Don't add filters to projection in TableScan #7670
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Dandandan
changed the title
github-actions
bot
added
logical-expr
Dandandan
commented
Sep 28, 2023
|
|
||
| let expected = "\ | ||
| Projection: Int32(1) AS a\ | ||
| \n TableScan: test projection=[a], full_filters=[b = Int32(1)]"; |
There was a problem hiding this comment.
Here is the important fix - before this would add b to the projection (even if it was not needed in the plan above). This would lead to unnecessary scanning columns supported by the filter.
Dandandan
changed the title
Dandandan
changed the title
Dandandan
commented
Sep 28, 2023
| @@ -147,10 +147,6 @@ impl OptimizerRule for PushDownProjection { | |||
| if !scan.projected_schema.fields().is_empty() => | |||
| { | |||
| let mut used_columns: HashSet<Column> = HashSet::new(); | |||
| // filter expr may not exist in expr in projection. | |||
There was a problem hiding this comment.
This added filters to the projection since #5188
FYI @jackwener
Add test WIP fix Fix filter after scan Totally reemove filter to column extraction Fix test Update tests 1 Update tests 2 Update tests 3
Dandandan
force-pushed
the
dont_add_filters_to_projection
branch
from
b283946
to
27069f0
Compare
|
FYI @alamb this might help with enabling parquet filter pushdown as well by default |
alamb
approved these changes
Sep 28, 2023
| @@ -2759,8 +2705,7 @@ Projection: a, b | |||
| // For right anti, filter of the left side can be pushed down. | |||
| let expected = "RightAnti Join: test1.a = test2.a Filter: test2.b > UInt32(2)\ | |||
| \n Projection: test1.a, test1.b\ | |||
| \n Filter: test1.b > UInt32(1)\ | |||
| \n TableScan: test1\ | |||
| \n TableScan: test1, full_filters=[test1.b > UInt32(1)]\ | |||
There was a problem hiding this comment.
it is great to see the filters pushed into the scans as part of this test
Ted-Jiang
pushed a commit
to Ted-Jiang/arrow-datafusion
that referenced
this pull request
Oct 7, 2023
Add test WIP fix Fix filter after scan Totally reemove filter to column extraction Fix test Update tests 1 Update tests 2 Update tests 3
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Which issue does this PR close?
Closes #7683
Rationale for this change
We don't want to scan columns not needed by the rest of the plan.
What changes are included in this PR?
We change the scan implementations to apply the filters before the projection.
This way we don't need to add the filters to the projection in to make the plan correct.
Are these changes tested?
Existing tests + new test.
Are there any user-facing changes?