mikebridge commented on PR #40130: URL: https://github.com/apache/superset/pull/40130#issuecomment-4861924022
Thanks @rusackas — all five addressed: 1. **Squashed** ✅ — the PR is now a single commit (`123db2618c`); the tree is byte-identical to the last green CI run, only history changed. **Bot threads**: doing one clean triage pass now (the recent pushes triggered bot re-reviews, so sweeping once rather than chasing a moving target) — each will get a reply-with-rationale or a fix, same as the sweep on #39859. 2. **Already rebased** ✅ — the branch sits directly on current master (`393adc45`), which includes #41166, so CI has been running the gated path: the soft-delete suites run under `@with_feature_flags(SOFT_DELETE=True)`, and the full DB matrix is green (the one red ❌ on a previous run was an artifact-upload timeout *after* all 2393 MySQL tests passed — a re-run clears it). 3. **Confirmed intentional** ✅ — both are pre-existing-bug fixes that are correct regardless of the flag: the catalog-NULL-aware uniqueness matching (a `catalog=NULL` row *is* the default-catalog dataset; the old predicate let twin pairs slip through) and the duplicate-scope fix (collisions should be scoped to (database, catalog, schema), not global-by-name). Since they ship un-gated, they're called out explicitly in UPDATING.md under "Uniqueness-validation changes that apply regardless of the feature flag." 4. **Intended** ✅ — deliberate no-orphaned-rows/no-lost-restorables trade-off, documented in UPDATING.md including the operator workaround. The planned resolution is the deletion-retention purge (#41549), which is sequenced after the entity PRs. 5. **Good catch — extracted** ✅ — the `_escape_like` hardening is now #41653 and this branch is back to master's version. Bonus: splitting it out surfaced a real bug in the hardening — coercing `None` to `""` built a wildcard-only pattern (`%%`) that silently matched **every** row. The standalone PR fixes the semantics properly (`None` matches no rows, mirroring SQL's `x LIKE NULL`), with tests that reproduce the original `AttributeError` against master. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
