Github user xuchuanyin commented on the issue:
https://github.com/apache/carbondata/pull/2781
This PR is supposed to optimize the pruning for InExpression.
For a expression like colA in (1, 2, 3).
Previously it will create 3 bloom query model and for each query model it
will iterate all the bloomfilters; But now we only need to generate 1 bloom
query model.
Also this PR does an intersection for the pruned result generated by each
expression. This optimization is important if we have multiple index columns
and used them in one query.
---
