Prepare to upgrade to django-oauth-toolkit (module name oauth2_provider) 1.3.2 concurrently with the Django 2.2 upgrade (versions after 1.1.3 dropped support for Django 1.11). Key points:
* We can stop using a commit hash since versions 1.3.0 through 1.3.2 have been released since that commit.
* The validator in oauth_dispatch migration 0002 has been removed from DOT, but the whole table was removed in migration 0004 anyway. Newer DOT versions moved where redirect URI validation happens so it's no longer needed here.
* DOT 1.3.0 squashed its original 6 migrations and immediately deleted them. This required some changes in oauth_dispatch migrations 0006 and 0007 to deal with either numbering system.
When deploying to stage and production, we'll need to delete the history for oauth2_provider migrations 0002-0006 first (the new squashed 0001 migration has the same name as the original 0001). The deployment with DOT 1.3.2 will add a new 0002 migration which should then apply cleanly on the existing database state.
Should resolve https://openedx.atlassian.net/browse/BOM-1456 .
This looks like the last failing unit test under Django 2.2. One of a series of such failures resulting from Django no longer accepting None as POST data (since you can't really get it from a real request), most of which were already fixed by the Arbisoft team in the last few days.
This dashboard template is only used for edx.org so we didn't catch this
in testing. Updating it to use 'course' instead of 'course_id' to
instantiate a CourseEnrollment object. Using 'course_id' is the old
way.
We don't use this field in edx-platform since we rely on UserProfile
instead. But we need to do this to prevent accidental incorrect use of
this field. Django adds a migration between 1.11 and 2.2 to increase
the length of this field that we are opting to skip. This change
enforces the old limit.
We now either pass in the relevant courseoverview or when creating the
enrollement we use the factory which automatically creates the relevant
CourseOverview object for testing purposes.
Historically, the CourseEnrollment model used to have a `course_id`
field. A lot of tests still call the factory using that. Instead of a
`course_id` field this model now has a `course` field which is a foreign
key to the CourseOverview model.
The factory still accepts the course_id but uses it to create the
relevant CourseOverview object where necessary. This commit fixes two
issues with the factory.
1. If the course id is passed in as`course_id` instead of `course__id`
then, the generated CourseOverview does not have the correct course_id.
2. Even though the CourseEnrollment model no longer needs the
`course_id` parameter, we were still passing it through. We now remove
it so that it is not passed through to the CourseEnrollment model
instantiation.
It conflicts with an underlying related field on that model which seems
to be getting at the same value from the related table.
Add logging for incorrectly instantiating CourseEnrollment models.
This is to catch any places that might break this that are outside of
edx-platform. Django won't accept `course` values that aren't
course_overviews so we don't need extra logic to test that `course`
values are of the correct type.
fixup! Remove the course_id property on student.CourseEnrollment.
This adds contraints to:
kiwisolver: next release only works with python>=3.6
The was run to upgrade:
django-splash
django-wiki
edx-search
The previous versions were constraining django below 2.0.
Co-authored-by: Robert Raposa <rraposa@edx.org>
