[
https://issues.apache.org/jira/browse/DRILL-5796?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16515847#comment-16515847
]
ASF GitHub Bot commented on DRILL-5796:
---------------------------------------
jbimbert commented on a change in pull request #1298: DRILL-5796: Filter
pruning for multi rowgroup parquet file
URL: https://github.com/apache/drill/pull/1298#discussion_r196114329
##########
File path:
exec/java-exec/src/main/java/org/apache/drill/exec/expr/stat/ParquetBooleanPredicate.java
##########
@@ -66,25 +83,38 @@ public boolean canDrop(RangeExprEvaluator<C> evaluator) {
ExpressionPosition pos
) {
return new ParquetBooleanPredicate<C>(name, args, pos) {
+ /**
+ * Evaluates a compound "OR" filter on the statistics of a RowGroup (the
filter reads "filterA or filterB").
+ * Return value :<ul>
+ * <li>NONE : only if all filters return NONE
+ * <li>ALL : if one filter at least returns ALL
+ * <li>SOME : all other cases
+ * </ul>
+ */
@Override
- public boolean canDrop(RangeExprEvaluator<C> evaluator) {
+ public RowsMatch matches(RangeExprEvaluator<C> evaluator) {
+ RowsMatch childMatch;
+ RowsMatch resultMatch = RowsMatch.NONE;
for (LogicalExpression child : this) {
- // "or" : as long as one branch is NOT ok to drop, we can NOT drop
it.
- if (!(child instanceof ParquetFilterPredicate) ||
!((ParquetFilterPredicate)child).canDrop(evaluator)) {
- return false;
+ if (child instanceof ParquetFilterPredicate) {
+ childMatch = ((ParquetFilterPredicate) child).matches(evaluator);
+ if (childMatch != RowsMatch.NONE) {
+ if (childMatch == RowsMatch.ALL) {
+ return RowsMatch.ALL; // One at least is ALL => can drop
filter but not RG
+ }
+ resultMatch = childMatch;
+ }
}
}
- return true;
+ if (resultMatch != RowsMatch.NONE) {
+ return RowsMatch.SOME; // One at least was SOME or SOME => can not
drop RG nor filter
+ }
+ return RowsMatch.NONE; // All are NONE => can drop RG
}
};
}
- public static <C extends Comparable<C>> LogicalExpression
createBooleanPredicate(
- String function,
- String name,
- List<LogicalExpression> args,
- ExpressionPosition pos
- ) {
+ public static <C extends Comparable<C>> LogicalExpression
createBooleanPredicate(String function, String name, List<LogicalExpression>
args, ExpressionPosition pos) {
Review comment:
No. I have reverted the changes (???) I don't have these lines again in my
code. Maybe refresh the diffs, I just resubmit the PR with the changes you
asked.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
> Filter pruning for multi rowgroup parquet file
> ----------------------------------------------
>
> Key: DRILL-5796
> URL: https://issues.apache.org/jira/browse/DRILL-5796
> Project: Apache Drill
> Issue Type: Improvement
> Components: Storage - Parquet
> Reporter: Damien Profeta
> Assignee: Jean-Blas IMBERT
> Priority: Major
> Fix For: 1.14.0
>
>
> Today, filter pruning use the file name as the partitioning key. This means
> you can remove a partition only if the whole file is for the same partition.
> With parquet, you can prune the filter if the rowgroup make a partition of
> your dataset as the unit of work if the rowgroup not the file.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
