[
https://issues.apache.org/jira/browse/DRILL-5796?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16514138#comment-16514138
]
ASF GitHub Bot commented on DRILL-5796:
---------------------------------------
vrozov 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_r195814485
##########
File path:
exec/java-exec/src/main/java/org/apache/drill/exec/expr/stat/ParquetComparisonPredicate.java
##########
@@ -83,37 +80,40 @@ private ParquetComparisonPredicate(
* where Column1 and Column2 are from same parquet table.
*/
@Override
- public boolean canDrop(RangeExprEvaluator<C> evaluator) {
+ public RowsMatch matches(RangeExprEvaluator<C> evaluator) {
Statistics<C> leftStat = left.accept(evaluator, null);
if (isNullOrEmpty(leftStat)) {
- return false;
+ return RowsMatch.SOME;
}
-
Statistics<C> rightStat = right.accept(evaluator, null);
if (isNullOrEmpty(rightStat)) {
- return false;
+ return RowsMatch.SOME;
}
-
- // if either side is ALL null, = is evaluated to UNKNOWN -> canDrop
if (isAllNulls(leftStat, evaluator.getRowCount()) || isAllNulls(rightStat,
evaluator.getRowCount())) {
- return true;
+ return RowsMatch.NONE;
}
+ return predicate.apply(leftStat, rightStat);
+ }
- return (leftStat.hasNonNullValue() && rightStat.hasNonNullValue()) &&
predicate.test(leftStat, rightStat);
+ /**
+ * If the tested rowgroup contains some null values, change the
RowsMatch.ALL into RowsMatch.SOME (null values should be discarded by filter)
+ *
+ * @param leftStat the rowgroup statistics to check
+ * @return the corresponding RowsMatch
+ */
+ static RowsMatch checkNull(Statistics leftStat) {
+ return leftStat.getNumNulls() > 0 ? RowsMatch.SOME : RowsMatch.ALL;
}
/**
* EQ (=) predicate
*/
- private static <C extends Comparable<C>> LogicalExpression
createEqualPredicate(
- LogicalExpression left,
- LogicalExpression right
- ) {
+ private static <C extends Comparable<C>> LogicalExpression
createEqualPredicate(LogicalExpression left, LogicalExpression right) {
return new ParquetComparisonPredicate<C>(left, right, (leftStat,
rightStat) -> {
- // can drop when left's max < right's min, or right's max < left's min
- final C leftMin = leftStat.genericGetMin();
- final C rightMin = rightStat.genericGetMin();
- return (leftStat.compareMaxToValue(rightMin) < 0) ||
(rightStat.compareMaxToValue(leftMin) < 0);
+ if ((leftStat.genericGetMax().compareTo(rightStat.genericGetMin()) < 0
|| rightStat.genericGetMax().compareTo(leftStat.genericGetMin()) < 0)) {
Review comment:
Use `?`.
----------------------------------------------------------------
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:
us...@infra.apache.org
> 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)
