[GitHub] [iceberg] aokolnychyi commented on a change in pull request #1664: Parquet: Optimize IN predicates in ParquetDictionaryRowGroupFilter

[GitBox <gi...@apache.org>]

aokolnychyi commented on a change in pull request #1664:
URL: https://github.com/apache/iceberg/pull/1664#discussion_r512213739



##########
File path: parquet/src/main/java/org/apache/iceberg/parquet/ParquetDictionaryRowGroupFilter.java
##########
@@ -278,8 +278,27 @@ public Boolean or(Boolean leftResult, Boolean rightResult) {
 
       Set<T> dictionary = dict(id, ref.comparator());
 
-      // ROWS_CANNOT_MATCH if all values of the dictionary are not in the set (the intersection is empty)
-      return Sets.intersection(dictionary, literalSet).isEmpty() ? ROWS_CANNOT_MATCH : ROWS_MIGHT_MATCH;
+      // we need to find out the smaller set to iterate through
+      Set<T> smallerSet;
+      Set<T> biggerSet;
+
+      if (literalSet.size() < dictionary.size()) {
+        smallerSet = literalSet;
+        biggerSet = dictionary;
+      } else {
+        smallerSet = dictionary;
+        biggerSet = literalSet;
+      }
+
+      for (T e : smallerSet) {

Review comment:
       We could rely on `Collections$disjoint` here knowing how it behaves but I'd prefer to have full control here and be sure this logic does not change.




----------------------------------------------------------------
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.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@iceberg.apache.org
For additional commands, e-mail: issues-help@iceberg.apache.org