aokolnychyi commented on code in PR #2276:
URL: https://github.com/apache/iceberg/pull/2276#discussion_r1000030700
##########
core/src/main/java/org/apache/iceberg/util/TableScanUtil.java:
##########
@@ -87,11 +105,44 @@ public static CloseableIterable<CombinedScanTask>
planTasks(
+
file.deletes().stream().mapToLong(ContentFile::fileSizeInBytes).sum(),
(1 + file.deletes().size()) * openFileCost);
- return CloseableIterable.transform(
- CloseableIterable.combine(
- new BinPacking.PackingIterable<>(splitFiles, splitSize, lookback,
weightFunc, true),
- splitFiles),
- BaseCombinedScanTask::new);
+ if (preservedPartitionIds != null && !preservedPartitionIds.isEmpty()) {
+ Preconditions.checkArgument(
+ spec != null, "spec can't be null when " + "preservedPartitionIds is
not null");
+ StructProjection projectedStruct =
+ StructProjection.create(spec.partitionType(), preservedPartitionIds);
+ Types.StructType projectedPartitionType = projectedStruct.type();
+ ListMultimap<StructLikeWrapper, FileScanTask> groupedFiles =
+ Multimaps.newListMultimap(Maps.newHashMap(), Lists::newArrayList);
+
+ splitFiles.forEach(
+ f -> {
+ StructLikeWrapper wrapper =
+ StructLikeWrapper.forType(projectedPartitionType)
+ .set(projectedStruct.copy().wrap(f.file().partition()));
+ groupedFiles.put(wrapper, f);
+ });
+
+ List<Iterable<BaseCombinedScanTask>> groupedTasks =
+ groupedFiles.asMap().values().stream()
+ .map(
+ t ->
+ Iterables.transform(
+ new BinPacking.PackingIterable<>(
+ CloseableIterable.withNoopClose(t),
Review Comment:
Suppose we preserve boundaries for `p1` but our partition spec is `p1`,
`p2`, `p3`. It may be beneficial to combine files with the exact partition
`p1`, `p2`, `p3` rather than simply with the same `p1`.
--
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: issues-unsubscr...@iceberg.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@iceberg.apache.org
For additional commands, e-mail: issues-h...@iceberg.apache.org
