saitharun15 commented on code in PR #11615:
URL: https://github.com/apache/iceberg/pull/11615#discussion_r1854296151
##########
spark/v3.5/spark/src/main/java/org/apache/iceberg/spark/source/SparkScan.java:
##########
@@ -194,10 +205,40 @@ protected Statistics estimateStatistics(Snapshot
snapshot) {
Map<NamedReference, ColumnStatistics> colStatsMap = Collections.emptyMap();
if (readConf.reportColumnStats() && cboEnabled) {
colStatsMap = Maps.newHashMap();
+ Map<Integer, Long> ndvs = Maps.newHashMap();
+ Map<Integer, Long> nullCounts = Maps.newHashMap();
+ Map<Integer, Object> minValues = Maps.newHashMap();
+ Map<Integer, Object> maxValues = Maps.newHashMap();
List<StatisticsFile> files = table.statisticsFiles();
if (!files.isEmpty()) {
List<BlobMetadata> metadataList = (files.get(0)).blobMetadata();
+ if (readConf.deriveStatsFromManifestSessionConf()
+ || readConf.deriveStatsFromManifestTableProperty()) {
+ Map<String, Map<Integer, Long>> distinctDataFilesNullCount =
Maps.newHashMap();
Review Comment:
We observed a scenario where different FileScanTask objects in the taskGroup
were pointing to the same data file, causing duplicates. I used putIfAbsent to
ensure a single entry per file and renamed the maps to distinctDataFiles.
However after switching to collect FileScanTask objects as a set, putIfAbsent
is no longer needed. @RussellSpitzer , can you please suggest if there is an
alternate approach?
--
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]
