[afonsobspinto]

Purpose

Draft for internal review before opening against upstream catmaid/CATMAID:dev. The worker condition fix is already included in feature/new-ingestion (PR #7). This PR isolates that fix and adds focused regression coverage for upstream preparation.

Problem and resulting behavior

The normal grid cache worker checks n_last_edited_skeletons_limit before reading n_largest_skeletons_limit, although the filters are independently configurable. As a result, configuring only the largest-skeleton filter silently drops it; configuring only the last-edited filter attempts int(None) and fails before refreshing the cell.

Grid settings	Before	After
Largest = 5; last edited unset	Largest-skeleton filter is ignored.	Cell generation receives n_largest_skeletons_limit=5.
Largest unset; last edited = 3	TypeError from int(None) prevents refresh.	Cell generation receives only n_last_edited_skeletons_limit=3.
Largest = 5; last edited = 3	Both filters are applied.	Same.
Both unset	No skeleton-count filter.	Same.

Changes

• Gate the largest-skeleton parameter on its own setting and explain why the condition is independent.
• Add four database-free worker unit tests covering each combination of the two filters, with the database and cell-generation collaborators mocked.

This PR contains no importer, cache transaction, migration or schema changes. It is independent of the cell publication changes in PR #9.

Validation

• All four regression tests passed in an isolated Python 3.12 / Django 4.2 harness importing the real worker with its web/database collaborators mocked. No PostgreSQL or full deployment startup was exercised.
• Negative control: restoring the previous condition causes the largest-only case to fail and the last-edited-only case to raise TypeError; the other two cases still pass.
• Python syntax compilation and git diff --check passed.

In a configured CATMAID environment, run:

python manage.py test catmaid.tests.test_cache_update_worker

The affected worker condition is also present in current upstream dev. After internal review, cherry-pick this commit onto upstream dev for the upstream PR.