Uh oh!
There was an error while loading. Please reload this page.
Filter channelsetlist by user - #2613
Conversation
…esults where the current user is the editor
Codecov Report
@@ Coverage Diff @@## develop #2613 +/- ##
===========================================
- Coverage 84.61% 84.56% -0.05%
===========================================
Files 295 296 +1 Lines 15018 15093 +75 ===========================================
+ Hits 12707 12764 +57 - Misses 2311 2329 +18
Continue to review full report at Codecov.
|
rtibbles
left a comment
There was a problem hiding this comment.
Suggestion for a slight change of approach.
| filter_class = ChannelSetFilter | ||
| permission_classes = [IsAuthenticated] | ||
| values = ("id", "name", "description", "channels", "secret_token__token") | ||
| values = ("id", "name", "description", "channels", "secret_token__token", "editors") |
There was a problem hiding this comment.
There is an interesting gotcha here, where adding editors to the values call, because editors is a many to many field, if there is more than one editor, then this will cause the same channelset to be returned multiple times in the resultant query (each with a different editors value corresponding to each editor).
Although currently we have no way in the interface of adding additional editors to a collection, I imagine we will do so at some point in the future.
I would instead do this consistently with how we filter Channels, and instead annotate an edit flag on the viewset, which is then different on a per user basis.
This is annotated onto the Channels queryset here: https://github.com/learningequality/studio/blob/develop/contentcuration/contentcuration/models.py#L780 (this method is called for getting the queryset inside these viewsets, so doing something parallel in the method on ChannelSet would mean this annotation can be used in the FilterSet and returned as a value).
There was a problem hiding this comment.
Thanks for the feedback - I've made some updates based on these notes and our conversation yesterday, and I think it's now using the approach you suggest. Let me know if you have any other comments!
| queryset = super(ChannelSetViewSet, self).get_queryset() | ||
| user_id = not self.request.user.is_anonymous() and self.request.user.id | ||
| edit = Exists(User.channel_sets.through.objects.filter(user_id=user_id, channelset_id=OuterRef("id"))) | ||
| return queryset.annotate(edit=edit).filter(edit=True) |
There was a problem hiding this comment.
Should remove this: .filter(edit=True) otherwise we are building in the filter you've added here into the base queryset.
Beyond that, this is good to go!
rtibbles
left a comment
There was a problem hiding this comment.
Manually tested, filtering is working as intended.
What does this PR do? Briefly describe in 1-2 sentences
This PR filters the ChannelSetList (aka "Collections") so that only the current user's Collections are displayed.
Issue Addressed (if applicable)
Addresses #2542 - Admins can see all Channelsets/Collections, even if they do not have edit access to them
Steps to Test
At a high level, how did you implement this?
I created a FilterSet on the backend for ChannelSetList and passed a parameter in the front end .where() query and updated tests for this change.