lokiore commented on code in PR #2563:
URL: https://github.com/apache/phoenix/pull/2563#discussion_r3615722943


##########
phoenix-core-client/src/main/java/org/apache/phoenix/schema/PTable.java:
##########


Review Comment:
   Good call — done in `77060fca3a`. `getSerializedValue()` now throws 
`IllegalStateException` on `UNKNOWN`, so any persist attempt fails loudly at 
the type boundary regardless of call path, instead of silently serializing 
`-1`. The message names the constant for log/stack clarity. Added a test that 
the accessor throws for `UNKNOWN`, and reworked the round-trip test to loop 
over `values()` so every known type is covered automatically as new ones are 
added.
   
   One deliberate divergence from the suggestion: I kept the existing per-site 
guards (`doCutover`, `finishForceCutover`, 
`TransformMonitorTask.skipIfUnknownTransformType`, `validateTransform`) rather 
than treating the throw as subsuming them. Those aren't persist-time checks — 
they're *graceful-degradation* gates that skip-with-warn (or fail-fast with an 
operator message) so an UNKNOWN record doesn't starve other transforms or get 
falsely marked COMPLETED. They compare `== UNKNOWN` and never reach the 
accessor, so the throw is a complementary backstop at the persist boundary 
rather than a replacement. Happy to reconsider if you'd prefer consolidating 
any of them.
   
   FYI @tkhurana
   
   _(Posted by Loki's Claude companion)_



-- 
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]

Reply via email to