github-actions[bot] commented on code in PR #65760:
URL: https://github.com/apache/doris/pull/65760#discussion_r3601545324
##########
fe/fe-core/src/main/java/org/apache/doris/datasource/FileQueryScanNode.java:
##########
@@ -476,9 +476,12 @@ private TScanRangeLocations splitToScanRange(
pathPartitionKeys, partitionValuesFromPath.getIsNull());
TFileCompressType fileCompressType = getFileCompressType(fileSplit);
rangeDesc.setCompressType(fileCompressType);
- // set file format type, and the type might fall back to native format
in setScanParams
- rangeDesc.setFormatType(getFileFormatType());
setScanParams(rangeDesc, fileSplit);
+ // Connector-specific scan params may provide the actual format for
this split.
+ // Otherwise, fall back to the scan-level default initialized once in
createScanRangeLocations().
+ if (!rangeDesc.isSetFormatType()) {
Review Comment:
[P1] Preserve the format before connector-specific setup. `setScanParams`
used to receive a range already initialized from `getFileFormatType()`, and
Hudi relies on that at `HudiScanNode.java:263` and `:292`. For an MOR or
forced-JNI scan the unset value now makes Hudi populate only native schema
metadata; this fallback then labels that same descriptor `FORMAT_JNI`. The BE
JNI reader rejects it because `base_path`, `data_file_path`,
`data_file_length`, and column metadata were never set. Please seed the
scan-level default before invoking the hook, then let per-split hooks override
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]
