pzhdfy commented on code in PR #66227:
URL: https://github.com/apache/doris/pull/66227#discussion_r4060186276
##########
fe/fe-core/src/main/java/org/apache/doris/datasource/paimon/source/PaimonScanNode.java:
##########
@@ -411,10 +412,102 @@ private void setPaimonParams(TFileRangeDesc rangeDesc,
PaimonSplit paimonSplit)
String fileFormat = getFileFormat(paimonSplit.getPathString());
if (split != null) {
+ // use jni reader / paimon-cpp reader / paimon-rust reader
rangeDesc.setFormatType(TFileFormatType.FORMAT_JNI);
- // A logical DataSplit may span multiple files, so keep it intact
for the JNI reader.
- fileDesc.setReaderType(TPaimonReaderType.PAIMON_JNI);
- fileDesc.setPaimonSplit(PaimonUtil.encodeObjectToString(split));
+ // paimon-cpp and paimon-rust both consume Paimon native binary
serialization,
+ // which only supports DataSplit. Any other split type falls back
to JNI.
+ boolean nativeSplit = split instanceof DataSplit;
Review Comment:
Fixed — canUseRust gates fallback reads to JNI through both shapes: a
wrapped split (split instanceof FallbackReadFileStoreTable.FallbackSplit, the
public interface every FallbackDataSplit implements) and the table wrapper
itself (processedTable instanceof FallbackReadFileStoreTable — both read sides
of a fallback table produce wrapped splits, so the whole wrapper routes to JNI
until the rust ABI can represent both sides). The routing test exercises the
genuine wire shape rather than a mock: an ordinary DataSplit serialized, the
isFallback byte appended exactly like FallbackDataSplit.serialize, and the
result deserialized through the public FallbackDataSplit.deserialize factory —
asserting JNI routing for both the split and the table-wrapper shapes;
PaimonScanNodeTest 51/51.
--
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]
