amaannawab923 opened a new pull request, #41391:
URL: https://github.com/apache/superset/pull/41391

   ### SUMMARY
   
   When a dataset is configured with a non-zero **Hour Offset**, date filters 
were landing on the wrong calendar day — e.g. filtering for a given date could 
return the previous day's rows.
   
   The root cause is an asymmetry: result timestamps are shifted by `+offset` 
hours for display (via `normalize_df` / `DateColumn`), but the time-range 
filter was still comparing against the raw stored values. So the data you saw 
was offset-adjusted while the filter wasn't, and the two disagreed by the 
offset.
   
   The fix shifts the time-filter bounds by `-offset` hours in 
`get_time_filter`, so the filter is evaluated in the same offset-adjusted space 
that's displayed. This covers the dashboard/explore time-range filter path. 
(The AG Grid column-level date filter builds its own SQL and isn't part of this 
change — I've left a note to follow that up separately.)
   
   ### TESTING INSTRUCTIONS
   
   1. Edit a dataset's **Settings** and set **Hour Offset** to a non-zero value 
(e.g. `-4`).
   2. Build a chart on that dataset with a visible date column and apply a 
date-range filter.
   3. The rows returned should match the selected calendar day after the offset 
is applied, instead of being shifted by a day.
   
   Unit tests added (no offset → raw bounds; positive offset → bounds shifted 
back; negative offset → bounds shifted forward). Existing time-filter tests 
still pass.
   
   ### ADDITIONAL INFORMATION
   
   - [ ] Changes UI
   
   This affects any dataset configured with a non-zero hour offset.
   


-- 
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]

Reply via email to