `Cache.set` while I'm here so the former also
properly initializes the cache directory.
Closes#12167.
It makes more sense, also, we have a long term idea of generalizing
fixture support to items defined by other plugins, not just python, in
which case `--fixtures` would definitely not be python-plugin
specific.
Based on pylint's message ``consider-iterating-dictionary`` suggestion.
Surprisingly using a dict or set comprehension instead of a new temp var is
actually consistently slower here, which was not intuitive for me.
```python
from timeit import timeit
families = {1: {"testcase": [1, 2, 3, 5, 8]}}
attrs = {1: "a", 2: "b", 3: "c", 4: "d", 5: "e", 6: "f", 7: "g", 8: "h"}
class Old:
def old(self):
self.attrs = attrs
temp_attrs = {}
for key in self.attrs.keys():
if key in families[1]["testcase"]:
temp_attrs[key] = self.attrs[key]
self.attrs = temp_attrs
class OldBis:
def old(self):
self.attrs = attrs
temp_attrs = {}
for key in self.attrs:
if key in families[1]["testcase"]:
temp_attrs[key] = self.attrs[key]
self.attrs = temp_attrs
class New:
def new(self):
self.attrs = attrs
self.attrs = { # Even worse with k: v for k in self.attrs.items()
k: self.attrs[k] for k in self.attrs if k in families[1]["testcase"]
}
if __name__ == "__main__":
n = 1000000
print(f"Old: {timeit(Old().old, number=n)}")
print(f"Just removing the keys(): {timeit(OldBis().old, number=n)}")
print(f"List comp, no temp var: {timeit(New().new, number=n)}")
```
Result:
Old: 0.9493889989680611
Just removing the keys(): 0.9042672360083088
List comp, no temp var: 0.9916125109884888
It's also true for the other example with similar benchmark, but the exact
code probably does not need to be in the commit message.
The plugin is abandoned and no longer working with new pytest versions.
I also reordered a bit to put pytest-rerunfailures first since it seems most maintained and is under pytest-dev.
and also fixes a regression in pytest 8.0.0 where `setup_method` crashes
if the class has static or class method tests.
It is allowed to have a test class with static/class methods which
request non-static/class method fixtures (including `setup_method`
xunit-fixture). I take it as a given that we need to support this
somewhat odd scenario (stdlib unittest also supports it).
This raises a question -- when a staticmethod test requests a bound
fixture, what is that fixture's `self`?
stdlib unittest says - a fresh instance for the test.
Previously, pytest said - some instance that is shared by all
static/class methods. This is definitely broken since it breaks test
isolation.
Change pytest to behave like stdlib unittest here.
In practice, this means stopping to rely on `self.obj.__self__` to get
to the instance from the test function's binding. This doesn't work
because staticmethods are not bound to anything.
Instead, keep the instance explicitly and use that.
BTW, I think this will allow us to change `Class`'s fixture collection
(`parsefactories`) to happen on the class itself instead of a class
instance, allowing us to avoid one class instantiation. But needs more
work.
Fixes#12065.
- Separate the requesting from the requested.
- Avoid the term "factory", I think most people don't distinguish
between "fixture" and "fixture function" (i.e. "factory") and would
find the term "factory" unfamiliar.
There are two non-optimal things in the current way scope checking is
done:
- It runs on `SubRequest`, but doesn't use the `SubRequest's scope,
which is confusing. Instead it takes `invoking_scope` and
`requested_scope`.
- Because `_check_scope` is only defined on `SubRequest` and not
`TopRequest`, `_compute_fixture_value` first creates the `SubRequest`
only then checks the scope (hence the need for the previous point).
Instead, also define `_check_scope` on `TopRequest` (always valid), and
remove `invoking_scope`, using `self._scope` instead.
Previously, the `obj` of a `TestCaseFunction` (the unittest plugin item
type) was the unbound method. This is unlike regular `Class` where the
`obj` is a bound method to a fresh instance.
This difference necessitated several special cases in in places outside
of the unittest plugin, such as `FixtureDef` and `FixtureRequest`, and
made things a bit harder to understand.
Instead, match how the python plugin does it, including collecting
fixtures from a fresh instance.
The downside is that now this instance for fixture-collection is kept
around in memory, but it's the same as `Class` so nothing new. Users
should only initialize stuff in `setUp`/`setUpClass` and similar
methods, and not in `__init__` which is generally off-limits in
`TestCase` subclasses.
I am not sure why there was a difference in the first place, though I
will say the previous unittest approach is probably the preferable one,
but first let's get consistency.
This adds a GHA job that reliably determines if all the required
dependencies have succeeded or not.
It also allows to reduce the list of required branch protection CI
statuses to just one — `check`. This reduces the maintenance burden
by a lot and have been battle-tested across a small bunch of projects
in its action form and in-house implementations of other people.
I was curious about the spread of use. And I've just discovered that
it is now in use in aiohttp (and other aio-libs projects), CherryPy,
some of the Ansible repositories, all of the jaraco's projects (like
`setuptools`, `importlib_metadata`), some PyCQA, PyCA and pytest
projects, a few AWS Labs projects. Admittedly, I maintain a few of
these but it seems to address some of the pain folks have:
https://github.com/jaraco/skeleton/pull/55#issuecomment-1106638475.
The story behind this is explained in more detail at
https://github.com/marketplace/actions/alls-green#why.
2023-12-07 01:29:02 +01:00
339 changed files with 7934 additions and 4276 deletions
stale-issue-message:"This issue is stale because it has been open for 14 days with no activity."
close-issue-message:"This issue was closed because it has been inactive for 7 days since being marked as stale."
stale-issue-message:"This issue is stale because it has the `status: needs information` label and requested follow-up information was not provided for 14 days."
close-issue-message:"This issue was closed because it has the `status: needs information` label and follow-up information has not been provided for 7 days since being marked as stale."
Added `--xfail-tb` flag, which turns on traceback output for XFAIL results.
* If the `--xfail-tb` flag is not sent, tracebacks for XFAIL results are NOT shown.
* The style of traceback for XFAIL is set with `--tb`, and can be `auto|long|short|line|native|no`.
* Note: Even if you have `--xfail-tb` set, you won't see them if `--tb=no`.
Some history:
With pytest 8.0, `-rx` or `-ra` would not only turn on summary reports for xfail, but also report the tracebacks for xfail results. This caused issues with some projects that utilize xfail, but don't want to see all of the xfail tracebacks.
This change detaches xfail tracebacks from `-rx`, and now we turn on xfail tracebacks with `--xfail-tb`. With this, the default `-rx`/ `-ra` behavior is identical to pre-8.0 with respect to xfail tracebacks. While this is a behavior change, it brings default behavior back to pre-8.0.0 behavior, which ultimately was considered the better course of action.
Fix crash with `assert testcase is not None` assertion failure when re-running unittest tests using plugins like pytest-rerunfailures. Regressed in 8.2.2.
Fixed a crash when returning category ``"error"`` or ``"failed"`` with a custom test status from :hook:`pytest_report_teststatus` hook -- :user:`pbrezina`.
pytest is actively evolving and is a project that has been decades in the making,
we keep learning about new and better structures to express different details about testing.
Pytest is an actively evolving project that has been decades in the making.
We keep learning about new and better structures to express different details about testing.
While we implement those modifications we try to ensure an easy transition and don't want to impose unnecessary churn on our users and community/plugin authors.
While we implement those modifications, we try to ensure an easy transition and don't want to impose unnecessary churn on our users and community/plugin authors.
As of now, pytest considers multiple types of backward compatibility transitions:
a) trivial: APIs which trivially translate to the new mechanism,
and do not cause problematic changes.
a) trivial: APIs that trivially translate to the new mechanism and do not cause problematic changes.
We try to support those indefinitely while encouraging users to switch to newer/better mechanisms through documentation.
We try to support those indefinitely while encouraging users to switch to newer or better mechanisms through documentation.
b) transitional: the old and new API don't conflict
and we can help users transition by using warnings, while supporting both for a prolonged time.
b) transitional: the old and new APIs don't conflict, and we can help users transition by using warnings while supporting both for a prolonged period of time.
We will only start the removal of deprecated functionality in major releases (e.g. if we deprecate something in 3.0 we will start to remove it in 4.0), and keep it around for at least two minor releases (e.g. if we deprecate something in 3.9 and 4.0 is the next release, we start to remove it in 5.0, not in 4.0).
We will only start the removal of deprecated functionality in major releases (e.g., if we deprecate something in 3.0, we will start to remove it in 4.0), and keep it around for at least two minor releases (e.g., if we deprecate something in 3.9 and 4.0 is the next release, we start to remove it in 5.0, not in 4.0).
A deprecated feature scheduled to be removed in major version X will use the warning class `PytestRemovedInXWarning` (a subclass of :class:`~pytest.PytestDeprecationWarning`).
When the deprecation expires (e.g. 4.0 is released), we won't remove the deprecated functionality immediately, but will use the standard warning filters to turn `PytestRemovedInXWarning` (e.g. `PytestRemovedIn4Warning`) into **errors** by default. This approach makes it explicit that removal is imminent, and still gives you time to turn the deprecated feature into a warning instead of an error so it can be dealt with in your own time. In the next minor release (e.g. 4.1), the feature will be effectively removed.
When the deprecation expires (e.g., 4.0 is released), we won't remove the deprecated functionality immediately but will use the standard warning filters to turn `PytestRemovedInXWarning` (e.g.,`PytestRemovedIn4Warning`) into **errors** by default. This approach makes it explicit that removal is imminent and still gives you time to turn the deprecated feature into a warning instead of an error so it can be dealt with in your own time. In the next minor release (e.g., 4.1), the feature will be effectively removed.
c) true breakage: should only be considered when normal transition is unreasonably unsustainable and would offset important development/features by years.
In addition, they should be limited to APIs where the number of actual users is very small (for example only impacting some plugins), and can be coordinated with the community in advance.
c) True breakage should only be considered when a normal transition is unreasonably unsustainable and would offset important developments or features by years. In addition, they should be limited to APIs where the number of actual users is very small (for example, only impacting some plugins) and can be coordinated with the community in advance.
Keeping backwards compatibility has a very high priority in the pytest project. Although we have deprecated functionality over the years, most of it is still supported. All deprecations in pytest were done because simpler or more efficient ways of accomplishing the same tasks have emerged, making the old way of doing things unnecessary.
With the pytest 3.0 release we introduced a clear communication scheme for when we will actually remove the old busted joint and politely ask you to use the new hotness instead, while giving you enough time to adjust your tests or raise concerns if there are valid reasons to keep deprecated functionality around.
With the pytest 3.0 release, we introduced a clear communication scheme for when we will actually remove the old busted joint and politely ask you to use the new hotness instead, while giving you enough time to adjust your tests or raise concerns if there are valid reasons to keep deprecated functionality around.
To communicate changes we issue deprecation warnings using a custom warning hierarchy (see :ref:`internal-warnings`). These warnings may be suppressed using the standard means: ``-W`` command-line flag or ``filterwarnings`` ini options (see :ref:`warnings`), but we suggest to use these sparingly and temporarily, and heed the warnings when possible.
To communicate changes, we issue deprecation warnings using a custom warning hierarchy (see :ref:`internal-warnings`). These warnings may be suppressed using the standard means: ``-W`` command-line flag or ``filterwarnings`` ini options (see :ref:`warnings`), but we suggest to use these sparingly and temporarily, and heed the warnings when possible.
We will only start the removal of deprecated functionality in major releases (e.g. if we deprecate something in 3.0 we will start to remove it in 4.0), and keep it around for at least two minor releases (e.g. if we deprecate something in 3.9 and 4.0 is the next release, we start to remove it in 5.0, not in 4.0).
We will only start the removal of deprecated functionality in major releases (e.g. if we deprecate something in 3.0, we will start to remove it in 4.0), and keep it around for at least two minor releases (e.g. if we deprecate something in 3.9 and 4.0 is the next release, we start to remove it in 5.0, not in 4.0).
When the deprecation expires (e.g. 4.0 is released), we won't remove the deprecated functionality immediately, but will use the standard warning filters to turn them into **errors** by default. This approach makes it explicit that removal is imminent, and still gives you time to turn the deprecated feature into a warning instead of an error so it can be dealt with in your own time. In the next minor release (e.g. 4.1), the feature will be effectively removed.
-`#12355 <https://github.com/pytest-dev/pytest/issues/12355>`_: Fix possible catastrophic performance slowdown on a certain parametrization pattern involving many higher-scoped parameters.
-`#12367 <https://github.com/pytest-dev/pytest/issues/12367>`_: Fix a regression in pytest 8.2.0 where unittest class instances (a fresh one is created for each test) were not released promptly on test teardown but only on session teardown.
-`#12381 <https://github.com/pytest-dev/pytest/issues/12381>`_: Fix possible "Directory not empty" crashes arising from concurent cache dir (``.pytest_cache``) creation. Regressed in pytest 8.2.0.
Improved Documentation
----------------------
-`#12290 <https://github.com/pytest-dev/pytest/issues/12290>`_: Updated Sphinx theme to use Furo instead of Flask, enabling Dark mode theme.
-`#12356 <https://github.com/pytest-dev/pytest/issues/12356>`_: Added a subsection to the documentation for debugging flaky tests to mention
lack of thread safety in pytest as a possible source of flakyness.
-`#12363 <https://github.com/pytest-dev/pytest/issues/12363>`_: The documentation webpages now links to a canonical version to reduce outdated documentation in search engine results.
pytest 8.2.1 (2024-05-19)
=========================
Improvements
------------
-`#12334 <https://github.com/pytest-dev/pytest/issues/12334>`_: Support for Python 3.13 (beta1 at the time of writing).
Bug Fixes
---------
-`#12120 <https://github.com/pytest-dev/pytest/issues/12120>`_: Fix `PermissionError` crashes arising from directories which are not selected on the command-line.
-`#12191 <https://github.com/pytest-dev/pytest/issues/12191>`_: Keyboard interrupts and system exits are now properly handled during the test collection.
-`#12300 <https://github.com/pytest-dev/pytest/issues/12300>`_: Fixed handling of 'Function not implemented' error under squashfuse_ll, which is a different way to say that the mountpoint is read-only.
-`#12308 <https://github.com/pytest-dev/pytest/issues/12308>`_: Fix a regression in pytest 8.2.0 where the permissions of automatically-created ``.pytest_cache`` directories became ``rwx------`` instead of the expected ``rwxr-xr-x``.
Trivial/Internal Changes
------------------------
-`#12333 <https://github.com/pytest-dev/pytest/issues/12333>`_: pytest releases are now attested using the recent `Artifact Attestation <https://github.blog/2024-05-02-introducing-artifact-attestations-now-in-public-beta/>`_ support from GitHub, allowing users to verify the provenance of pytest's sdist and wheel artifacts.
pytest 8.2.0 (2024-04-27)
=========================
Breaking Changes
----------------
-`#12089 <https://github.com/pytest-dev/pytest/pull/12089>`_: pytest now requires that :class:`unittest.TestCase` subclasses can be instantiated freely using ``MyTestCase('runTest')``.
If the class doesn't allow this, you may see an error during collection such as ``AttributeError: 'MyTestCase' object has no attribute 'runTest'``.
Classes which do not override ``__init__``, or do not access the test method in ``__init__`` using ``getattr`` or similar, are unaffected.
Classes which do should take care to not crash when ``"runTest"`` is given, as is shown in `unittest.TestCases's implementation <https://github.com/python/cpython/blob/51aefc5bf907ddffaaf083ded0de773adcdf08c8/Lib/unittest/case.py#L419-L426>`_.
Alternatively, consider using :meth:`setUp <unittest.TestCase.setUp>` instead of ``__init__``.
If you run into this issue using ``tornado.AsyncTestCase``, please see `issue 12263 <https://github.com/pytest-dev/pytest/issues/12263>`_.
If you run into this issue using an abstract ``TestCase`` subclass, please see `issue 12275 <https://github.com/pytest-dev/pytest/issues/12275>`_.
Historical note: the effect of this change on custom TestCase implementations was not properly considered initially, this is why it was done in a minor release. We apologize for the inconvenience.
Deprecations
------------
-`#12069 <https://github.com/pytest-dev/pytest/issues/12069>`_: A deprecation warning is now raised when implementations of one of the following hooks request a deprecated ``py.path.local`` parameter instead of the ``pathlib.Path`` parameter which replaced it:
- :hook:`pytest_ignore_collect` - the ``path`` parameter - use ``collection_path`` instead.
-:hook:`pytest_collect_file` - the ``path`` parameter - use ``file_path`` instead.
-:hook:`pytest_pycollect_makemodule` - the ``path`` parameter - use ``module_path`` instead.
-:hook:`pytest_report_header` - the ``startdir`` parameter - use ``start_path`` instead.
-:hook:`pytest_report_collectionfinish` - the ``startdir`` parameter - use ``start_path`` instead.
The replacement parameters are available since pytest 7.0.0.
The old parameters will be removed in pytest 9.0.0.
See :ref:`legacy-path-hooks-deprecated` for more details.
Features
--------
-`#11871 <https://github.com/pytest-dev/pytest/issues/11871>`_: Added support for reading command line arguments from a file using the prefix character ``@``, like e.g.: ``pytest @tests.txt``. The file must have one argument per line.
See :ref:`Read arguments from file <args-from-file>` for details.
Improvements
------------
-`#11523 <https://github.com/pytest-dev/pytest/issues/11523>`_: :func:`pytest.importorskip` will now issue a warning if the module could be found, but raised :class:`ImportError` instead of :class:`ModuleNotFoundError`.
The warning can be suppressed by passing ``exc_type=ImportError`` to :func:`pytest.importorskip`.
See :ref:`import-or-skip-import-error` for details.
-`#11728 <https://github.com/pytest-dev/pytest/issues/11728>`_: For ``unittest``-based tests, exceptions during class cleanup (as raised by functions registered with :meth:`TestCase.addClassCleanup <unittest.TestCase.addClassCleanup>`) are now reported instead of silently failing.
-`#11777 <https://github.com/pytest-dev/pytest/issues/11777>`_: Text is no longer truncated in the ``short test summary info`` section when ``-vv`` is given.
-`#12112 <https://github.com/pytest-dev/pytest/issues/12112>`_: Improved namespace packages detection when :confval:`consider_namespace_packages` is enabled, covering more situations (like editable installs).
-`#9502 <https://github.com/pytest-dev/pytest/issues/9502>`_: Added :envvar:`PYTEST_VERSION` environment variable which is defined at the start of the pytest session and undefined afterwards. It contains the value of ``pytest.__version__``, and among other things can be used to easily check if code is running from within a pytest run.
Bug Fixes
---------
-`#12065 <https://github.com/pytest-dev/pytest/issues/12065>`_: Fixed a regression in pytest 8.0.0 where test classes containing ``setup_method`` and tests using ``@staticmethod`` or ``@classmethod`` would crash with ``AttributeError: 'NoneType' object has no attribute 'setup_method'``.
Now the :attr:`request.instance <pytest.FixtureRequest.instance>` attribute of tests using ``@staticmethod`` and ``@classmethod`` is no longer ``None``, but a fresh instance of the class, like in non-static methods.
Previously it was ``None``, and all fixtures of such tests would share a single ``self``.
-`#12135 <https://github.com/pytest-dev/pytest/issues/12135>`_: Fixed issue where fixtures adding their finalizer multiple times to fixtures they request would cause unreliable and non-intuitive teardown ordering in some instances.
-`#12194 <https://github.com/pytest-dev/pytest/issues/12194>`_: Fixed a bug with ``--importmode=importlib`` and ``--doctest-modules`` where child modules did not appear as attributes in parent modules.
-`#1489 <https://github.com/pytest-dev/pytest/issues/1489>`_: Fixed some instances where teardown of higher-scoped fixtures was not happening in the reverse order they were initialized in.
Trivial/Internal Changes
------------------------
-`#12069 <https://github.com/pytest-dev/pytest/issues/12069>`_: ``pluggy>=1.5.0`` is now required.
-`#12167 <https://github.com/pytest-dev/pytest/issues/12167>`_: :ref:`cache <cache>`: create supporting files (``CACHEDIR.TAG``, ``.gitignore``, etc.) in a temporary directory to provide atomic semantics.
pytest 8.1.2 (2024-04-26)
=========================
@@ -69,7 +232,7 @@ Improvements
-`#11311 <https://github.com/pytest-dev/pytest/issues/11311>`_: When using ``--override-ini`` for paths in invocations without a configuration file defined, the current working directory is used
as the relative directory.
Previoulsy this would raise an :class:`AssertionError`.
Previously this would raise an :class:`AssertionError`.
-`#11475 <https://github.com/pytest-dev/pytest/issues/11475>`_: :ref:`--import-mode=importlib <import-mode-importlib>` now tries to import modules using the standard import mechanism (but still without changing :py:data:`sys.path`), falling back to importing modules directly only if that fails.
@@ -105,7 +268,7 @@ Bug Fixes
-`#11904 <https://github.com/pytest-dev/pytest/issues/11904>`_: Fixed a regression in pytest 8.0.0 that would cause test collection to fail due to permission errors when using ``--pyargs``.
This change improves the collection tree for tests specified using ``--pyargs``, see :pull:`12043` for a comparison with pytest 8.0 and <8.
This change improves the collection tree for tests specified using ``--pyargs``, see :pr:`12043` for a comparison with pytest 8.0 and <8.
-`#12011 <https://github.com/pytest-dev/pytest/issues/12011>`_: Fixed a regression in 8.0.1 whereby ``setup_module`` xunit-style fixtures are not executed when ``--doctest-modules`` is passed.
@@ -1259,7 +1422,7 @@ Bug Fixes
tests/link -> tests/real
running ``pytest tests`` now imports the conftest twice, once as ``tests/real/conftest.py`` and once as ``tests/link/conftest.py``.
This is a fix to match a similar change made to test collection itself in pytest 6.0 (see :pull:`6523` for details).
This is a fix to match a similar change made to test collection itself in pytest 6.0 (see :pr:`6523` for details).
-`#9626 <https://github.com/pytest-dev/pytest/issues/9626>`_: Fixed count of selected tests on terminal collection summary when there were errors or skipped modules.
@@ -1312,7 +1475,7 @@ Deprecations
``__init__`` method, they should take ``**kwargs``. See
:ref:`uncooperative-constructors-deprecated` for details.
Note that a deprection warning is only emitted when there is a conflict in the
Note that a deprecation warning is only emitted when there is a conflict in the
arguments pytest expected to pass. This deprecation was already part of pytest
7.0.0rc1 but wasn't documented.
@@ -1354,7 +1517,7 @@ Breaking Changes
-`#7259 <https://github.com/pytest-dev/pytest/issues/7259>`_: The :ref:`Node.reportinfo() <non-python tests>` function first return value type has been expanded from `py.path.local | str` to `os.PathLike[str] | str`.
Most plugins which refer to `reportinfo()` only define it as part of a custom :class:`pytest.Item` implementation.
Since `py.path.local` is an `os.PathLike[str]`, these plugins are unaffacted.
Since `py.path.local` is an `os.PathLike[str]`, these plugins are unaffected.
Plugins and users which call `reportinfo()`, use the first return value and interact with it as a `py.path.local`, would need to adjust by calling `py.path.local(fspath)`.
Although preferably, avoid the legacy `py.path.local` and use `pathlib.Path`, or use `item.location` or `item.path`, instead.
@@ -1862,7 +2025,7 @@ Bug Fixes
the ``tmp_path``/``tmpdir`` fixture). Now the directories are created with
private permissions.
pytest used to silently use a pre-existing ``/tmp/pytest-of-<username>`` directory,
pytest used to silently use a preexisting ``/tmp/pytest-of-<username>`` directory,
even if owned by another user. This means another user could pre-create such a
directory and gain control of another user's temporary directory. Now such a
condition results in an error.
@@ -2428,7 +2591,7 @@ Breaking Changes
Resolving symlinks for the current directory and during collection was introduced as a bugfix in 3.9.0, but it actually is a new feature which had unfortunate consequences in Windows and surprising results in other platforms.
The team decided to step back on resolving symlinks at all, planning to review this in the future with a more solid solution (see discussion in
:pull:`6523` for details).
:pr:`6523` for details).
This might break test suites which made use of this feature; the fix is to create a symlink
for the entire test tree, and not only to partial files/tress as it was possible previously.
@@ -2589,7 +2752,7 @@ Features
also changes ``sys.modules`` as a side-effect), which works but has a number of drawbacks, like requiring test modules
that don't live in packages to have unique names (as they need to reside under a unique name in ``sys.modules``).
``--import-mode=importlib`` uses more finegrained import mechanisms from ``importlib`` which don't
``--import-mode=importlib`` uses more fine-grained import mechanisms from ``importlib`` which don't
require pytest to change ``sys.path`` or ``sys.modules`` at all, eliminating much of the drawbacks
of the previous mode.
@@ -2606,7 +2769,7 @@ Improvements
------------
-:issue:`4375`: The ``pytest`` command now suppresses the ``BrokenPipeError`` error message that
is printed to stderr when the output of ``pytest`` is piped and and the pipe is
is printed to stderr when the output of ``pytest`` is piped and the pipe is
closed by the piped-to program (common examples are ``less`` and ``head``).
@@ -2711,7 +2874,7 @@ Bug Fixes
-:issue:`6871`: Fix crash with captured output when using :fixture:`capsysbinary`.
-:issue:`6909`: Revert the change introduced by :pull:`6330`, which required all arguments to ``@pytest.mark.parametrize`` to be explicitly defined in the function signature.
-:issue:`6909`: Revert the change introduced by :pr:`6330`, which required all arguments to ``@pytest.mark.parametrize`` to be explicitly defined in the function signature.
The intention of the original change was to remove what was expected to be an unintended/surprising behavior, but it turns out many people relied on it, so the restriction has been reverted.
@@ -2881,7 +3044,7 @@ pytest 5.4.1 (2020-03-13)
Bug Fixes
---------
-:issue:`6909`: Revert the change introduced by :pull:`6330`, which required all arguments to ``@pytest.mark.parametrize`` to be explicitly defined in the function signature.
-:issue:`6909`: Revert the change introduced by :pr:`6330`, which required all arguments to ``@pytest.mark.parametrize`` to be explicitly defined in the function signature.
The intention of the original change was to remove what was expected to be an unintended/surprising behavior, but it turns out many people relied on it, so the restriction has been reverted.
@@ -2908,7 +3071,7 @@ Breaking Changes
This hook has been marked as deprecated and not been even called by pytest for over 10 years now.
-:issue:`6673`: Reversed / fix meaning of "+/-" in error diffs. "-" means that sth. expected is missing in the result and "+" means that there are unexpected extras in the result.
-:issue:`6673`: Reversed / fix meaning of "+/-" in error diffs. "-" means that something expected is missing in the result and "+" means that there are unexpected extras in the result.
-:issue:`6737`: The ``cached_result`` attribute of ``FixtureDef`` is now set to ``None`` when
@@ -3197,7 +3360,9 @@ Bug Fixes
-:issue:`5914`: pytester: fix :py:func:`~pytest.LineMatcher.no_fnmatch_line` when used after positive matching.
-:issue:`6082`: Fix line detection for doctest samples inside:py:class:`python:property` docstrings, as a workaround to :bpo:`17446`.
-:issue:`6082`: Fix line detection for doctest samples inside
:py:class:`python:property` docstrings, as a workaround to
:issue:`python/cpython#61648`.
-:issue:`6254`: Fix compatibility with pytest-parallel (regression in pytest 5.3.0).
@@ -3904,7 +4069,7 @@ Bug Fixes
(``--collect-only``) when ``--log-cli-level`` is used.
-:issue:`5389`: Fix regressions of :pull:`5063` for ``importlib_metadata.PathDistribution`` which have their ``files`` attribute being ``None``.
-:issue:`5389`: Fix regressions of :pr:`5063` for ``importlib_metadata.PathDistribution`` which have their ``files`` attribute being ``None``.
-:issue:`5390`: Fix regression where the ``obj`` attribute of ``TestCase`` items was no longer bound to methods.
@@ -4105,7 +4270,7 @@ Bug Fixes
(``--collect-only``) when ``--log-cli-level`` is used.
-:issue:`5389`: Fix regressions of :pull:`5063` for ``importlib_metadata.PathDistribution`` which have their ``files`` attribute being ``None``.
-:issue:`5389`: Fix regressions of :pr:`5063` for ``importlib_metadata.PathDistribution`` which have their ``files`` attribute being ``None``.
-:issue:`5390`: Fix regression where the ``obj`` attribute of ``TestCase`` items was no longer bound to methods.
@@ -4513,7 +4678,7 @@ Bug Fixes
Improved Documentation
----------------------
-:issue:`4974`: Update docs for ``pytest_cmdline_parse`` hook to note availability liminations
-:issue:`4974`: Update docs for ``pytest_cmdline_parse`` hook to note availability limitations
@@ -6371,7 +6536,7 @@ Features
Bug Fixes
---------
- Fix hanging pexpect test on MacOS by using flush() instead of wait().
- Fix hanging pexpect test on macOS by using flush() instead of wait().
(:issue:`2022`)
- Fix restoring Python state after in-process pytest runs with the
@@ -6419,7 +6584,7 @@ Trivial/Internal Changes
------------------------
- Show a simple and easy error when keyword expressions trigger a syntax error
(for example, ``"-k foo and import"`` will show an error that you cannot use
(for example, ``"-k foo and import"`` will show an error that you cannot use
the ``import`` keyword in expressions). (:issue:`2953`)
- Change parametrized automatic test id generation to use the ``__name__``
@@ -7066,10 +7231,10 @@ New Features
* Added ``junit_suite_name`` ini option to specify root ``<testsuite>`` name for JUnit XML reports (:issue:`533`).
* Added an ini option ``doctest_encoding`` to specify which encoding to use for doctest files.
Thanks :user:`wheerd` for the PR (:pull:`2101`).
Thanks :user:`wheerd` for the PR (:pr:`2101`).
*``pytest.warns`` now checks for subclass relationship rather than
class equality. Thanks :user:`lesteve` for the PR (:pull:`2166`)
class equality. Thanks :user:`lesteve` for the PR (:pr:`2166`)
*``pytest.raises`` now asserts that the error message matches a text or regex
with the ``match`` keyword argument. Thanks :user:`Kriechi` for the PR.
@@ -7097,7 +7262,7 @@ Changes
the failure. (:issue:`2228`) Thanks to :user:`kkoukiou` for the PR.
* Testcase reports with a ``url`` attribute will now properly write this to junitxml.
Thanks :user:`fushi` for the PR (:pull:`1874`).
Thanks :user:`fushi` for the PR (:pr:`1874`).
* Remove common items from dict comparison output when verbosity=1. Also update
the truncation message to make it clearer that pytest truncates all
@@ -7106,7 +7271,7 @@ Changes
*``--pdbcls`` no longer implies ``--pdb``. This makes it possible to use
``addopts=--pdbcls=module.SomeClass`` on ``pytest.ini``. Thanks :user:`davidszotten` for
the PR (:pull:`1952`).
the PR (:pr:`1952`).
* fix :issue:`2013`: turn RecordedWarning into ``namedtuple``,
to give it a comprehensible repr while preventing unwarranted modification.
@@ -7360,7 +7525,7 @@ Bug Fixes
a sequence of strings) when modules are considered for assertion rewriting.
Due to this bug, much more modules were being rewritten than necessary
if a test suite uses ``pytest_plugins`` to load internal plugins (:issue:`1888`).
Thanks :user:`jaraco` for the report and :user:`nicoddemus` for the PR (:pull:`1891`).
Thanks :user:`jaraco` for the report and :user:`nicoddemus` for the PR (:pr:`1891`).
* Do not call tearDown and cleanups when running tests from
``unittest.TestCase`` subclasses with ``--pdb``
@@ -7415,12 +7580,12 @@ time or change existing behaviors in order to make them less surprising/more use
*``--nomagic``: use ``--assert=plain`` instead;
*``--report``: use ``-r`` instead;
Thanks to :user:`RedBeardCode` for the PR (:pull:`1664`).
Thanks to :user:`RedBeardCode` for the PR (:pr:`1664`).
* ImportErrors in plugins now are a fatal error instead of issuing a
pytest warning (:issue:`1479`). Thanks to :user:`The-Compiler` for the PR.
* Removed support code for Python 3 versions < 3.3 (:pull:`1627`).
* Removed support code for Python 3 versions < 3.3 (:pr:`1627`).
* Removed all ``py.test-X*`` entry points. The versioned, suffixed entry points
were never documented and a leftover from a pre-virtualenv era. These entry
@@ -7431,19 +7596,19 @@ time or change existing behaviors in order to make them less surprising/more use
*``pytest.skip()`` now raises an error when used to decorate a test function,
as opposed to its original intent (to imperatively skip a test inside a test function). Previously
this usage would cause the entire module to be skipped (:issue:`607`).
Thanks :user:`omarkohl` for the complete PR (:pull:`1519`).
Thanks :user:`omarkohl` for the complete PR (:pr:`1519`).
* Exit tests if a collection error occurs. A poll indicated most users will hit CTRL-C
anyway as soon as they see collection errors, so pytest might as well make that the default behavior (:issue:`1421`).
A ``--continue-on-collection-errors`` option has been added to restore the previous behaviour.
Thanks :user:`olegpidsadnyi` and :user:`omarkohl` for the complete PR (:pull:`1628`).
Thanks :user:`olegpidsadnyi` and :user:`omarkohl` for the complete PR (:pr:`1628`).
* Renamed the pytest ``pdb`` module (plugin) into ``debugging`` to avoid clashes with the builtin ``pdb`` module.
* Raise a helpful failure message when requesting a parametrized fixture at runtime,
e.g. with ``request.getfixturevalue``. Previously these parameters were simply
never defined, so a fixture decorated like ``@pytest.fixture(params=[0, 1, 2])``
only ran once (:pull:`460`).
only ran once (:pr:`460`).
Thanks to :user:`nikratio` for the bug report, :user:`RedBeardCode` and :user:`tomviner` for the PR.
*``_pytest.monkeypatch.monkeypatch`` class has been renamed to ``_pytest.monkeypatch.MonkeyPatch``
@@ -7461,7 +7626,7 @@ time or change existing behaviors in order to make them less surprising/more use
* New ``doctest_namespace`` fixture for injecting names into the
namespace in which doctests run.
Thanks :user:`milliams` for the complete PR (:pull:`1428`).
Thanks :user:`milliams` for the complete PR (:pr:`1428`).
* New ``--doctest-report`` option available to change the output format of diffs
when running (failing) doctests (implements :issue:`1749`).
@@ -7469,23 +7634,23 @@ time or change existing behaviors in order to make them less surprising/more use
* New ``name`` argument to ``pytest.fixture`` decorator which allows a custom name
for a fixture (to solve the funcarg-shadowing-fixture problem).
Thanks :user:`novas0x2a` for the complete PR (:pull:`1444`).
Thanks :user:`novas0x2a` for the complete PR (:pr:`1444`).
* New ``approx()`` function for easily comparing floating-point numbers in
tests.
Thanks :user:`kalekundert` for the complete PR (:pull:`1441`).
Thanks :user:`kalekundert` for the complete PR (:pr:`1441`).
* Ability to add global properties in the final xunit output file by accessing
the internal ``junitxml`` plugin (experimental).
Thanks :user:`tareqalayan` for the complete PR :pull:`1454`).
Thanks :user:`tareqalayan` for the complete PR :pr:`1454`).
* New ``ExceptionInfo.match()`` method to match a regular expression on the
string representation of an exception (:issue:`372`).
Thanks :user:`omarkohl` for the complete PR (:pull:`1502`).
Thanks :user:`omarkohl` for the complete PR (:pr:`1502`).
*``__tracebackhide__`` can now also be set to a callable which then can decide
whether to filter the traceback based on the ``ExceptionInfo`` object passed
to it. Thanks :user:`The-Compiler` for the complete PR (:pull:`1526`).
to it. Thanks :user:`The-Compiler` for the complete PR (:pr:`1526`).
* New ``pytest_make_parametrize_id(config, val)`` hook which can be used by plugins to provide
friendly strings for custom types.
@@ -7503,7 +7668,7 @@ time or change existing behaviors in order to make them less surprising/more use
* Introduce ``pytest`` command as recommended entry point. Note that ``py.test``
still works and is not scheduled for removal. Closes proposal
:issue:`1629`. Thanks :user:`obestwalter` and :user:`davehunt` for the complete PR
(:pull:`1633`).
(:pr:`1633`).
* New cli flags:
@@ -7547,19 +7712,19 @@ time or change existing behaviors in order to make them less surprising/more use
* Change ``report.outcome`` for ``xpassed`` tests to ``"passed"`` in non-strict
mode and ``"failed"`` in strict mode. Thanks to :user:`hackebrot` for the PR
(:pull:`1795`) and :user:`gprasad84` for report (:issue:`1546`).
(:pr:`1795`) and :user:`gprasad84` for report (:issue:`1546`).
* Tests marked with ``xfail(strict=False)`` (the default) now appear in
JUnitXML reports as passing tests instead of skipped.
Thanks to :user:`hackebrot` for the PR (:pull:`1795`).
Thanks to :user:`hackebrot` for the PR (:pr:`1795`).
* Highlight path of the file location in the error report to make it easier to copy/paste.
Thanks :user:`suzaku` for the PR (:pull:`1778`).
Thanks :user:`suzaku` for the PR (:pr:`1778`).
* Fixtures marked with ``@pytest.fixture`` can now use ``yield`` statements exactly like
those marked with the ``@pytest.yield_fixture`` decorator. This change renders
``@pytest.yield_fixture`` deprecated and makes ``@pytest.fixture`` with ``yield`` statements
the preferred way to write teardown code (:pull:`1461`).
the preferred way to write teardown code (:pr:`1461`).
Thanks :user:`csaftoiu` for bringing this to attention and :user:`nicoddemus` for the PR.
* Explicitly passed parametrize ids do not get escaped to ascii (:issue:`1351`).
@@ -7570,11 +7735,11 @@ time or change existing behaviors in order to make them less surprising/more use
Thanks :user:`nicoddemus` for the PR.
*``pytest_terminal_summary`` hook now receives the ``exitstatus``
of the test session as argument. Thanks :user:`blueyed` for the PR (:pull:`1809`).
of the test session as argument. Thanks :user:`blueyed` for the PR (:pr:`1809`).
* Parametrize ids can accept ``None`` as specific test id, in which case the
automatically generated id for that argument will be used.
Thanks :user:`palaviv` for the complete PR (:pull:`1468`).
Thanks :user:`palaviv` for the complete PR (:pr:`1468`).
* The parameter to xunit-style setup/teardown methods (``setup_method``,
``setup_module``, etc.) is now optional and may be omitted.
@@ -7582,32 +7747,32 @@ time or change existing behaviors in order to make them less surprising/more use
* Improved automatic id generation selection in case of duplicate ids in
parametrize.
Thanks :user:`palaviv` for the complete PR (:pull:`1474`).
Thanks :user:`palaviv` for the complete PR (:pr:`1474`).
* Now pytest warnings summary is shown up by default. Added a new flag
``--disable-pytest-warnings`` to explicitly disable the warnings summary (:issue:`1668`).
* Make ImportError during collection more explicit by reminding
the user to check the name of the test module/package(s) (:issue:`1426`).
Thanks :user:`omarkohl` for the complete PR (:pull:`1520`).
Thanks :user:`omarkohl` for the complete PR (:pr:`1520`).
* Add ``build/`` and ``dist/`` to the default ``--norecursedirs`` list. Thanks
:user:`mikofski` for the report and :user:`tomviner` for the PR (:issue:`1544`).
*``pytest.raises`` in the context manager form accepts a custom
``message`` to raise when no exception occurred.
Thanks :user:`palaviv` for the complete PR (:pull:`1616`).
Thanks :user:`palaviv` for the complete PR (:pr:`1616`).
*``conftest.py`` files now benefit from assertion rewriting; previously it
was only available for test modules. Thanks :user:`flub`, :user:`sober7` and
:user:`nicoddemus` for the PR (:issue:`1619`).
* Text documents without any doctests no longer appear as "skipped".
Thanks :user:`graingert` for reporting and providing a full PR (:pull:`1580`).
Thanks :user:`graingert` for reporting and providing a full PR (:pr:`1580`).
* Ensure that a module within a namespace package can be found when it
is specified on the command line together with the ``--pyargs``
option. Thanks to :user:`taschini` for the PR (:pull:`1597`).
option. Thanks to :user:`taschini` for the PR (:pr:`1597`).
* Always include full assertion explanation during assertion rewriting. The previous behaviour was hiding
sub-expressions that happened to be ``False``, assuming this was redundant information.
@@ -7623,20 +7788,20 @@ time or change existing behaviors in order to make them less surprising/more use
Thanks :user:`nicoddemus` for the PR.
*``[pytest]`` sections in ``setup.cfg`` files should now be named ``[tool:pytest]``
to avoid conflicts with other distutils commands (see :pull:`567`). ``[pytest]`` sections in
to avoid conflicts with other distutils commands (see :pr:`567`). ``[pytest]`` sections in
``pytest.ini`` or ``tox.ini`` files are supported and unchanged.
Thanks :user:`nicoddemus` for the PR.
* Using ``pytest_funcarg__`` prefix to declare fixtures is considered deprecated and will be
removed in pytest-4.0 (:pull:`1684`).
removed in pytest-4.0 (:pr:`1684`).
Thanks :user:`nicoddemus` for the PR.
* Passing a command-line string to ``pytest.main()`` is considered deprecated and scheduled
for removal in pytest-4.0. It is recommended to pass a list of arguments instead (:pull:`1723`).
for removal in pytest-4.0. It is recommended to pass a list of arguments instead (:pr:`1723`).
* Rename ``getfuncargvalue`` to ``getfixturevalue``. ``getfuncargvalue`` is
still present but is now considered deprecated. Thanks to :user:`RedBeardCode` and :user:`tomviner`
for the PR (:pull:`1626`).
for the PR (:pr:`1626`).
*``optparse`` type usage now triggers DeprecationWarnings (:issue:`1740`).
@@ -7694,11 +7859,11 @@ time or change existing behaviors in order to make them less surprising/more use
:user:`tomviner` for the PR.
*``ConftestImportFailure`` now shows the traceback making it easier to
identify bugs in ``conftest.py`` files (:pull:`1516`). Thanks :user:`txomon` for
identify bugs in ``conftest.py`` files (:pr:`1516`). Thanks :user:`txomon` for
the PR.
* Text documents without any doctests no longer appear as "skipped".
Thanks :user:`graingert` for reporting and providing a full PR (:pull:`1580`).
Thanks :user:`graingert` for reporting and providing a full PR (:pr:`1580`).
* Fixed collection of classes with custom ``__new__`` method.
Fixes :issue:`1579`. Thanks to :user:`Stranger6667` for the PR.
@@ -7706,7 +7871,7 @@ time or change existing behaviors in order to make them less surprising/more use
# run this with $ pytest --collect-only test_collectonly.py
#
from__future__importannotations
deftest_function():
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
