This should cut down the API rate limiting we see when large numbers of
PRs are submitted.
Note when we used mypyc-compiled mypy in mypy_primer, this took <10m.
This is currently not the case, but hopefully will once again be.
Co-authored-by: hauntsaninja <>
Fixes#5270
The comment workflow waits for the mypy primer to run. I think with
Github's changes to workflow running, these runs get marked as
completed, but have the conclusion "action_required".
E.g., see https://github.com/python/typeshed/pull/5269/checks
I recently published https://github.com/marketplace/actions/run-pyright, which wraps pyright and uses GHA magic to leave diagnostics on commits / PRs (like https://github.com/jakebailey/pyright-action-test/pull/1/files), plus a faster startup time thanks to GHA tool caching and piggy backing off of the `node` install used in GHA itself (so no `setup-node` needed).
This PR switches to that action and leaves a comment noting that the version number is pinned in two places.
This action is a prototype, but I'm pretty confident in it so far.
Run the mypy test suite using Ubuntu 18.04. Ubuntu 20.04 doesn't have a package for virtualenv on Python 2.
Also, curses.color_pair() now has a named argument.
For now, the test requires changes that have been merged into the pytype master branch but not yet released. I'll update requirements-tests-py3.txt again once I've cut a new release (hopefully later this week).
* Fix pytype_test.
* Re-enable pytype_test.
* Remove continue-on-error from pytype_test.
* Temporarily install pytype from its master branch, as the code to
support the new directory structure hasn't been released yet.
* Set TYPESHED_HOME before using pytype.pytd.Typeshed().
* typeshed: update stubtest version
Includes changes from https://github.com/python/mypy/pull/9680
I've already fixed all the true positives on typeshed.
* attempt to fix windows' _WarnFunction
Co-authored-by: hauntsaninja <>
We use tee and no longer try to delete the file based on exit code,
since we use data.trim now.
mypy_primer started failing silently because of changes in aiohttp, so
now we'll fail if the exit codes are not 0 or 1.
Note mypy_primer has --project-date to get around the problem of
breaking changes in projects, but I want to try and keep mypy_primer up
to date / get a feel for how bad this problem actually is.
Co-authored-by: hauntsaninja <>
In #4696 I bumped stubtest to a commit that hasn't been released. We
should have these stubtest versions match. In
https://github.com/python/mypy/pull/9426 I changed the output of
stubtest here; it's good to get this change in before I forget and this
action silently stops working.
Co-authored-by: hauntsaninja <>
* Find unused stubtest whitelist entries
This is currently a GitHub workflow that runs daily and lists all unused
whitelist entries found by running stubtest against current Python
versions on Linux and Windows. The workflow run will succeed if there
are no such entries, and fail otherwise.
In a second step, this should collate the output of the various runs and
create a PR to remove the entries. In that case, the workflow should
probably only run weekly or even monthly to keep the noise down.
Cf. #3728
