szehon-ho commented on code in PR #4520:
URL: https://github.com/apache/iceberg/pull/4520#discussion_r846330540
##########
core/src/main/java/org/apache/iceberg/BaseFilesTable.java:
##########
@@ -92,7 +94,26 @@ public TableScan appendsAfter(long fromSnapshotId) {
protected CloseableIterable<FileScanTask> planFiles(TableOperations ops,
Snapshot snapshot, Expression rowFilter,
boolean
ignoreResiduals, boolean caseSensitive,
boolean colStats) {
- CloseableIterable<ManifestFile> filtered = filterManifests(manifests(),
rowFilter, caseSensitive);
+ Map<Integer, PartitionSpec> specsById = table().specs();
+
+ LoadingCache<Integer, ManifestEvaluator> evalCache =
Caffeine.newBuilder().build(specId -> {
+ PartitionSpec spec = specsById.get(specId);
+ PartitionSpec transformedSpec = transformSpec(fileSchema, spec,
PARTITION_FIELD_PREFIX);
+ return ManifestEvaluator.forRowFilter(rowFilter, transformedSpec,
caseSensitive);
+ });
+
+ CloseableIterable<ManifestFile> filtered = CloseableIterable.filter(
+ manifests(),
+ manifest -> {
+ PartitionSpec spec = specsById.get(manifest.partitionSpecId());
+
+ if (spec.fields().stream().anyMatch(f ->
f.transform().equals(Transforms.alwaysNull()))) {
Review Comment:
Thought: haven't looked but maybe we could potentially push this logic down
to ManifestEvaluator, return "ROWS_MIGHT_MATCH" match for any partition
comparison involving a partition field that has become VoidTransformation
(removed in V1). (I am thinking this case of querying removed partition field
should only affect metadata tables primarily).
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
