cshuo commented on code in PR #20671:
URL: https://github.com/apache/hudi/pull/20671#discussion_r4594931423


##########
hudi-client/hudi-flink-client/src/main/java/org/apache/hudi/io/storage/row/HoodieRowDataFileWriterFactory.java:
##########
@@ -102,7 +102,7 @@ public HoodieFileWriter newParquetFileWriter(
     HoodieConfig hoodieConfig = injectedConfigs.getRight();
 
     Configuration conf = (Configuration) 
storageConfiguration.unwrapAs(Configuration.class);
-    BloomFilter filter = enableBloomFilter(metaFieldsMode, hoodieConfig) ? 
createBloomFilter(hoodieConfig) : null;
+    BloomFilter filter = enableParquetBloomFilter(metaFieldsMode, 
hoodieConfig, storagePath) ? createBloomFilter(hoodieConfig) : null;

Review Comment:
   The existing flag alone would not cover this: Parquet still enables bloom 
filters for BLOOM/GLOBAL_BLOOM indexes even when the flag is false, and ORC 
writer does not use that flag.
   
   Moving this decision to the native log writer would therefore require an 
additional override propagated through the writer factories. The current 
implementation keeps the fix scoped: the Parquet path check is already 
centralized in a shared helper used by Avro/Spark/Flink, with one separate 
check for ORC. This preserves the existing base-file behavior without 
introducing another writer option.



-- 
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