Migrate inbound_group_session2 to fix keys incorrectly copied from old store version

[andybalaam]

Fixes element-hq/element-web#26782

In migrate_data_for_v6, we incorrectly copied the keys in inbound_group_sessions verbatim into inbound_group_sessions2. What we should have done is re-encrypt them using the new table name, so we fix that up with a new migration here.

This caused the bug because we were looking for sessions to mark as backed up by calculating their key (from room_id and session_id) but that key did not exist, because the old sessions were stored under the incorrect keys. So no sessions were marked as backed up, and we repeatedly tried to re-mark them.

Depending on your taste you might like to review this commit by commit.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (320b868) 83.50% compared to head (efea767) 83.50%.
Report is 17 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2957   +/-   ##
=======================================
  Coverage   83.50%   83.50%           
=======================================
  Files         221      221           
  Lines       23007    23007           
=======================================
  Hits        19211    19211           
  Misses       3796     3796           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[andybalaam]

Running a git bisect to try and figure out what I did to the tests...

[andybalaam]

Running a git bisect to try and figure out what I did to the tests...

OK, Rich helped me figure out it was due to missing calls to db.close(). In a future PR I will attempt to protect against forgetting this.

[richvdh]

lgtm otherwise.

[Hywan]

I don't have all the knowledge to validate this patch, but the code looks good (modulo 2 tiny comments).

[richvdh]

I don't have all the knowledge to validate this patch, but the code looks good (modulo 2 tiny comments).

Thanks @Hywan! I'm pretty happy with this from the functionality side, so if you're happy from the code style side, I think we're good to go.