Stop vendoring pluggy #2719
Stop vendoring pluggy #2719
Conversation
setup.py
Outdated
| @@ -43,7 +43,7 @@ def has_environment_marker_support(): | |||
|
|
|||
|
|
|||
| def main(): | |||
| install_requires = ['py>=1.4.33', 'six>=1.10.0','setuptools'] # pluggy is vendored in _pytest.vendored_packages | |||
| install_requires = ['py>=1.4.33', 'six>=1.10.0','setuptools', 'pluggy==0.4.0'] | |||
There was a problem hiding this comment.
Perhaps 'pluggy>=0.4.0,<0.5'?
_pytest/_pluggy.py
Outdated
| @@ -1,11 +1,6 @@ | |||
| """ | |||
| imports symbols from vendored "pluggy" if available, otherwise | |||
| falls back to importing "pluggy" from the default namespace. | |||
| Import symbols from ``pluggy`` | |||
There was a problem hiding this comment.
I think we can get rid of this module altogether and just import pluggy docs normally no?
doc/en/writing_plugins.rst
Outdated
| @@ -454,7 +454,7 @@ hook wrappers and passes the same arguments as to the regular hooks. | |||
|
|
|||
| At the yield point of the hook wrapper pytest will execute the next hook | |||
| implementations and return their result to the yield point in the form of | |||
| a :py:class:`CallOutcome <_pytest.vendored_packages.pluggy._CallOutcome>` instance which encapsulates a result or | |||
| a :py:class:`CallOutcome <_pytest._pluggy._CallOutcome>` instance which encapsulates a result or | |||
There was a problem hiding this comment.
Could you please create the docs locally (tox -e docs) to ensure the changes links still work?
There was a problem hiding this comment.
@nicoddemus yep I'll test it out
_pytest/_pluggy.py
Outdated
| from pluggy import * # noqa | ||
| from pluggy import __version__ # noqa | ||
| from pluggy import * | ||
| from pluggy import __version__ |
There was a problem hiding this comment.
we might want to consider just directly inlining the imports as a followup
There was a problem hiding this comment.
I think we can do it in the same PR if @tgoodlet is willing
There was a problem hiding this comment.
@nicoddemus @RonnyPfannschmidt yup no problem guys.
I'll take a look this morning.
|
@RonnyPfannschmidt @nicoddemus alright made the changes. |
|
personally i like the history intact, but please add a changelog fragment |
There was a problem hiding this comment.
Thanks, looks good!
Just missing adding a trivial changelog entry like @RonnyPfannschmidt asked and then is good to merge from my POV.
Regarding history, I like to tidy the history (and always do on my PRs), so we can leave that to you. 😁
|
@nicoddemus @RonnyPfannschmidt I noticed in the
Are we doing that as well or do I just need to add a |
|
IMO just the |
|
@nicoddemus good enough? |
|
Yep, thanks! |
|
@nicoddemus I'm assuming those CI failures are expected? |
|
That seems odd - I opened HypothesisWorks/hypothesis#822 |
|
Thanks @The-Compiler :) |
|
Once you guys merge this I'll put up a new PR to bring in |
|
another item to consider, we could now cython-ize pluggy for cpython ^^ |
|
@RonnyPfannschmidt good idea. |
If this is too much let me know guys ;)
Pretty sure I got everything but someone should double check.