github-actions[bot] commented on code in PR #69163:
URL: https://github.com/apache/doris/pull/69163#discussion_r4309907512
##########
fe/be-java-extensions/hadoop-hudi-scanner/src/main/java/org/apache/doris/hudi/HadoopHudiJniScanner.java:
##########
@@ -111,6 +112,7 @@ public HadoopHudiJniScanner(int fetchSize, Map<String,
String> params) {
this.hudiColumnNames = params.get("hudi_column_names");
this.hudiColumnTypes = params.get("hudi_column_types").split("#");
+ this.requiredTypeNames = params.get("columns_types").split("#");
Review Comment:
[P2] Supply requested types for the legacy Hudi caller
With `force_jni_scanner=true` and `enable_file_scanner_v2=false`, an
ordinary `SELECT id FROM hudi_mor_table` still uses
`be/src/format/table/hudi_jni_reader.cpp`. That caller supplies
`hudi_column_types` and `required_fields`, but no `columns_types`;
`JniConnector` forwards the map unchanged apart from `time_zone`. This new
dereference therefore throws `NullPointerException` during scanner
construction, before reading even an INT-only table. Populate the requested
types in that sender, or preserve initialization for callers that omit the
parameter. The new constructor tests always supply the key and miss this
existing query path.
--
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]
