fix: preserve aggregate scope when unparsing - #23327
Merged
Merged
Conversation
Phoenix500526
marked this pull request as draft
July 5, 2026 13:30
Phoenix500526
marked this pull request as ready for review
July 5, 2026 14:59
kosiew
requested changes
Jul 9, 2026
kosiew
left a comment
Contributor
There was a problem hiding this comment.
@Phoenix500526
Thanks for the fix here.
I think there is one remaining correctness gap around aggregate expression normalization in HAVING.
I also left one test suggestion that would make this regression coverage a bit stronger.
| } else if let Some(agg) = agg { | ||
| let unprojected = | ||
| unproject_agg_exprs(filter.predicate.clone(), agg, None)?; | ||
| let filter_expr = self.expr_to_sql(&unprojected)?; |
Contributor
There was a problem hiding this comment.
Nice work applying the new aggregate input normalization to SELECT and GROUP BY.
I think the same path is needed for HAVING predicates that come from a Filter over an Aggregate.
For the same plan shape being fixed here, where Aggregate reads from an unnamed derived Projection, a query like ... GROUP BY 1 HAVING count(DISTINCT cs.customer_id) > 0 still unparses to HAVING (count(DISTINCT \"cs\".\"customer_id\") > 0). The FROM item is (SELECT \"cs\".\"customer_id\", ... FROM ...), so the cs qualifier is no longer in scope.
Could we apply the same normalize_agg_input_columns path after unproject_agg_exprs here, and also in the QUALIFY branch when it unprojects aggregate expressions? That should make all expressions rendered in the aggregate SELECT resolve against the derived projection output.
| FROM | ||
| "warehouse"."main"."sales" cs | ||
| JOIN "warehouse"."main"."customers" c USING (customer_id) | ||
| GROUP BY |
Contributor
There was a problem hiding this comment.
Small suggestion for the regression coverage: it would be great if the issue_23317 tests executed, or at least parsed, the unparsed SQL rather than only checking string fragments.
The bug here is invalid SQL caused by out-of-scope aliases, so a parse or roundtrip execution assertion would be more likely to catch similar missed clauses like HAVING or ORDER BY even if the exact formatting changes later.
Contributor
Author
There was a problem hiding this comment.
Good catch. The sql has already parsed in assert_issue_23317_unparsed_sql_plans now.
The unparser can fold a Projection into its Aggregate and then keep walking into the Aggregate input. When that input is another Aggregate or an unnamed derived Projection, the generated SQL moved references above the SELECT block that defined them. Optimizer aliases such as `group_alias_0` and base-table qualifiers such as `c.signup_date` were no longer in scope. Stop crossing those scope boundaries blindly. Render nested aggregate inputs as derived relations, and when an aggregate reads from a derived Projection, rewrite only input-schema columns to the derived output names before rendering SELECT and GROUP BY expressions. CLOSES apache#23317 Signed-off-by: Jiawei Zhao <Phoenix500526@163.com>
HAVING and QUALIFY predicates can reference aggregate inputs after unprojection. Normalize those inputs when the aggregate reads from a derived projection so unparsed SQL does not leak inner aliases. Refs apache#23317
kosiew
approved these changes
Jul 11, 2026
alamb
enabled auto-merge
July 16, 2026 18:36
Contributor
|
I took a quick look and it looks reasonable to me -- than k you @Phoenix500526 and @kosiew |
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #23327 +/- ##
=======================================
Coverage ? 80.65%
=======================================
Files ? 1086
Lines ? 366165
Branches ? 366165
=======================================
Hits ? 295334
Misses ? 53238
Partials ? 17593 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
6 tasks
Omega359
pushed a commit
to Omega359/arrow-datafusion
that referenced
this pull request
Jul 18, 2026
## Which issue does this PR close? <!-- We generally require a GitHub issue to be filed for all bug fixes and enhancements and this helps us generate change logs for our releases. You can link an issue to this PR using the GitHub syntax. For example `Closes apache#123` indicates that this PR will close issue apache#123. --> - Closes apache#23317. ## Rationale for this change <!-- Why are you proposing this change? If this is already explained clearly in the issue then this section is not needed. Explaining clearly why changes are proposed helps reviewers understand your changes and offer better suggestions for fixes. --> The unparser can fold a Projection into its Aggregate and then keep walking into the Aggregate input. When that input is another Aggregate or an unnamed derived Projection, the generated SQL moved references above the SELECT block that defined them. Optimizer aliases such as `group_alias_0` and base-table qualifiers such as `c.signup_date` were no longer in scope. Stop crossing those scope boundaries blindly. Render nested aggregate inputs as derived relations, and when an aggregate reads from a derived Projection, rewrite only input-schema columns to the derived output names before rendering SELECT and GROUP BY expressions. ## What changes are included in this PR? <!-- There is no need to duplicate the description in the issue here but it is sometimes worth providing a summary of the individual changes in this PR. --> ## Are these changes tested? <!-- We typically require tests for all PRs in order to: 1. Prevent the code from being accidentally broken by subsequent changes 2. Serve as another way to document the expected behavior of the code If tests are not included in your PR, please explain why (for example, are they covered by existing tests)? --> Yes ## Are there any user-facing changes? <!-- If there are user-facing changes then we may require documentation to be updated before approving the PR. --> <!-- If there are any breaking changes to public APIs, please add the `api change` label. --> No --------- Signed-off-by: Jiawei Zhao <Phoenix500526@163.com> Co-authored-by: Andrew Lamb <andrew@nerdnetworks.org>
kosiew
pushed a commit
to kosiew/datafusion
that referenced
this pull request
Aug 12, 2026
## Which issue does this PR close? <!-- We generally require a GitHub issue to be filed for all bug fixes and enhancements and this helps us generate change logs for our releases. You can link an issue to this PR using the GitHub syntax. For example `Closes apache#123` indicates that this PR will close issue apache#123. --> - Closes apache#23317. ## Rationale for this change <!-- Why are you proposing this change? If this is already explained clearly in the issue then this section is not needed. Explaining clearly why changes are proposed helps reviewers understand your changes and offer better suggestions for fixes. --> The unparser can fold a Projection into its Aggregate and then keep walking into the Aggregate input. When that input is another Aggregate or an unnamed derived Projection, the generated SQL moved references above the SELECT block that defined them. Optimizer aliases such as `group_alias_0` and base-table qualifiers such as `c.signup_date` were no longer in scope. Stop crossing those scope boundaries blindly. Render nested aggregate inputs as derived relations, and when an aggregate reads from a derived Projection, rewrite only input-schema columns to the derived output names before rendering SELECT and GROUP BY expressions. ## What changes are included in this PR? <!-- There is no need to duplicate the description in the issue here but it is sometimes worth providing a summary of the individual changes in this PR. --> ## Are these changes tested? <!-- We typically require tests for all PRs in order to: 1. Prevent the code from being accidentally broken by subsequent changes 2. Serve as another way to document the expected behavior of the code If tests are not included in your PR, please explain why (for example, are they covered by existing tests)? --> Yes ## Are there any user-facing changes? <!-- If there are user-facing changes then we may require documentation to be updated before approving the PR. --> <!-- If there are any breaking changes to public APIs, please add the `api change` label. --> No --------- Signed-off-by: Jiawei Zhao <Phoenix500526@163.com> Co-authored-by: Andrew Lamb <andrew@nerdnetworks.org>
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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?
Rationale for this change
The unparser can fold a Projection into its Aggregate and then keep walking into the Aggregate input. When that input is another Aggregate or an unnamed derived Projection, the generated SQL moved references above the SELECT block that defined them. Optimizer aliases such as
group_alias_0and base-table qualifiers such asc.signup_datewere no longer in scope.Stop crossing those scope boundaries blindly. Render nested aggregate inputs as derived relations, and when an aggregate reads from a derived Projection, rewrite only input-schema columns to the derived output names before rendering SELECT and GROUP BY expressions.
What changes are included in this PR?
Are these changes tested?
Yes
Are there any user-facing changes?
No