* issue a warning when Items and Collector form a diamond
addresses #8435
* Apply suggestions from code review
Co-authored-by: Ran Benita <ran@unusedvar.com>
* Return support for the broken File/Item hybrids
* adds deprecation
* ads necessary support code in node construction
* fix incorrect mypy based assertions
* add docs for deprecation of Item/File inheritance
* warn when a non-cooperative ctor is encountered
* use getattr instead of cast to get the class __init__ for legacy ctors
* update documentation references for node inheritance
* clean up file+item inheritance test
enhance docs
move import upwards
Co-authored-by: Ran Benita <ran@unusedvar.com>
* cache: Move repetitive code to fixture
* cache: Explicitly test for chmod result
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Fix lint
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
In the "Getting Started" doc, the test class instance example for
instance sharing doesn't actually demonstrate anything about the
reinstantiation of the class. This change shows clearly how the instance
data isn't retained between test runs.
* [pre-commit.ci] pre-commit autoupdate
updates:
- [github.com/asottile/pyupgrade: v2.16.0 → v2.18.2](https://github.com/asottile/pyupgrade/compare/v2.16.0...v2.18.2)
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
`Parser` is used by many plugins and custom hooks. `OptionGroup` is
exposed by the `parser.addgroup` API.
The constructors of both are marked private, they are not meant to be
constructed directly.
Given a `RecursionError` traceback, the `Traceback.recursionindex()`
method returns the index of the frame which started the recursion
(repeated set of frames). To do so it attempts to check whether two
frames are equivalent. Just checking the function/line is not enough
because the recursion variable(s) might differ (e.g. imagine the numeric
value in a recursive factorial implementation). So it also compares the
`f_locals` (local variables) of each frame for equivalence.
For some reason, the locals comparison is wrapped in an `eval` whose
purpose is to evaluate the comparison in one of the compared frame's
context (locals + globals in scope). However, I can not think of any way
in which the global scope could affect the evaluation. It would have an
affect when the locals are bound but that's already done. So this seems
unnecessary - remove it.
* add feature to view fixture source location in invocations with --fixtures-per-test option
* remove unrelated changes to show_fixtures_per_test::test_doctest_items
* eshew the extraneous else in _show_fixtures_per_test.write_fixture
* enable the accommodation of multi-line docstring with --fixtures-per-test option
* add feature to view fixture source location in invocations with --fixtures
* add colour encoding to fixture location paths
* add changelog for #8606 fixing
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Replace for loop using the operator
* Replace for loop with a generator expression inside any()
* Replace for loop with a dictionary comprehension
* Use list comprehension
* Simplify arguments for range()
* Change newfuncargs variable to in-line dictionary comprehension
* is_ancestor: return base.is_relative_to(query)
* Remove unneeded import of pathlib
* Try using PurePath
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* Import PurePath on new line
* Revert and remove is_relative_to
Co-authored-by: Zachary Kneupper <zacharykneupper@Zacharys-MBP.lan>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Remove unnecessary else clause in repr_failure()
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: Zachary Kneupper <zacharykneupper@Zacharys-MBP.lan>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Expand command line argument examples with validation examples
This shows how to pass more detailed error messages back to the user when
they've provided an incorrect value for a custom command line option.
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
* Fix issue where TestCase.setUpClass is not called for test methods with a / in its name by checking if there is :: before the selected / or any :: after. Also added a test case for this.
* removed print statement that was added
* Change iterparentnodeids to consume / parts until the first ::. Then consider ::. Tests were changed to reflect this.
* Update changelog/8509.improvement.rst
Co-authored-by: Ran Benita <ran@unusedvar.com>
The `Sphinx-PyPI-upload` tool the section corresponds to hasn't been updated since 2009, and the new (c. 2018) PyPI doesn't support documentation uploads anyway.
* [pre-commit.ci] pre-commit autoupdate
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Ran Benita <ran@unusedvar.com>
* Add tmp_path_factory methods to the docs
Add the methods of tmp_path_factory to the docs.
Currently, only the class description appears in the docs.
* Add tmpdir_factory methods to the docs
pytest uses a root temp directory named `/tmp/pytest-of-<username>`. The
name is predictable, and the directory might already exists from a
previous run, so that's allowed.
This makes it possible for my_user to pre-create
`/tmp/pytest-of-another_user`, thus giving my_user control of
another_user's tempdir.
Prevent this scenario by adding a couple of safety checks. I believe
they are sufficient.
Testing the first check requires changing the owner, which requires
root permissions, so can't be unit-tested easily, but I checked it
manually.
(Written for a Unix system, but might be applicable to Windows as well).
pytest creates a root temporary directory under /tmp, named
`pytest-of-<username>`, and creates tmp_path's and other under it.
/tmp is shared between all users of the system.
This root temporary directory was created with 0o777&~umask permissions,
which usually becomes 0o755, meaning any user in the system could list
and read the files, which is undesirable.
Use 0o700 permissions instead. Also for subdirectories, because the root
dir is adjustable.
Calling pkg_resources.fixup_namespace_packages() is only needed for packages
that use pkg_resources.declare_namespace() and hence they already imported
pkg_resources. When pkg_resources is not imported, we don't need to use it.
This avoids an unneeded runtime dependency on setuptools.
The code is tested by test_syspath_prepend_with_namespace_packages,
behavior should remain unchanged, hence no new test was added.
When people drop pkg_resources from sys.modules, they are on their own.
If someone has a actual use case making this valid to support,
they can come in and provide a test, a reference and a fix.
Last time I "fixed" this I left a `\` at the start of the string
to avoid an initial newline, but didn't realize it was a raw string.
This should fix it now for good.
* Changed "Table Of Contents" link to a plain heading.
* Removed Customize, API Reference, 3rd party plugins links altogether.
* Added an "About the project" heading
* re-ordered and grouped how-to guides in how-to landing page
* retitled moved documents appropriately
* separated writing plugins/writing hook functions into two documents
Warnings like this:
python3.9/site-packages/setuptools/dist.py:634: UserWarning: Usage of
dash-separated 'upload-dir' will not be supported in future versions.
Please use the underscore name 'upload_dir' instead
This type is most prominent in `pytest.raises` and we should allow to
refer to it by a public name.
The type is not in a perfectly "exposable" state. In particular:
- The `traceback` property with type `Traceback` which is derived from
the `py.code` API and exposes a bunch more types transitively. This
stuff is *not* exported and probably won't be.
- The `getrepr` method which probably should be private.
But they're already used in the wild so no point in just hiding them
now.
The __init__ API is hidden -- the public API for this are the `from_*`
classmethods.
The prefixes make the API Reference docs (for e.g. `pytest.raises`,
`pytest.fixture`) uglier.
Being under `_pytest` is sufficient from a privacy perspective, so let's
drop them.
Sections have been moved to:
* reference/fixtures.rst
* how-to/fixtures.rst
* fixtures.rst
according to their function. Further refinement and rewriting will be required.
Some material has been moved to a new "Anatomy of a test" document, in
anticipation that material from other sections will also find a natural
home there later.
Removed several unneeded reference targets from fixtures documentation.
Moved various documents into subdirectories, how-to and
reference.
Updated multiple links to use `:ref:` instead of `:doc:`,
meaning that files can henceforth be moved around without
breaking references.
This configures Sphinx autodoc to include the type annotations
along with the description of the function/method, instead of including
it into the signature.
Fix#8405
* Fix test_strict_and_skip
The `--strict` argument was removed in #2552, but the removal wasn't
actually correct - see #1472.
* Fix argument handling in pytest.mark.skip
See #8384
* Raise from None
* Fix test name
* retry writing pytest-of dir when invalid chars are in directory name
* add unit tests for getbasetemp() and changelog
* patch _basetemp & _given_basetemp for testing basetemp()
* Tweak changelog for #8317, tidy up comments
Similarly to #7143, at work we have a project with a custom pytest.Class
subclass, adding an additional argument to the constructor.
All from_parent implementations in pytest accept and forward *kw, except
Class (before this change) and DoctestItem - since I'm not familiar with
doctest support, I've left the latter as-is.
doc This paragraph looks like it is a more verbose version of the sentence right above it. Removing it doesn't reduce the amount of information here but does make the section flow a little better.
In assert.rst at line 175 and further there is an example of an assert encountering comparisons. The code-block for this example starts with a comment (line 177) and then it has 2 empty lines. Comparing this example code (test_assert2.py) with the previously mentioned example code on the same page (i.e. test_assert1.py) you can see that there should not be 2 empty lines after the comment. These 2 empty lines are removed.
* Type annotation polishing for symbols around Pytester.run
Hopefully these will help document readers understand pertinent methods
and constants better.
Following up #8294
* Use NOTSET instead of object
Also update these tests ensure they pass, and be explicit about the test
file called in an existing test to avoid unintentional calls to the
added tests
SetupState maintains its own state, so it can store the exception
itself, instead of using the node's store, which is better avoided when
possible.
This also reduces the lifetime of the reference-cycle-inducing exception
objects which is never a bad thing.
The assertion ensures that when `addfinalizer(finalizer, node)` is
called, the node is in the stack. This then would ensure that the
finalization is actually properly executed properly during the node's
teardown. Anything else indicates something is wrong.
Previous commits fixed all of the tests which previously failed this, so
can be reenabeld now.
Previously the tests (probably unintentionally) mixed a fresh SetupState
and the generated item Session's SetupState, which led to some serious
head scratching when prodding it a bit.
When the stack is empty, the finalizers which are supposed to be
attached to nodes in the stack really ought to be empty as well. So the
code here is dead. If this doesn't happen, the assert will trigger.
It seems the code that would not install pytest's faulthandler support
if it was already enabled is not really needed at all, and even detrimental
when using `python -X dev -m pytest` to run Python in "dev" mode.
Also simplified the plugin by removing the hook class, now the hooks
will always be active so there's no need to delay the hook definitions anymore.
Fix#8258
The Twisted Logger will return an invalid file descriptor since it is
not backed by an FD. So, let's also forward this to the same code path
as with `pytest-xdist`.
as PyObjMixin is always supposed to be mixed in the mro
before nodes.Node the behavior doesn't change,
but all the typing information carry over to help mypy.
extracted from #8037
It is not very clear why this code exists -- we are not running any
unittest or nose code during collection, and really these frameworks
don't have the concept of collection at all, and just raising these
exceptions at e.g. the module level would cause an error. So unless I'm
missing something, I don't think anyone is using this.
Deprecate it so we can eventually clear up this code and keep unittest
more tightly restricted to its plugin.
When `pytest.skip()` is called inside a test function, the skip location
should be reported as the line that made the call, however when
`pytest.skip()` is called by the `pytest.mark.skip` and similar
mechanisms, the location should be reported at the item's location,
because the exact location is some irrelevant internal code.
Currently the item-location case is implemented by the caller setting a
boolean key on the item's store and the `skipping` plugin checking it
and fixing up the location if needed. This is really roundabout IMO and
breaks encapsulation.
Instead, allow the caller to specify directly on the skip exception
whether to use the item's location or not. For now, this is entirely
private.
When line_index was a large negative number, get_source failed
on `source.lines[line_index]`.
Use the same dummy Source as with a large positive line_index.
Pytest document currently does not index the top-level package name
`pytest`, which causes some trouble when building documentation that
cross-refers to the pytest package via ``:mod:`pytest` ``.
In https://docs.pytest.org/en/stable/reference.html#testdir, it is
suggested:
> New code should avoid using testdir in favor of pytester.
Multiple spots in the documents still use testdir and they can be quite
confusing (especially the plugin writing guide).
* [pre-commit.ci] pre-commit autoupdate
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
* manual fixes after configuration update
* [pre-commit.ci] auto fixes from pre-commit.com hooks
for more information, see https://pre-commit.ci
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Anthony Sottile <asottile@umich.edu>
The type cannot be constructed directly, but is exported for use in type
annotations, since it is reachable through existing public API.
This also documents `from_call` as public, because at least
pytest-forked uses it, so we must treat it as public already anyway.
The env var effects all of the pip installs, including regendoc which
also uses setuptools-scm, so it gets the wrong version, and fails to
install with the new pip resolver:
ERROR: Requested regendoc from 206e495142/regendoc-0.6.1.tar.gz (sha256)=db1e8c9ae02c1af559eae105bfd77ba41ed07fc8ca7030ea59db5f3f161236a4 has different version in metadata: '6.2.0'
Add a new hook , `pytest_markeval_namespace` which should return a dictionary.
This dictionary will be used to augment the "global" variables available to evaluate skipif/xfail/xpass markers.
Pseudo example
``conftest.py``:
.. code-block:: python
def pytest_markeval_namespace():
return {"color": "red"}
``test_func.py``:
.. code-block:: python
@pytest.mark.skipif("color == 'blue'", reason="Color is not red")
def test_func():
assert False
On Windows, os.path.samefile returns false for paths mounted in UNC paths which
point to the same location.
I couldn't reproduce the actual case reported, but looking at the code it seems
this commit should fix the issue.
Fix#7678Fix#8076
* Custom multiple marker execution order
https://github.com/pytest-dev/pytest/issues/8020 issue stated that ordering of multiple custom markers is from inside - out. I have added example for the same in the documentation. Please let me know for further changes / concerns.
* remove trailing spaces
The last commit was failing due to extra spaces
* Ran tox tests locally to debug white space trimming issues
* Resolve: ERROR: docs: commands failed for tox -e docs
* Update doc/en/reference.rst
Committed PR suggestions.
Co-authored-by: Florian Bruhin <me@the-compiler.org>
* Added reference to Node.iter_markers_with_node in documentation
* Add myself to Authors
Co-authored-by: Shubham <shubham.adep@wsu.edu>
Co-authored-by: Florian Bruhin <me@the-compiler.org>
Explanation: The default handling of these lifetimes is done in
`tmpdir.TempPathFactory.getbasetemp`, which passes `keep=3` to
`pathlib.make_numbered_dir_with_cleanup`.
GH Issue: #8036
* Prefix contextmanagers with module name in doc examples
* Import pytest explicitly for doctests
Co-authored-by: Bruno Oliveira <nicoddemus@gmail.com>
Python 3.7 changes the pyc format by adding a flags byte. Even though it
is not necessary for us to match it, it is nice to be able to read pyc
files we emit for debugging the rewriter.
Update our custom pyc files to use that format. We write flags==0
meaning we still use the mtime+size format rather the newer hash format.
In order to allow users to type annotate fixtures they request, the
types need to be imported from the `pytest` namespace. They are/were
always available to import from the `_pytest` namespace, but that is
not guaranteed to be stable.
These types are only exported for the purpose of typing. Specifically,
the following are *not* public:
- Construction (`__init__`)
- Subclassing
- staticmethods and classmethods
We try to combat them being used anyway by:
- Marking the classes as `@final` when possible (already done).
- Not documenting private stuff in the API Reference.
- Using `_`-prefixed names or marking as `:meta private:` for private
stuff.
- Adding a keyword-only `_ispytest=False` to private constructors,
warning if False, and changing pytest itself to pass True. In the
future it will (hopefully) become a hard error.
Hopefully that will be enough.
An upcoming commit wants to import from `_pytest.pytester` in the public
`pytest` module. This means that `_pytest.pytester` would start to get
imported during import time, which it hasn't up to now -- it was
imported by the plugin loader (if requested). When a plugin is loaded,
it is subjected to assertion rewriting, but only if the module isn't
imported yet, it issues a warning "Module already imported so cannot be
rewritten" and skips the rewriting. So we'd end up with the pytester
plugin not being rewritten, but it wants to be.
Absent better ideas, the solution here is to split the pytester
assertions to their own plugin (which will always only be imported by
the plugin loader) and exclude pytester itself from plugin rewriting.
When --doctest-modules is used, an `__init__.py` file is not a `Package`
but a `DoctestModule`, but some collection code assumed that
`__init__.py` implies a `Package`. That code caused only a single test
to be collected in the scenario in the subject.
Tighten up this check to explicitly check for `Package`. There are
better solutions, but for another time.
Report & test by Nick Gates <nickgatzgates@gmail.com>.
We want to export `pytest.MonkeyPatch` for the purpose of
type-annotating the `monkeypatch` fixture. For other fixtures we export
in this way, we also make direct construction of them (e.g.
`MonkeyPatch()`) private. But unlike the others, `MonkeyPatch` is also
widely used directly already, mostly because the `monkeypatch` fixture
only works in `function` scope (issue #363), but also in other cases. So
making it private will be annoying and we don't offer a decent
replacement yet.
So, let's just make direct construction public & documented.
Don't import `pytest` from within some `_pytest` modules since an
upcoming commit will import from them into `pytest`.
It would have been nice not to have to do it, so that internal plugins
look more like external plugins, but with the existing layout this seems
unavoidable.
When pytest was run on a directory containing a recursive symlink it failed
with ELOOP as the library was not able to determine the type of the
direntry:
src/_pytest/main.py:685: in collect
if not direntry.is_file():
E OSError: [Errno 40] Too many levels of symbolic links: '/home/florian/proj/pytest/tests/recursive'
This is fixed by handling ELOOP and other errors in the visit function in
pathlib.py, so the entries whose is_file() call raises an OSError with the
pre-defined list of error numbers will be exluded from the result.
The _ignore_errors function was copied from Lib/pathlib.py of cpython 3.9.
Fixes#7951
* adding --sw-skip shorthand for stepwise skip
* be explicit rather than implicit with default args for stepwise
* add constant for sw cache dir; only register plugin if necessary rather check check activity always;
* use str format; remove unused args in hooks
* assert cache upfront, allow stepwise to have a reference to the cache
* type hinting lf, skip, move literal strings into module constants
* convert parametrized option into a list
* add a sessionfinish hook for stepwise to keep backwards behaviour the same
* add changelog for #7938
* Improve performance of stepwise modifyitems & address PR feedback
* add test for stepwise deselected based on performance enhancements
* Apply suggestions from code review
* delete from items, account for edge case where failed_index = 0
Co-authored-by: Bruno Oliveira <nicoddemus@gmail.com>
We run mypy through pre-commit, and we don't keep duplicate targets in
tox for all of the other linters. Since this adds some (small)
maintenance overhead, remove it.
In pytester tests, pytest stashes & restores the sys.modules for each
test. So if the test imports a new module, it is initialized anew each
time.
Turns out the readline module isn't multi-init safe, which causes
pytester.spawn to crash or hang. So preserve it as a workaround.
It turns out all autouse fixtures are kept in a global list, and thinned
out for a particular node using a linear scan of the entire list each
time.
Change the list to a dict, and only take the nodes we need.
ischildnode can be quite hot in some cases involving many fixtures.
However it is always used in a way that the nodeid is constant and the
baseid is iterated. So we can save work by pre-computing the parents of
the nodeid and use a simple containment test.
The `_getautousenames` function has the same stuff open-coded, so change
it to use the new function as well.
Since commit 0f918b1a9d pytest uses auto-generated autouse
pytest fixtures for the xunit fixtures
{setup,teardown}_{module,class,method,function}. All of these fixtures
were given the same name.
Unfortunately, pytest fixture lookup for a name works by grabbing all of
the fixtures globally declared with a name and filtering to only those
which match the specific node. So each xunit-using item iterates over a
list (of fixturedefs) of a size of all previous same-xunit-using items,
i.e. quadratic.
Fixing this properly to use a better data structure is likely to take
some effort, but we can avoid the immediate problem by just using
a different name for each item's autouse fixture, so it only matches
itself.
A benchmark is added to demonstrate the issue. It is still way too slow
after the fix and possibly still quadratic, but for a different reason
which is another matter.
Running --collect-only, before (snipped):
202533232 function calls (201902604 primitive calls) in 86.379 seconds
ncalls tottime percall cumtime percall filename:lineno(function)
1 0.000 0.000 85.688 85.688 main.py:320(pytest_collection)
1 0.000 0.000 85.688 85.688 main.py:567(perform_collect)
80557/556 0.021 0.000 85.050 0.153 {method 'extend' of 'list' objects}
85001/15001 0.166 0.000 85.045 0.006 main.py:785(genitems)
10002 0.050 0.000 84.219 0.008 runner.py:455(collect_one_node)
10002 0.049 0.000 83.763 0.008 runner.py:340(pytest_make_collect_report)
10002 0.079 0.000 83.668 0.008 runner.py:298(from_call)
10002 0.019 0.000 83.541 0.008 runner.py:341(<lambda>)
5001 0.184 0.000 81.922 0.016 python.py:412(collect)
5000 0.020 0.000 81.072 0.016 python.py:842(collect)
30003 0.118 0.000 78.478 0.003 python.py:218(pytest_pycollect_makeitem)
30000 0.190 0.000 77.957 0.003 python.py:450(_genfunctions)
40001 0.081 0.000 76.664 0.002 nodes.py:183(from_parent)
30000 0.087 0.000 76.629 0.003 python.py:1595(from_parent)
40002 0.092 0.000 76.583 0.002 nodes.py:102(_create)
30000 0.305 0.000 76.404 0.003 python.py:1533(__init__)
15000 0.132 0.000 74.765 0.005 fixtures.py:1439(getfixtureinfo)
15000 0.165 0.000 73.664 0.005 fixtures.py:1492(getfixtureclosure)
15000 0.044 0.000 57.584 0.004 fixtures.py:1653(getfixturedefs)
30000 18.840 0.001 57.540 0.002 fixtures.py:1668(_matchfactories)
37507500 31.352 0.000 38.700 0.000 nodes.py:76(ischildnode)
15000 10.464 0.001 15.806 0.001 fixtures.py:1479(_getautousenames)
112930587/112910019 7.333 0.000 7.339 0.000 {built-in method builtins.len}
After:
51890333 function calls (51259706 primitive calls) in 27.306 seconds
ncalls tottime percall cumtime percall filename:lineno(function)
1 0.000 0.000 26.783 26.783 main.py:320(pytest_collection)
1 0.000 0.000 26.783 26.783 main.py:567(perform_collect)
80557/556 0.020 0.000 26.108 0.047 {method 'extend' of 'list' objects}
85001/15001 0.151 0.000 26.103 0.002 main.py:785(genitems)
10002 0.047 0.000 25.324 0.003 runner.py:455(collect_one_node)
10002 0.045 0.000 24.888 0.002 runner.py:340(pytest_make_collect_report)
10002 0.069 0.000 24.805 0.002 runner.py:298(from_call)
10002 0.017 0.000 24.690 0.002 runner.py:341(<lambda>)
5001 0.168 0.000 23.150 0.005 python.py:412(collect)
5000 0.019 0.000 22.223 0.004 python.py:858(collect)
30003 0.101 0.000 19.818 0.001 python.py:218(pytest_pycollect_makeitem)
30000 0.161 0.000 19.368 0.001 python.py:450(_genfunctions)
30000 0.302 0.000 18.236 0.001 python.py:1611(from_parent)
40001 0.084 0.000 18.051 0.000 nodes.py:183(from_parent)
40002 0.116 0.000 17.967 0.000 nodes.py:102(_create)
30000 0.308 0.000 17.770 0.001 python.py:1549(__init__)
15000 0.117 0.000 16.111 0.001 fixtures.py:1439(getfixtureinfo)
15000 0.134 0.000 15.135 0.001 fixtures.py:1492(getfixtureclosure)
15000 9.320 0.001 14.738 0.001 fixtures.py:1479(_getautousenames)
--lf has an optimization where it skips collecting Modules (python
files) which don't contain failing tests. The optimization works by
getting the paths of all cached failed tests and skipping the collection
of Modules whose path is not included in that list.
In pytest, Package nodes are Module nodes with the fspath being the file
`<package dir>/__init__.py`. Since it's a Module the logic above
triggered for it, and because it's an `__init__.py` file which is
unlikely to have any failing tests in it, it is skipped, which causes
its entire directory to be skipped, including any Modules inside it with
failing tests.
Fix by special-casing Packages to never filter. This means entire
Packages are never filtered, the Modules themselves are always checked.
It is reasonable to consider an optimization which does filter entire
packages bases on parent paths etc. but this wouldn't actually save any
real work so is really not worth it.
Regressed in 6.1.0 in 62e249a1f9.
The `x` is an `str` but is expected to be a `pathlib.Path`. Not caught
by mypy because `config.getini()` returns `Any`.
Fix by just removing the `bestrelpath` call:
- testpaths are always relative to the rootdir, it thus would be very
unusual to specify an absolute path there.
- The code was wrong even before the regression: `py.path.local`'s
`bestrelpath` function expects a `py.path.local`, not an `str`. But it
had some weird `try ... except AttributeError` fallback which just
returns the argument, i.e. it was a no-op. So there is no behavior
change.
- It seems reasonable to me to just print the full path if that's what
the ini specifies.
When switching from py.path.local to pathlib (70f3ad1c1f),
`local.parts(reverse=True)` was translated incorrectly, leading to the
wrong rootdir being determined in some non-trivial cases where parent
directories have config files as well.
<!-- What are you trying to do, that you are unable to achieve with pytest as it currently stands? -->
#### Describe the solution you'd like
<!-- A clear and concise description of what you want to happen. -->
<!-- Provide examples of real-world use cases that this would enable and how it solves the problem described above. -->
#### Alternative Solutions
<!-- Have you tried to workaround the problem using a pytest plugin or other tools? Or a different approach to solving this issue? Please elaborate here. -->
#### Additional context
<!-- Add any other context, links, etc. about the feature here. -->
@@ -13,7 +13,7 @@ If this change fixes an issue, please:
Unless your change is trivial or a small documentation fix (e.g., a typo or reword of a small section) please:
- [ ] Create a new changelog file in the `changelog` folder, with a name like `<ISSUE NUMBER>.<TYPE>.rst`. See [changelog/README.rst](https://github.com/pytest-dev/pytest/blob/master/changelog/README.rst) for details.
- [ ] Create a new changelog file in the `changelog` folder, with a name like `<ISSUE NUMBER>.<TYPE>.rst`. See [changelog/README.rst](https://github.com/pytest-dev/pytest/blob/main/changelog/README.rst) for details.
Write sentences in the **past or present tense**, examples:
@@ -313,32 +307,37 @@ Here is a simple overview, with pytest-specific bits:
#. Add yourself to ``AUTHORS`` file if not there yet, in alphabetical order.
#. Commit and push once your tests pass and you are happy with your change(s)::
$ git commit -a -m "<commit message>"
$ git push -u
#. Finally, submit a pull request through the GitHub website using this data::
head-fork: YOUR_GITHUB_USERNAME/pytest
compare: your-branch-name
base-fork: pytest-dev/pytest
base: master
base: main
Writing Tests
~~~~~~~~~~~~~
Writing tests for plugins or for pytest itself is often done using the `testdir fixture <https://docs.pytest.org/en/stable/reference.html#testdir>`_, as a "black-box" test.
Writing tests for plugins or for pytest itself is often done using the `pytester fixture <https://docs.pytest.org/en/stable/reference.html#pytester>`_, as a "black-box" test.
For example, to ensure a simple test passes you can write:
..code-block::python
deftest_true_assertion(testdir):
testdir.makepyfile(
deftest_true_assertion(pytester):
pytester.makepyfile(
"""
def test_foo():
assert True
"""
)
result=testdir.runpytest()
result=pytester.runpytest()
result.assert_outcomes(failed=0,passed=1)
@@ -347,14 +346,14 @@ Alternatively, it is possible to make checks based on the actual output of the t
When choosing a file where to write a new test, take a look at the existing files and see if there's
@@ -388,15 +387,15 @@ Suppose for example that the latest release was 1.2.3, and you want to include
a bug fix in 1.2.4 (check https://github.com/pytest-dev/pytest/releases for the
actual latest release). The procedure for this is:
#. First, make sure the bug is fixed the ``master`` branch, with a regular pull
#. First, make sure the bug is fixed the ``main`` branch, with a regular pull
request, as described above. An exception to this is if the bug fix is not
applicable to ``master`` anymore.
applicable to ``main`` anymore.
#.``git checkout origin/1.2.x -b backport-XXXX`` # use the master PR number here
#.``git checkout origin/1.2.x -b backport-XXXX`` # use the main PR number here
#. Locate the merge commit on the PR, in the *merged* message, for example:
nicoddemus merged commit 0f8b462 into pytest-dev:master
nicoddemus merged commit 0f8b462 into pytest-dev:main
#.``git cherry-pick -x -m1 REVISION`` # use the revision you found above (``0f8b462``).
@@ -409,8 +408,8 @@ actual latest release). The procedure for this is:
Who does the backporting
~~~~~~~~~~~~~~~~~~~~~~~~
As mentioned above, bugs should first be fixed on ``master`` (except in rare occasions
that a bug only happens in a previous release). So who should do the backport procedure described
As mentioned above, bugs should first be fixed on ``main`` (except in rare occasions
that a bug only happens in a previous release). So, who should do the backport procedure described
above?
1. If the bug was fixed by a core developer, it is the main responsibility of that core developer
@@ -418,8 +417,8 @@ above?
2. However, often the merge is done by another maintainer, in which case it is nice of them to
do the backport procedure if they have the time.
3. For bugs submitted by non-maintainers, it is expected that a core developer will to do
the backport, normally the one that merged the PR on ``master``.
4. If a non-maintainers notices a bug which is fixed on ``master`` but has not been backported
the backport, normally the one that merged the PR on ``main``.
4. If a non-maintainers notices a bug which is fixed on ``main`` but has not been backported
(due to maintainers forgetting to apply the *needs backport* label, or just plain missing it),
they are also welcome to open a PR with the backport. The procedure is simple and really
helps with the maintenance of the project.
@@ -448,7 +447,7 @@ can always reopen the issue/pull request in their own time later if it makes sen
When to close
~~~~~~~~~~~~~
Here are a few general rules the maintainers use to decide when to close issues/PRs because
Here are a few general rules the maintainers use deciding when to close issues/PRs because
of lack of inactivity:
* Issues labeled ``question`` or ``needs information``: closed after 14 days inactive.
@@ -460,15 +459,15 @@ The above are **not hard rules**, but merely **guidelines**, and can be (and oft
Closing pull requests
~~~~~~~~~~~~~~~~~~~~~
When closing a Pull Request, it needs to be acknowledge the time, effort, and interest demonstrated by the person which submitted it. As mentioned previously, it is not the intent of the team to dismiss stalled pull request entirely but to merely to clear up our queue, so a message like the one below is warranted when closing a pull request that went stale:
When closing a Pull Request, it needs to be acknowledging the time, effort, and interest demonstrated by the person which submitted it. As mentioned previously, it is not the intent of the team to dismiss a stalled pull request entirely but to merely to clear up our queue, so a message like the one below is warranted when closing a pull request that went stale:
Hi <contributor>,
First of all we would like to thank you for your time and effort on working on this, the pytest team deeply appreciates it.
First of all, we would like to thank you for your time and effort on working on this, the pytest team deeply appreciates it.
We noticed it has been awhile since you have updated this PR, however. pytest is a high activity project, with many issues/PRs being opened daily, so it is hard for us maintainers to track which PRs are ready for merging, for review, or need more attention.
So for those reasons we think it is best to close the PR for now, but with the only intention to cleanup our queue, it is by no means a rejection of your changes. We still encourage you to re-open this PR (it is just a click of a button away) when you are ready to get back to it.
So for those reasons we, think it is best to close the PR for now, but with the only intention to cleanup our queue, it is by no means a rejection of your changes. We still encourage you to re-open this PR (it is just a click of a button away) when you are ready to get back to it.
Again we appreciate your time for working on this, and hope you might get back to this at a later time!
As per our policy, the following features which have been deprecated in the 5.X series are now
removed:
* The ``funcargnames`` read-only property of ``FixtureRequest``, ``Metafunc``, and ``Function`` classes. Use ``fixturenames`` attribute.
*``@pytest.fixture`` no longer supports positional arguments, pass all arguments by keyword instead.
* Direct construction of ``Node`` subclasses now raise an error, use ``from_parent`` instead.
* The default value for ``junit_family`` has changed to ``xunit2``. If you require the old format, add ``junit_family=xunit1`` to your configuration file.
* The ``TerminalReporter`` no longer has a ``writer`` attribute. Plugin authors may use the public functions of the ``TerminalReporter`` instead of accessing the ``TerminalWriter`` object directly.
* The ``--result-log`` option has been removed. Users are recommended to use the `pytest-reportlog <https://github.com/pytest-dev/pytest-reportlog>`__ plugin instead.
For more information consult
`Deprecations and Removals <https://docs.pytest.org/en/stable/deprecations.html>`__ in the docs.
Internal pytest warnings issued during the early stages of initialization are now properly handled and can filtered through :confval:`filterwarnings` or ``--pythonwarnings/-W``.
This also fixes a number of long standing issues: `#2891 <https://github.com/pytest-dev/pytest/issues/2891>`__, `#7620 <https://github.com/pytest-dev/pytest/issues/7620>`__, `#7426 <https://github.com/pytest-dev/pytest/issues/7426>`__.
The types of objects used in pytest's API are now exported so they may be used in type annotations.
The newly-exported types are:
-``pytest.Config`` for :class:`Config <pytest.Config>`.
-``pytest.Mark`` for :class:`marks <pytest.Mark>`.
-``pytest.MarkDecorator`` for :class:`mark decorators <pytest.MarkDecorator>`.
-``pytest.MarkGenerator`` for the :class:`pytest.mark <pytest.MarkGenerator>` singleton.
-``pytest.Metafunc`` for the :class:`metafunc <pytest.MarkGenerator>` argument to the :func:`pytest_generate_tests <pytest.hookspec.pytest_generate_tests>` hook.
-``pytest.CallInfo`` for the :class:`CallInfo <pytest.CallInfo>` type passed to various hooks.
-``pytest.PytestPluginManager`` for :class:`PytestPluginManager <pytest.PytestPluginManager>`.
-``pytest.ExceptionInfo`` for the :class:`ExceptionInfo <pytest.ExceptionInfo>` type returned from :func:`pytest.raises` and passed to various hooks.
-``pytest.Parser`` for the :class:`Parser <pytest.Parser>` type passed to the :func:`pytest_addoption <pytest.hookspec.pytest_addoption>` hook.
-``pytest.OptionGroup`` for the :class:`OptionGroup <pytest.OptionGroup>` type returned from the :func:`parser.addgroup <pytest.Parser.getgroup>` method.
Constructing them directly is not supported; they are only meant for use in type annotations.
Doing so will emit a deprecation warning, and may become a hard-error in pytest 7.0.
Subclassing them is also not supported. This is not currently enforced at runtime, but is detected by type-checkers such as mypy.
When a plugin listed in ``required_plugins`` is missing or an unknown config key is used with ``--strict-config``, a simple error message is now shown instead of a stacktrace.
Fixed test collection when a full path without a drive letter was passed to pytest on Windows (for example ``\projects\tests\test.py`` instead of ``c:\projects\tests\pytest.py``).
Fix handling of command-line options that appear as paths but trigger an OS-level syntax error on Windows, such as the options used internally by ``pytest-xdist``.
New ``--durations-min`` command-line flag controls the minimal duration for inclusion in the slowest list of tests shown by ``--durations``. Previously this was hard-coded to ``0.005s``.
Added two new attributes :attr:`rootpath <_pytest.config.Config.rootpath>` and :attr:`inipath <_pytest.config.Config.inipath>` to :class:`Config <_pytest.config.Config>`.
These attributes are :class:`pathlib.Path` versions of the existing :attr:`rootdir <_pytest.config.Config.rootdir>` and :attr:`inifile <_pytest.config.Config.inifile>` attributes,
The ``@pytest.mark.skip`` decorator now correctly handles its arguments. When the ``reason`` argument is accidentally given both positional and as a keyword (e.g. because it was confused with ``skipif``), a ``TypeError`` now occurs. Before, such tests were silently skipped, and the positional argument ignored. Additionally, ``reason`` is now documented correctly as positional or keyword (rather than keyword-only).
Use private names for internal fixtures that handle classic setup/teardown so that they don't show up with the default ``--fixtures`` invocation (but they still show up with ``--fixtures -v``).
The :confval:`required_plugins` config option now works correctly when pre-releases of plugins are installed, rather than falsely claiming that those plugins aren't installed at all.
Fixed issue where `TestCase.setUpClass` is not called when a test has `/` in its name since pytest 6.2.0.
This refers to the path part in pytest node IDs, e.g. `TestClass::test_it` in the node ID `tests/test_file.py::TestClass::test_it`.
Now, instead of assuming that the test name does not contain ``/``, it is assumed that test path does not contain ``::``. We plan to hopefully make both of these work in the future.
New :ref:`version-tuple` attribute, which makes it simpler for users to do something depending on the pytest version (such as declaring hooks which are introduced in later versions).
@@ -10,7 +10,7 @@ we keep learning about new and better structures to express different details ab
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 multipe types of backward compatibility transitions:
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.
@@ -25,7 +25,7 @@ b) transitional: the old and new API don't conflict
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.
c) true breakage: should only to be considered when normal transition is unreasonably unsustainable and would offset important development/features by years.
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.
Examples for such upcoming changes:
@@ -42,7 +42,7 @@ c) true breakage: should only to be considered when normal transition is unreaso
After there's no hard *-1* on the issue it should be followed up by an initial proof-of-concept Pull Request.
This POC serves as both a coordination point to assess impact and potential inspriation to come up with a transitional solution after all.
This POC serves as both a coordination point to assess impact and potential inspiration to come up with a transitional solution after all.
After a reasonable amount of time the PR can be merged to base a new major release.
@@ -28,6 +28,455 @@ with advance notice in the **Deprecations** section of releases.
.. towncrier release notes start
pytest 6.2.4 (2021-05-04)
=========================
Bug Fixes
---------
-`#8539 <https://github.com/pytest-dev/pytest/issues/8539>`_: Fixed assertion rewriting on Python 3.10.
pytest 6.2.3 (2021-04-03)
=========================
Bug Fixes
---------
-`#8414 <https://github.com/pytest-dev/pytest/issues/8414>`_: pytest used to create directories under ``/tmp`` with world-readable
permissions. This means that any user in the system was able to read
information written by tests in temporary directories (such as those created by
the ``tmp_path``/``tmpdir`` fixture). Now the directories are created with
private permissions.
pytest used to silenty use a pre-existing ``/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.
pytest 6.2.2 (2021-01-25)
=========================
Bug Fixes
---------
-`#8152 <https://github.com/pytest-dev/pytest/issues/8152>`_: Fixed "(<Skipped instance>)" being shown as a skip reason in the verbose test summary line when the reason is empty.
-`#8249 <https://github.com/pytest-dev/pytest/issues/8249>`_: Fix the ``faulthandler`` plugin for occasions when running with ``twisted.logger`` and using ``pytest --capture=no``.
pytest 6.2.1 (2020-12-15)
=========================
Bug Fixes
---------
-`#7678 <https://github.com/pytest-dev/pytest/issues/7678>`_: Fixed bug where ``ImportPathMismatchError`` would be raised for files compiled in
the host and loaded later from an UNC mounted path (Windows).
-`#8132 <https://github.com/pytest-dev/pytest/issues/8132>`_: Fixed regression in ``approx``: in 6.2.0 ``approx`` no longer raises
``TypeError`` when dealing with non-numeric types, falling back to normal comparison.
Before 6.2.0, array types like tf.DeviceArray fell through to the scalar case,
and happened to compare correctly to a scalar if they had only one element.
After 6.2.0, these types began failing, because they inherited neither from
standard Python number hierarchy nor from ``numpy.ndarray``.
``approx`` now converts arguments to ``numpy.ndarray`` if they expose the array
protocol and are not scalars. This treats array-like objects like numpy arrays,
regardless of size.
pytest 6.2.0 (2020-12-12)
=========================
Breaking Changes
----------------
-`#7808 <https://github.com/pytest-dev/pytest/issues/7808>`_: pytest now supports python3.6+ only.
Deprecations
------------
-`#7469 <https://github.com/pytest-dev/pytest/issues/7469>`_: Directly constructing/calling the following classes/functions is now deprecated:
These have always been considered private, but now issue a deprecation warning, which may become a hard error in pytest 7.0.0.
-`#7530 <https://github.com/pytest-dev/pytest/issues/7530>`_: The ``--strict`` command-line option has been deprecated, use ``--strict-markers`` instead.
We have plans to maybe in the future to reintroduce ``--strict`` and make it an encompassing flag for all strictness
related options (``--strict-markers`` and ``--strict-config`` at the moment, more might be introduced in the future).
-`#7988 <https://github.com/pytest-dev/pytest/issues/7988>`_: The ``@pytest.yield_fixture`` decorator/function is now deprecated. Use :func:`pytest.fixture` instead.
``yield_fixture`` has been an alias for ``fixture`` for a very long time, so can be search/replaced safely.
Features
--------
-`#5299 <https://github.com/pytest-dev/pytest/issues/5299>`_: pytest now warns about unraisable exceptions and unhandled thread exceptions that occur in tests on Python>=3.8.
See :ref:`unraisable` for more information.
-`#7425 <https://github.com/pytest-dev/pytest/issues/7425>`_: New :fixture:`pytester` fixture, which is identical to :fixture:`testdir` but its methods return :class:`pathlib.Path` when appropriate instead of ``py.path.local``.
This is part of the movement to use :class:`pathlib.Path` objects internally, in order to remove the dependency to ``py`` in the future.
Internally, the old :class:`Testdir <_pytest.pytester.Testdir>` is now a thin wrapper around :class:`Pytester <_pytest.pytester.Pytester>`, preserving the old interface.
-`#7695 <https://github.com/pytest-dev/pytest/issues/7695>`_: A new hook was added, `pytest_markeval_namespace` which should return a dictionary.
This dictionary will be used to augment the "global" variables available to evaluate skipif/xfail/xpass markers.
Pseudo example
``conftest.py``:
..code-block::python
defpytest_markeval_namespace():
return{"color":"red"}
``test_func.py``:
..code-block::python
@pytest.mark.skipif("color == 'blue'",reason="Color is not red")
deftest_func():
assertFalse
-`#8006 <https://github.com/pytest-dev/pytest/issues/8006>`_: It is now possible to construct a :class:`~pytest.MonkeyPatch` object directly as ``pytest.MonkeyPatch()``,
in cases when the :fixture:`monkeypatch` fixture cannot be used. Previously some users imported it
from the private `_pytest.monkeypatch.MonkeyPatch` namespace.
Additionally, :meth:`MonkeyPatch.context <pytest.MonkeyPatch.context>` is now a classmethod,
and can be used as ``with MonkeyPatch.context() as mp: ...``. This is the recommended way to use
``MonkeyPatch`` directly, since unlike the ``monkeypatch`` fixture, an instance created directly
is not ``undo()``-ed automatically.
Improvements
------------
-`#1265 <https://github.com/pytest-dev/pytest/issues/1265>`_: Added an ``__str__`` implementation to the :class:`~pytest.pytester.LineMatcher` class which is returned from ``pytester.run_pytest().stdout`` and similar. It returns the entire output, like the existing ``str()`` method.
-`#2044 <https://github.com/pytest-dev/pytest/issues/2044>`_: Verbose mode now shows the reason that a test was skipped in the test's terminal line after the "SKIPPED", "XFAIL" or "XPASS".
-`#7469 <https://github.com/pytest-dev/pytest/issues/7469>`_ The types of builtin pytest fixtures are now exported so they may be used in type annotations of test functions.
The newly-exported types are:
-``pytest.FixtureRequest`` for the :fixture:`request` fixture.
-``pytest.Cache`` for the :fixture:`cache` fixture.
-``pytest.CaptureFixture[str]`` for the :fixture:`capfd` and :fixture:`capsys` fixtures.
-``pytest.CaptureFixture[bytes]`` for the :fixture:`capfdbinary` and :fixture:`capsysbinary` fixtures.
-``pytest.LogCaptureFixture`` for the :fixture:`caplog` fixture.
-``pytest.Pytester`` for the :fixture:`pytester` fixture.
-``pytest.Testdir`` for the :fixture:`testdir` fixture.
-``pytest.TempdirFactory`` for the :fixture:`tmpdir_factory` fixture.
-``pytest.TempPathFactory`` for the :fixture:`tmp_path_factory` fixture.
-``pytest.MonkeyPatch`` for the :fixture:`monkeypatch` fixture.
-``pytest.WarningsRecorder`` for the :fixture:`recwarn` fixture.
Constructing them is not supported (except for `MonkeyPatch`); they are only meant for use in type annotations.
Doing so will emit a deprecation warning, and may become a hard-error in pytest 7.0.
Subclassing them is also not supported. This is not currently enforced at runtime, but is detected by type-checkers such as mypy.
-`#7527 <https://github.com/pytest-dev/pytest/issues/7527>`_: When a comparison between :func:`namedtuple <collections.namedtuple>` instances of the same type fails, pytest now shows the differing field names (possibly nested) instead of their indexes.
-`#7615 <https://github.com/pytest-dev/pytest/issues/7615>`_: :meth:`Node.warn <_pytest.nodes.Node.warn>` now permits any subclass of :class:`Warning`, not just :class:`PytestWarning <pytest.PytestWarning>`.
-`#7701 <https://github.com/pytest-dev/pytest/issues/7701>`_: Improved reporting when using ``--collected-only``. It will now show the number of collected tests in the summary stats.
-`#7710 <https://github.com/pytest-dev/pytest/issues/7710>`_: Use strict equality comparison for non-numeric types in :func:`pytest.approx` instead of
raising :class:`TypeError`.
This was the undocumented behavior before 3.7, but is now officially a supported feature.
-`#7938 <https://github.com/pytest-dev/pytest/issues/7938>`_: New ``--sw-skip`` argument which is a shorthand for ``--stepwise-skip``.
-`#8023 <https://github.com/pytest-dev/pytest/issues/8023>`_: Added ``'node_modules'`` to default value for :confval:`norecursedirs`.
-`#8032 <https://github.com/pytest-dev/pytest/issues/8032>`_: :meth:`doClassCleanups <unittest.TestCase.doClassCleanups>` (introduced in :mod:`unittest` in Python and 3.8) is now called appropriately.
Bug Fixes
---------
-`#4824 <https://github.com/pytest-dev/pytest/issues/4824>`_: Fixed quadratic behavior and improved performance of collection of items using autouse fixtures and xunit fixtures.
-`#7758 <https://github.com/pytest-dev/pytest/issues/7758>`_: Fixed an issue where some files in packages are getting lost from ``--lf`` even though they contain tests that failed. Regressed in pytest 5.4.0.
-`#7911 <https://github.com/pytest-dev/pytest/issues/7911>`_: Directories created by by :fixture:`tmp_path` and :fixture:`tmpdir` are now considered stale after 3 days without modification (previous value was 3 hours) to avoid deleting directories still in use in long running test suites.
-`#7913 <https://github.com/pytest-dev/pytest/issues/7913>`_: Fixed a crash or hang in :meth:`pytester.spawn <_pytest.pytester.Pytester.spawn>` when the :mod:`readline` module is involved.
-`#7951 <https://github.com/pytest-dev/pytest/issues/7951>`_: Fixed handling of recursive symlinks when collecting tests.
-`#7981 <https://github.com/pytest-dev/pytest/issues/7981>`_: Fixed symlinked directories not being followed during collection. Regressed in pytest 6.1.0.
-`#8016 <https://github.com/pytest-dev/pytest/issues/8016>`_: Fixed only one doctest being collected when using ``pytest --doctest-modules path/to/an/__init__.py``.
Improved Documentation
----------------------
-`#7429 <https://github.com/pytest-dev/pytest/issues/7429>`_: Add more information and use cases about skipping doctests.
-`#7780 <https://github.com/pytest-dev/pytest/issues/7780>`_: Classes which should not be inherited from are now marked ``final class`` in the API reference.
-`#7872 <https://github.com/pytest-dev/pytest/issues/7872>`_: ``_pytest.config.argparsing.Parser.addini()`` accepts explicit ``None`` and ``"string"``.
-`#7878 <https://github.com/pytest-dev/pytest/issues/7878>`_: In pull request section, ask to commit after editing changelog and authors file.
Trivial/Internal Changes
------------------------
-`#7802 <https://github.com/pytest-dev/pytest/issues/7802>`_: The ``attrs`` dependency requirement is now >=19.2.0 instead of >=17.4.0.
-`#8014 <https://github.com/pytest-dev/pytest/issues/8014>`_: `.pyc` files created by pytest's assertion rewriting now conform to the newer PEP-552 format on Python>=3.7.
(These files are internal and only interpreted by pytest itself.)
pytest 6.1.2 (2020-10-28)
=========================
Bug Fixes
---------
-`#7758 <https://github.com/pytest-dev/pytest/issues/7758>`_: Fixed an issue where some files in packages are getting lost from ``--lf`` even though they contain tests that failed. Regressed in pytest 5.4.0.
-`#7911 <https://github.com/pytest-dev/pytest/issues/7911>`_: Directories created by `tmpdir` are now considered stale after 3 days without modification (previous value was 3 hours) to avoid deleting directories still in use in long running test suites.
Improved Documentation
----------------------
-`#7815 <https://github.com/pytest-dev/pytest/issues/7815>`_: Improve deprecation warning message for ``pytest._fillfuncargs()``.
pytest 6.1.1 (2020-10-03)
=========================
Bug Fixes
---------
-`#7807 <https://github.com/pytest-dev/pytest/issues/7807>`_: Fixed regression in pytest 6.1.0 causing incorrect rootdir to be determined in some non-trivial cases where parent directories have config files as well.
-`#7814 <https://github.com/pytest-dev/pytest/issues/7814>`_: Fixed crash in header reporting when :confval:`testpaths` is used and contains absolute paths (regression in 6.1.0).
pytest 6.1.0 (2020-09-26)
=========================
Breaking Changes
----------------
-`#5585 <https://github.com/pytest-dev/pytest/issues/5585>`_: As per our policy, the following features which have been deprecated in the 5.X series are now
removed:
* The ``funcargnames`` read-only property of ``FixtureRequest``, ``Metafunc``, and ``Function`` classes. Use ``fixturenames`` attribute.
*``@pytest.fixture`` no longer supports positional arguments, pass all arguments by keyword instead.
* Direct construction of ``Node`` subclasses now raise an error, use ``from_parent`` instead.
* The default value for ``junit_family`` has changed to ``xunit2``. If you require the old format, add ``junit_family=xunit1`` to your configuration file.
* The ``TerminalReporter`` no longer has a ``writer`` attribute. Plugin authors may use the public functions of the ``TerminalReporter`` instead of accessing the ``TerminalWriter`` object directly.
* The ``--result-log`` option has been removed. Users are recommended to use the `pytest-reportlog <https://github.com/pytest-dev/pytest-reportlog>`__ plugin instead.
For more information consult
`Deprecations and Removals <https://docs.pytest.org/en/stable/deprecations.html>`__ in the docs.
Deprecations
------------
-`#6981 <https://github.com/pytest-dev/pytest/issues/6981>`_: The ``pytest.collect`` module is deprecated: all its names can be imported from ``pytest`` directly.
-`#7097 <https://github.com/pytest-dev/pytest/issues/7097>`_: The ``pytest._fillfuncargs`` function is deprecated. This function was kept
for backward compatibility with an older plugin.
It's functionality is not meant to be used directly, but if you must replace
it, use `function._request._fillfixtures()` instead, though note this is not
a public API and may break in the future.
-`#7210 <https://github.com/pytest-dev/pytest/issues/7210>`_: The special ``-k '-expr'`` syntax to ``-k`` is deprecated. Use ``-k 'not expr'``
instead.
The special ``-k 'expr:'`` syntax to ``-k`` is deprecated. Please open an issue
if you use this and want a replacement.
-`#7255 <https://github.com/pytest-dev/pytest/issues/7255>`_: The :func:`pytest_warning_captured <_pytest.hookspec.pytest_warning_captured>` hook is deprecated in favor
of :func:`pytest_warning_recorded <_pytest.hookspec.pytest_warning_recorded>`, and will be removed in a future version.
-`#7648 <https://github.com/pytest-dev/pytest/issues/7648>`_: The ``gethookproxy()`` and ``isinitpath()`` methods of ``FSCollector`` and ``Package`` are deprecated;
use ``self.session.gethookproxy()`` and ``self.session.isinitpath()`` instead.
This should work on all pytest versions.
Features
--------
-`#7667 <https://github.com/pytest-dev/pytest/issues/7667>`_: New ``--durations-min`` command-line flag controls the minimal duration for inclusion in the slowest list of tests shown by ``--durations``. Previously this was hard-coded to ``0.005s``.
Improvements
------------
-`#6681 <https://github.com/pytest-dev/pytest/issues/6681>`_: Internal pytest warnings issued during the early stages of initialization are now properly handled and can filtered through :confval:`filterwarnings` or ``--pythonwarnings/-W``.
This also fixes a number of long standing issues: `#2891 <https://github.com/pytest-dev/pytest/issues/2891>`__, `#7620 <https://github.com/pytest-dev/pytest/issues/7620>`__, `#7426 <https://github.com/pytest-dev/pytest/issues/7426>`__.
-`#7572 <https://github.com/pytest-dev/pytest/issues/7572>`_: When a plugin listed in ``required_plugins`` is missing or an unknown config key is used with ``--strict-config``, a simple error message is now shown instead of a stacktrace.
-`#7685 <https://github.com/pytest-dev/pytest/issues/7685>`_: Added two new attributes :attr:`rootpath <_pytest.config.Config.rootpath>` and :attr:`inipath <_pytest.config.Config.inipath>` to :class:`Config <_pytest.config.Config>`.
These attributes are :class:`pathlib.Path` versions of the existing :attr:`rootdir <_pytest.config.Config.rootdir>` and :attr:`inifile <_pytest.config.Config.inifile>` attributes,
and should be preferred over them when possible.
-`#7780 <https://github.com/pytest-dev/pytest/issues/7780>`_: Public classes which are not designed to be inherited from are now marked `@final <https://docs.python.org/3/library/typing.html#typing.final>`_.
Code which inherits from these classes will trigger a type-checking (e.g. mypy) error, but will still work in runtime.
Currently the ``final`` designation does not appear in the API Reference but hopefully will in the future.
Bug Fixes
---------
-`#1953 <https://github.com/pytest-dev/pytest/issues/1953>`_: Fixed error when overwriting a parametrized fixture, while also reusing the super fixture value.
.. code-block:: python
# conftest.py
import pytest
@pytest.fixture(params=[1, 2])
def foo(request):
return request.param
# test_foo.py
import pytest
@pytest.fixture
def foo(foo):
return foo * 2
-`#4984 <https://github.com/pytest-dev/pytest/issues/4984>`_: Fixed an internal error crash with ``IndexError: list index out of range`` when
collecting a module which starts with a decorated function, the decorator
raises, and assertion rewriting is enabled.
-`#7591 <https://github.com/pytest-dev/pytest/issues/7591>`_: pylint shouldn't complain anymore about unimplemented abstract methods when inheriting from :ref:`File <non-python tests>`.
-`#7628 <https://github.com/pytest-dev/pytest/issues/7628>`_: Fixed test collection when a full path without a drive letter was passed to pytest on Windows (for example ``\projects\tests\test.py`` instead of ``c:\projects\tests\pytest.py``).
-`#7638 <https://github.com/pytest-dev/pytest/issues/7638>`_: Fix handling of command-line options that appear as paths but trigger an OS-level syntax error on Windows, such as the options used internally by ``pytest-xdist``.
-`#7742 <https://github.com/pytest-dev/pytest/issues/7742>`_: Fixed INTERNALERROR when accessing locals / globals with faulty ``exec``.
Improved Documentation
----------------------
-`#1477 <https://github.com/pytest-dev/pytest/issues/1477>`_: Removed faq.rst and its reference in contents.rst.
Trivial/Internal Changes
------------------------
-`#7536 <https://github.com/pytest-dev/pytest/issues/7536>`_: The internal ``junitxml`` plugin has rewritten to use ``xml.etree.ElementTree``.
The order of attributes in XML elements might differ. Some unneeded escaping is
no longer performed.
-`#7587 <https://github.com/pytest-dev/pytest/issues/7587>`_: The dependency on the ``more-itertools`` package has been removed.
-`#7631 <https://github.com/pytest-dev/pytest/issues/7631>`_: The result type of :meth:`capfd.readouterr() <_pytest.capture.CaptureFixture.readouterr>` (and similar) is no longer a namedtuple,
but should behave like one in all respects. This was done for technical reasons.
-`#7671 <https://github.com/pytest-dev/pytest/issues/7671>`_: When collecting tests, pytest finds test classes and functions by examining the
attributes of python objects (modules, classes and instances). To speed up this
process, pytest now ignores builtin attributes (like ``__class__``,
``__delattr__`` and ``__new__``) without consulting the :confval:`python_classes` and
:confval:`python_functions` configuration options and without passing them to plugins
using the :func:`pytest_pycollect_makeitem <_pytest.hookspec.pytest_pycollect_makeitem>` hook.
pytest 6.0.2 (2020-09-04)
=========================
@@ -606,7 +1055,7 @@ Bug Fixes
-`#7110 <https://github.com/pytest-dev/pytest/issues/7110>`_: Fixed regression: ``asyncbase.TestCase`` tests are executed correctly again.
-`#7143 <https://github.com/pytest-dev/pytest/issues/7143>`_: Fix ``File.from_constructor`` so it forwards extra keyword arguments to the constructor.
-`#7143 <https://github.com/pytest-dev/pytest/issues/7143>`_: Fix ``File.from_parent`` so it forwards extra keyword arguments to the constructor.
-`#7145 <https://github.com/pytest-dev/pytest/issues/7145>`_: Classes with broken ``__getattribute__`` methods are displayed correctly during failures.
@@ -1682,6 +2131,44 @@ Improved Documentation
-`#5416 <https://github.com/pytest-dev/pytest/issues/5416>`_: Fix PytestUnknownMarkWarning in run/skip example.
pytest 4.6.11 (2020-06-04)
==========================
Bug Fixes
---------
-`#6334 <https://github.com/pytest-dev/pytest/issues/6334>`_: Fix summary entries appearing twice when ``f/F`` and ``s/S`` report chars were used at the same time in the ``-r`` command-line option (for example ``-rFf``).
The upper case variants were never documented and the preferred form should be the lower case.
-`#7310 <https://github.com/pytest-dev/pytest/issues/7310>`_: Fix ``UnboundLocalError: local variable 'letter' referenced before
assignment`` in ``_pytest.terminal.pytest_report_teststatus()``
when plugins return report objects in an unconventional state.
This was making ``pytest_report_teststatus()`` skip
entering if-block branches that declare the ``letter`` variable.
The fix was to set the initial value of the ``letter`` before
the if-block cascade so that it always has a value.
pytest 4.6.10 (2020-05-08)
==========================
Features
--------
-`#6870 <https://github.com/pytest-dev/pytest/issues/6870>`_: New ``Config.invocation_args`` attribute containing the unchanged arguments passed to ``pytest.main()``.
Remark: while this is technically a new feature and according to our `policy <https://docs.pytest.org/en/latest/py27-py34-deprecation.html#what-goes-into-4-6-x-releases>`_ it should not have been backported, we have opened an exception in this particular case because it fixes a serious interaction with ``pytest-xdist``, so it can also be considered a bugfix.
Trivial/Internal Changes
------------------------
-`#6404 <https://github.com/pytest-dev/pytest/issues/6404>`_: Remove usage of ``parser`` module, deprecated in Python 3.9.
pytest 4.6.9 (2020-01-04)
=========================
@@ -2630,12 +3117,12 @@ Features
will not issue the warning.
-`#3632 <https://github.com/pytest-dev/pytest/issues/3632>`_: Richer equality comparison introspection on ``AssertionError`` for objects created using `attrs <http://www.attrs.org/en/stable/>`__ or `dataclasses <https://docs.python.org/3/library/dataclasses.html>`_ (Python 3.7+, `backported to 3.6 <https://pypi.org/project/dataclasses>`__).
-`#3632 <https://github.com/pytest-dev/pytest/issues/3632>`_: Richer equality comparison introspection on ``AssertionError`` for objects created using `attrs <https://www.attrs.org/en/stable/>`__ or `dataclasses <https://docs.python.org/3/library/dataclasses.html>`_ (Python 3.7+, `backported to 3.6 <https://pypi.org/project/dataclasses>`__).
-`#4278 <https://github.com/pytest-dev/pytest/issues/4278>`_: ``CACHEDIR.TAG`` files are now created inside cache directories.
Those files are part of the `Cache Directory Tagging Standard <http://www.bford.info/cachedir/spec.html>`__, and can
Those files are part of the `Cache Directory Tagging Standard <https://bford.info/cachedir/spec.html>`__, and can
be used by backup or synchronization programs to identify pytest's cache directory as such.
@@ -4619,7 +5106,7 @@ Improved Documentation
- In one of the simple examples, use ``pytest_collection_modifyitems()`` to skip
tests based on a command-line option, allowing its sharing while preventing a
user error when acessing ``pytest.config`` before the argument parsing.
user error when accessing ``pytest.config`` before the argument parsing.
The accompanying ``py.path.local`` based paths have been deprecated: plugins which manually invoke those hooks should only pass the new ``pathlib.Path`` arguments, and users should change their hook implementations to use the new ``pathlib.Path`` arguments.
``Node.fspath`` in favor of ``pathlib`` and ``Node.path``
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.
