geoHeil commented on issue #27315:
URL: https://github.com/apache/datafusion/issues/27315#issuecomment-6779350270

   The vendored fix and worker-decode tests are in 
https://github.com/telekom/sutura/pull/1867.
   
   A hosted replay now checks the unchanged 21-byte input against both 
versions: 
https://github.com/telekom/sutura/actions/runs/42883840932/job/130066849710. 
With the original constructors, the input panics at the same first-element 
read. With the fix, the input returns without a panic, and every retained 
peer-stage seed passes. The input SHA256 is checked before and after: 
`13c0181b8c2a4fd334aef37df4d9696c00eeec9d300c7facf089d85c392ff14d`.
   
   The fix uses `window_expr.first().ok_or_else(...)` in each constructor and 
returns `DataFusionError::Plan` when the list is empty. The bounded constructor 
reuses that checked expression for its partition properties. This replay is 
limited to retained inputs; it is not a new fuzz search or a complete 
malformed-plan audit.
   


-- 
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