Python types have reference cycles to themselves when they are created. This is
partially caused by descriptors which get / set values from the __dict__
attribute for getattr / setattr on classes.
This is not normally an issue since types tend to remain referenced for the
lifetime of the Python process (and thus never become garbage).
However, in the case of PseudoFixtureDef, the class is generated in
_get_active_fixturedef and later discarded when pytest_fixture_setup returns.
As a result, the generated PseudoFixtureDef type becomes garbage.
This is not really a performance issue but it can lead to some problems when
making tests and assertions about garbage when using pytest.
This garbage creation problem can be rectified by returning a namedtuple
instance which is functionally the same. In the modified code, the namedtuple
is allocated / deallocated using reference counting rather than having to use
the garbage collector.
Adds a badge showing the number of people helping this repo on CodeTriage.
[](https://www.codetriage.com/pytest-dev/pytest)
## What is CodeTriage?
CodeTriage is an Open Source app that is designed to make contributing to Open Source projects easier. It works by sending subscribers a few open issues in their inbox. If subscribers get busy, there is an algorithm that backs off issue load so they do not get overwhelmed
[Read more about the CodeTriage project](https://www.codetriage.com/what).
## Why am I getting this PR?
Your project was picked by the human, @schneems. They selected it from the projects submitted to https://www.codetriage.com and hand edited the PR. How did your project get added to [CodeTriage](https://www.codetriage.com/what)? Roughly 11 months ago, [cacoze](https://github.com/cacoze) added this project to CodeTriage in order to start contributing. Since then, 32 people have subscribed to help this repo.
## What does adding a badge accomplish?
Adding a badge invites people to help contribute to your project. It also lets developers know that others are invested in the longterm success and maintainability of the project.
You can see an example of a CodeTriage badge on these popular OSS READMEs:
- [](https://www.codetriage.com/rails/rails) https://github.com/rails/rails
- [](https://www.codetriage.com/crystal-lang/crystal) https://github.com/crystal-lang/crystal
## Have a question or comment?
While I am a bot, this PR was manually reviewed and monitored by a human - @schneems. My job is writing commit messages and handling PR logistics.
If you have any questions, you can reply back to this PR and they will be answered by @schneems. If you do not want a badge right now, no worries, close the PR, you will not hear from me again.
Thanks for making your project Open Source! Any feedback is greatly appreciated.
This option is superseded by the --show-capture option. With --no-print-logs
it was possible to only disable the reporting of captured logs, which is no
longer possible with --show-capture. If --show-capture=no is used, no
captured content (stdout, stderr and logs) is reported for failed tests.
When a test with pytest.raises(ValueError, match='foo') doesn't raise, the
following error is printed:
Failed: DID NOT RAISE <class 'ValueError'> matching 'foo'
This error message is confusing as it implies a ValueError was raised, but the
message wasn't matching 'foo'.
I first considered rewording it somehow to preserve the match pattern in it, but
I don't think that's worthwhile as the pattern should usually be apparent from
the stacktrace anyways (hard-coded, as parametrization, or with --showlocals for
more sophisticated cases).
Ronny Pfannschmidt
Allan Feldman
Bruno Oliveira
mike
Carlos Jenkins
codetriage-readme-bot
Thomas Hisch
Brian Maissy
Jason R. Coombs
Jordan Speicher
Florian Bruhin
