Support push down sort after system limit #4657
Open
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Outdated suggestions cannot be applied.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.
Description
Support push down sort after limit if there is only limit pushed, since PPL or other database all don't promise sequence for only limit operator. Thus it should be acceptable to transform
limit + sorttosort + limit.However, we should avoid pushing down
sort, if there is existingsortbeforelimit. In such case, users intense to retrieve the Top-K values on the first sort fields from our index. The order will be overridden by the secondsortif we keep pushing down it. So this PR prevents this case by detecting whethertop-kis pushed down already.Since #4501 has always introduced a limit before sort for join or subsearch, which will block sort push down. This PR will also enhance these scenarios, especially for left join -- both sides will have limit and sort pushed down then. See
CalciteExplainIT::testExplainScalarCorrelatedSubqueryInSelectRelated Issues
Resolves #4570
Check List
--signoffor-s.By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.