[backend] Fix simulation to scenario grants

[RomuDeuxfois]

error → "could not execute statement [ERROR: duplicate key value violates unique constraint "grant"\n Detail: Key (grant_group, grant_exercise, grant_name)=(d5d8c183-e6d9-4d47-b041-25e5f3bdbc62, 375280f3-7207-4f6b-95f8-2b1aaa480697, PLANNER) already exists.] [insert into grants (grant_exercise,grant_group,grant_name,grant_scenario,grant_id) values (?,?,?,?,?)]"

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 41.22%. Comparing base (62b0a78) to head (702642d).
Report is 26 commits behind head on release/current.

Additional details and impacted files

@@                  Coverage Diff                  @@
##             release/current    #3055      +/-   ##
=====================================================
+ Coverage              41.20%   41.22%   +0.01%     
- Complexity              2286     2288       +2     
=====================================================
  Files                    682      683       +1     
  Lines                  20938    20944       +6     
  Branches                1424     1424              
=====================================================
+ Hits                    8628     8634       +6     
  Misses                 11824    11824              
  Partials                 486      486              

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[antoinemzs]

I'm concerned we are reimplementing the table constraint in the service layer: I understand this is to avoid triggering a constraint violation while inserting new grants.

I'd be more concerned with findijng where in the app we are adding a duplicate grant that ultimately conflicts; is this about default grants?

Somehow this should perhaps be handled in the repository if this is not possible to reconcile in code: https://www.postgresql.org/docs/current/sql-insert.html#SQL-ON-CONFLICT

This is a Postgres-specific clause (i.e. not standard SQL) which can ignore an insert upon a constraint violation (by name), which would fit the bill here.
Since we reassign the saved grants after the inserts we would have deduplicated objects eventually in the java collection.

[RomuDeuxfois]

I totally agree with you. I looked through the code and I don't understand how we got to this situation.
I had a recurring scenario, and the problem occurred at the end of the X recurrence.

However, I wasn't aware of SQL ON CONFLICT, which is much more appropriate. I'll look into it, thanks for the resource!

[RomuDeuxfois]

@antoinemzs I found the root cause. I made some changes.

[antoinemzs]

Cheers for finding a hole in our constraints !