dhruv-15-03 opened a new pull request, #3821:
URL: https://github.com/apache/parquet-java/pull/3821

   ### Rationale for this change
   
   The binary-statistics truncation test calls a static default overload 
through a configured converter instance. The constructor's truncation length is 
ignored, and the existing ordering-only assertions let untruncated bounds pass.
   
   ### What changes are included in this PR?
   
   - Call the explicit `toParquetStatistics(stats, truncateLen)` overload.
   - Use deterministic byte arrays and check exact truncated bytes and lengths, 
including the conservative increment of the maximum.
   - Check that combined bounds of 4095 bytes are accepted and 4096 bytes are 
omitted, with both default and finite truncation.
   - Preserve invalid-length constructor checks and current null-count omission 
behavior.
   
   Only `TestParquetMetadataConverter.java` changes. The personal-fork 
validation workflow is not included.
   
   ### Are these changes tested?
   
   Yes. On the contribution commit `a9940be5cfee6962dc0f3e064b4c619a07b2d3b7` 
(Temurin 17), the full `TestParquetMetadataConverter` class ran 78 tests with 0 
failures, 0 errors and 0 skipped, including `testBinaryStatsWithTruncation` and 
`testBinaryStatsWithTruncationAtSizeLimit`. `mvn -pl parquet-hadoop 
spotless:check` reported 262 files clean. Run: 
https://github.com/dhruv-15-03/parquet-java/actions/runs/36249258685
   
   ### Are there any user-facing changes?
   
   None. Production code and output semantics are unchanged. #3819 addresses 
null-count serialization; this PR repairs a separate test-coverage gap and does 
not fix or close #3574.
   
   Closes #3820
   


-- 
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