mikebridge commented on PR #43776: URL: https://github.com/apache/superset/pull/43776#issuecomment-5546394317
@aminghadersohi — your reading is right and mine is corrected at the new head. I traced your chain at the cited lines and it holds end to end: the rewrite is `datePickerInAdhocFilterMixin.initialValue` (`shared-controls/mixins.tsx` — the `granularity_sqla && time_range` branch returning the `TEMPORAL_RANGE` simple filter), evaluated inside `applyMapStateToPropsToControl` — reached through `getControlsState` (`hydrateExplore.ts:143`) and the explicit loop at `:217`, both **before** `hydratedFormData` (`:222`) — so `adhoc_filters` is a hydration transition, still equals `to_value` in the saved params, and `filter_normalization_records` drops it on a plain save. Your narrower surviving effect is also exactly what the code says: `hydrateExplore.ts:115` clears `granularity_sqla`, and `normalization.ts:93-102` deliberately doesn't cover disappearing keys (and bails on a non-JSON `to_value`), so the removal records while the `groupby`/`adhoc_filters` additions are suppressed. To answer your direct question honestly: no — I had not watched the entry appear on a plain save. My grounding was #43350's own PR body, whose "deliberate residual" list names the temporal migration alongside `dashboards`/`extra_form_data`; your trace supersedes that prose, which appears inaccurate on that one item. The docs now state the narrower truth: on a plain save the rewrite is suppressed, with at most a removed legacy key appearing as cleared plus the save's own additions; the broader set still belongs to the gated entry points. Thanks for re-asking instead of asserting — the sentence did hang on it. -- 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]
