Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #1144 +/- ##
==========================================
- Coverage 87.82% 87.78% -0.04%
==========================================
Files 89 89
Lines 6150 6150
Branches 111 111
==========================================
- Hits 5401 5399 -2
- Misses 693 695 +2
Partials 56 56 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
adombeck
left a comment
adombeck
left a comment
There was a problem hiding this comment.
Nice work! Just a few small comments. I'm glad we don't need a schema migration :)
Oh, I must have missed that in the review of #782. Do you know what the locked column is treated as? Is it also TEXT? 😵 |
|
Nice, migrations check on a more complete setup are already covered by the integration tests we already have, right? |
The change works without any schema migration which the tests added here prove, so I don't see the need for testing more complex setups - or am I missing something? |
It is treated as INT |
TestLoadOldSchemaV2Database ensures databases created with the old schema (v2, ugid INT) can be opened and read correctly by the current code without additional migrations.
TestMigrateOldSchemaV1 ensures databases created with schema v1 (no 'locked' column) are transparently migrated on open and readable via current code.
4 participants
ugid are ids provided by the broker that we do not control. They are rightfully treated as strings in the brokers and authd, but declared as INT in the database schema.
This PR uses the right type for UGID and adds 2 tests for the migration from v1 and old v2.
STRICT is not enforced at the moment, because it would require to change BOOLEAN of the locked column as well.
The casing of the table name has been fixed as well.
UDENG-7762