[SPARK-17357][SPARK-6624][SQL] Convert filter predicate to CNF in Optimizer for pushdown #15558
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
This PR is proposed to solve the problem #14912 tried to solve before. Simply said, currently some predicates can not be correctly pushdown through operators due to its format is a bunch of ORs.
A simple example is (a > 10) || (b > 2 && c == 3). If a datasource has attributes a and b, this filtering predicate cannot be pushdown. If we can convert it to CNF (a > 10 || b > 2) && (a > 10 || c == 3). Then we can push down (a > 10 || b > 2).
To convert the predicate to CNF format can solve this formally instead of a hacky way on #14912.
We have previous PRs for CNF conversion, such as #8200. Most of added tests in
CNFNormalizationSuite
are copied from #8200.How was this patch tested?
Jenkins tests.
Please review https://cwiki.apache.org/confluence/display/SPARK/Contributing+to+Spark before opening a pull request.