-
Notifications
You must be signed in to change notification settings - Fork 37
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
Release 0.11.0 #108
Release 0.11.0 #108
Conversation
Remove unnecessary debug prints
Create pre-commit setup
Also updates black version used, and sets max line length of 120
Fix all black and flake warnings
Bumps [django](https://github.com/django/django) from 2.2.27 to 2.2.28. - [Release notes](https://github.com/django/django/releases) - [Commits](django/django@2.2.27...2.2.28) --- updated-dependencies: - dependency-name: django dependency-type: direct:development ... Signed-off-by: dependabot[bot] <support@github.com>
fix custom pk incompatibility errors
`many_to_one_extras` -> `many_to_many_extras` in the relevant section.
Bumps [sqlparse](https://github.com/andialbrecht/sqlparse) from 0.4.2 to 0.4.4. - [Release notes](https://github.com/andialbrecht/sqlparse/releases) - [Changelog](https://github.com/andialbrecht/sqlparse/blob/master/CHANGELOG) - [Commits](andialbrecht/sqlparse@0.4.2...0.4.4) --- updated-dependencies: - dependency-name: sqlparse dependency-type: indirect ... Signed-off-by: dependabot[bot] <support@github.com>
Allow overwriting return ids in Batch Delete
Prior versions have reached EOL: https://endoflife.date/django
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.
These changes look good to me! I pip-installed this branch into a medium/large-sized project where I have dozens of unit-tests for various mutation behavior that utilizes graphene-django-cud (Create, BatchCreate, Patch, and Delete specifically) and all tests still pass.
I think it may be worth trying to merge #110 (and potentially #111 as well, since that seems to resolve an issue several people ran into), if possible. Thanks again for reviving graphene-django-cud and pushing another release forward 🙏
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.
Looks good to me, aside from all the checks failing (likely just because ubuntu-18.04
is deprecated, is probably fine just changing to ubuntu-latest
).
Change ubuntu version from 18 to 22
Remove testing for python version 3.6, and add for python version 3.9 and 3.10.
- Change ubuntu running versions from 22 to 20 (for potential compatability issues with older versions of python in Ubuntu 22) - Make python versions in matrix be strings - Make release-task use ubuntu 20 and python version 3.9
- Upgrade to newer way of installation poetry - Change caching strategy to doing it directly in the python setup
- Update some packages further - Fix issues after updating factoryboy
Fix issue #81 and #105 properly, by amending the way we handle already-converted choices field from graphene-django 3.0. The fix introduced in #111 made the error people experienced from the bad handling go away, but it didn't actually set the `is_required` flag correctly. This is because mounted BlankValue Field actually wrapped the underlying choices enum in `NonNull` when pertinent. Ths issue was related to graphene-django 3.0 taking the converted enum class and mounting it outright in a BlankValueField, instead of just saving the enum itself. We solve this by, for graphene-django 3.0, by retrieving the wrapped choices Enum, and reusing this.
Tests are running successfully again. Getting close to the release. The pre-commits are still failing. Care to check that out @mbuvarp? @sjdemartini Is your medium/large-size repo using graphene-django 2.x? If so, could you run another quick check from the latest HEAD here? I've had to implement another quick fix for the issue outlined in #81 and #105. |
@tOgg1 Thanks for all the effort here. We upgraded to graphene-django v3 a few weeks ago (though not the latest patch version yet due to the issue in #109). I haven't seen the error that was mentioned in #81 and #105. I noticed both of the folks who reported it were using Django 4, and we're using 3.2, so perhaps that accounts for the difference, but I haven't looked into it in any detail. I can try to investigate more what would be causing that if you'd like. |
Oh, it occurred to me just now that what's likely the key difference is that I use a separate enum class which I manually define for each of my graphene_django object types (since I want more control over enum naming whenever I use |
Thanks, @sjdemartini! I think it's very likely just a graphene-django v3 vs graphene-django v2 issue. I'll run some tests myself with graphene-django 2.0, and see if anything breaks. |
Fix compatibility issue with importing BlankValueField and assert_name from graphql Signed-off-by: Tormod Haugland <tormod.haugland@gmail.com>
Hm, I get a new error trying the latest commit:
These are the relevant package versions I'm using:
It seems graphql-core changed graphene-django-cud/graphene_django_cud/converter.py Lines 39 to 40 in 3a5e232
should seemingly be changed to be specific to 3.2+ (not 3+). Side note: I was already under the impression that graphene-django-cud dropped support for graphene-django 2 in its 0.10.0 release due to its version specification graphene-django-cud/pyproject.toml Line 10 in 82e3662
graphene_django v2 (folks using that could stick with <0.10.0 presumably). That seems to be the trend with most graphene-django third party packages, where they dropped support for v2 when adding support for v3.
|
Change the target graphql version from 3 to 3.2 when differing on whether or not to use "assert_name" or "assert_valid_name" Signed-off-by: Tormod Haugland <tormod.haugland@gmail.com>
Signed-off-by: Tormod Haugland <tormod.haugland@gmail.com>
Released. Thanks for all the help guys! |
Preparation of the 0.11.0 release.
@mbuvarp Would you be so kind as to do a final review?