danny0405 commented on code in PR #17580:
URL: https://github.com/apache/hudi/pull/17580#discussion_r2618430634
##########
hudi-flink-datasource/hudi-flink/src/main/java/org/apache/hudi/source/prune/ColumnStatsProbe.java:
##########
@@ -142,4 +149,12 @@ private static Object getValAsJavaObj(RowData indexRow,
int pos, LogicalType col
throw new UnsupportedOperationException("Unsupported type: " +
colType);
}
}
+
+ /**
+ * Only the following metadata columns have column stats
index:_hoodie_commit_time, _hoodie_record_key, _hoodie_partition_path
+ */
+ private static boolean isPrunableFilter(ResolvedExpression expression) {
+ String[] refs =
ExpressionUtils.referencedColumns(Collections.singletonList(expression));
Review Comment:
For regular data columns that didn't build col_stats index, the query will
skip the whole data skipping if the filter contains such columns, there is no
need to differentiate the metadata columns here, let's remove this and add a
TODO for the general fix of the missing col_stats columns.
--
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]
