-
-
Notifications
You must be signed in to change notification settings - Fork 7k
Updated dependencies #7589
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Updated dependencies #7589
Conversation
django-filter>=2.2.0, <2.3 | ||
pygments==2.7.1 | ||
django-guardian==2.3.0 | ||
django-filter>=2.4.0, <2.5 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why are some requirements semantically ranged, and others aren't? Seems like it would make sense to do them all.
@@ -1,7 +1,7 @@ | |||
# PEP8 code linting, which we run on all commits. | |||
flake8==3.8.3 | |||
flake8==3.8.4 | |||
flake8-tidy-imports==4.1.0 | |||
pycodestyle==2.6.0 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Each minor flake8 release only works with one pycodestyle
release. Imo you can remove this pin, since flake8
depends on pycodestyle
- and many other requirements
pycodestyle==2.6.0 |
pytest-cov>=2.7.1 | ||
pytest>=6.1,<6.2 | ||
pytest-django>=3.10.0,<3.11 | ||
pytest-cov>=2.10.1 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Add a max here, like the other semantic ranges, at least <3
?
I've widened some of the ranges:
I'm less sure about some of the others, we've seen breaking changes within Markdown 3.3 for example. coreapi & corescheme are archived, so I've left those as pinned versions. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, just need to fix the conflict with markdown.
4865771
to
05fd4c2
Compare
@adamchainz hope you had a good Christmas! 🎅 Could I ask you for a (hopefully) final review of this? |
pytest-cov>=2.7.1 | ||
pytest>=6.1,<7.0 | ||
pytest-cov>=2.10.1,<3.0 | ||
pytest-django>=3.10.0,<4.0 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
pytest-django can be 4.1+
Sorry I forgot to come back to this @smithdc1 . Could you rebase to fix the conflict and show the latest tests still pass? Then I'd be happy to merge. |
05fd4c2
to
cd4fa83
Compare
Hi Adam, A change in pygments version 2.7.3 causes a markdown test to fail. I suspect this is due to the test containing invalid json. The test was added in #5462, and I can't tell if this was deliberate to get the output with errors. It's this line here that I think is incorrect.
The image below shows the current output, the output with pygments>=2.7.3 and the output if I add a I think we should adjust the test, but I'd appreciate your thoughts. |
Yup, agreed. |
cd4fa83
to
e65aed0
Compare
semantically ranged dependencies alpahetical order dependencies widen allowed version ranges Sorted imports with isort 5.8
e65aed0
to
a8fece5
Compare
It's green so I'll merge. But what happened to the markdown test change? Did a newer pygments release revert the change in output? |
Also for future it could be convenient to move to a setup like I use on my repos: linting with pre-commit, pre-commit.ci running the tools and auto-updating them weekly, and test dependencies pinned with pip-compile, which I update fortnightly with a script. |
I fixed it in #7892 👍 Thanks for the merge. 🎉 |
Following [my comment here](encode#7589 (comment)) and [Django's own move to pre-commit](https://docs.djangoproject.com/en/dev/internals/contributing/writing-code/coding-style/#pre-commit-checks). * Add pre-commit config file to run flake8 and isort. * Add extra "common sense" hooks. * Run pre-commit on GitHub actions using the [official action](https://github.com/pre-commit/action/). This is a good way to get up-and-running but it would be better if we activated [pre-commit.ci](https://pre-commit.ci/), which is faster and will auto-update the hooks for us going forwards. * Remove `runtests.py` code for running linting tools. * Remove `runtests.py --fast` flag, since that would now just run `pytest -q`, which can be done with `runtests.py -q` instead. * Remove tox configuration and requirements files for linting. * Update the contributing guide to mention setting up pre-commit.
Aha yes, great. |
Following [my comment here](#7589 (comment)) and [Django's own move to pre-commit](https://docs.djangoproject.com/en/dev/internals/contributing/writing-code/coding-style/#pre-commit-checks). * Add pre-commit config file to run flake8 and isort. * Add extra "common sense" hooks. * Run pre-commit on GitHub actions using the [official action](https://github.com/pre-commit/action/). This is a good way to get up-and-running but it would be better if we activated [pre-commit.ci](https://pre-commit.ci/), which is faster and will auto-update the hooks for us going forwards. * Remove `runtests.py` code for running linting tools. * Remove `runtests.py --fast` flag, since that would now just run `pytest -q`, which can be done with `runtests.py -q` instead. * Remove tox configuration and requirements files for linting. * Update the contributing guide to mention setting up pre-commit.
Following [my comment here](encode#7589 (comment)) and [Django's own move to pre-commit](https://docs.djangoproject.com/en/dev/internals/contributing/writing-code/coding-style/#pre-commit-checks). * Add pre-commit config file to run flake8 and isort. * Add extra "common sense" hooks. * Run pre-commit on GitHub actions using the [official action](https://github.com/pre-commit/action/). This is a good way to get up-and-running but it would be better if we activated [pre-commit.ci](https://pre-commit.ci/), which is faster and will auto-update the hooks for us going forwards. * Remove `runtests.py` code for running linting tools. * Remove `runtests.py --fast` flag, since that would now just run `pytest -q`, which can be done with `runtests.py -q` instead. * Remove tox configuration and requirements files for linting. * Update the contributing guide to mention setting up pre-commit.
I've bumped all the dependency versions with the exception of Markdown. Bumping that version causes a test fail (separate PR incoming shortly).
The main change here is bumping Pytest to version 6.