projects / carbondata.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: ce860d0) | patch
[CARBONDATA-4243] Fixed si with column meta cache on same column
authornihal0107 <nihalnitn97@gmail.com>
Mon, 27 Sep 2021 13:06:52 +0000 (18:36 +0530)
committerIndhumathi27 <indhumathim27@gmail.com>
Thu, 7 Oct 2021 08:10:36 +0000 (13:40 +0530)
commit9944936d684f241c5e81ca809507f97d1ef82f26
tree616e4a1b298c56123fd5a813114274ba3e0b45b1tree | snapshot
parentce860d0431f41e0594cee848149cdc344f728dd3commit | diff
[CARBONDATA-4243] Fixed si with column meta cache on same column

Why is this PR needed?
Currently, the select query fails when table contains SI and column_meta_cache
on the same columns with to date() UDF. This is happening because pushdownfilters
is null in CarbonDataSourceScanHelper and it is causing null pointer exception.

What changes were proposed in this PR?
At place of passing null value for pushdownfilters in CarbonDataSourceScan.doCanonicalize passed Seq.empty.

This closes #4225
index/secondary-index/src/test/scala/org/apache/carbondata/spark/testsuite/secondaryindex/TestSIWithSecondaryIndex.scala diff | blob | history
integration/spark/src/main/scala/org/apache/spark/sql/execution/strategy/CarbonDataSourceScan.scala diff | blob | history
High performance data store solution