[SPARK-44219][SQL] Adds extra per-rule validations for optimization rewrites. #41763
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
YannisSismanis
changed the title
cloud-fan
reviewed
Jul 21, 2023
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/LogicalPlan.scala
Show resolved
Hide resolved
cloud-fan
reviewed
Jul 21, 2023
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/LogicalPlan.scala
Outdated
Show resolved
Hide resolved
cloud-fan
reviewed
Aug 24, 2023
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/LogicalPlan.scala
Outdated
Show resolved
Hide resolved
cloud-fan
reviewed
Aug 25, 2023
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/LogicalPlan.scala
Outdated
Show resolved
Hide resolved
cloud-fan
reviewed
Oct 4, 2023
| case expr: AggregateExpression => | ||
| val aggFunction = expr.aggregateFunction | ||
| aggFunction.children.foreach { | ||
| child => |
There was a problem hiding this comment.
code style nit to save one indentation level. It's also the style of the previous code: https://github.com/apache/spark/pull/41763/files#diff-583171e935b2dc349378063a5841c5b98b30a2d57ac3743a9eccfe7bffcb8f2aL432
aggFunction.children.foreach { child =>
...
}
cloud-fan
reviewed
Oct 4, 2023
Comment on lines
211
to
214
| msg = s"Non-deterministic expression '${ | ||
| toSQLExpr(expr) | ||
| }' should not appear in " + | ||
| "grouping expression.", |
There was a problem hiding this comment.
Suggested change
| msg = s"Non-deterministic expression '${ | |
| toSQLExpr(expr) | |
| }' should not appear in " + | |
| "grouping expression.", | |
| msg = s"Non-deterministic expression '${toSQLExpr(expr)}' should not appear in " + | |
| "grouping expression.", |
cloud-fan
reviewed
Oct 4, 2023
| def validateNoDanglingReferences(plan: LogicalPlan): Option[String] = { | ||
| plan.collectFirst { | ||
| // DML commands and multi instance relations (like InMemoryRelation caches) | ||
| // have different output semantics than typical queries. |
There was a problem hiding this comment.
Do we really need to special-case these two? QueryPlan#missingInputs should have already taken care of them
final def missingInput: AttributeSet = references -- inputSet
There was a problem hiding this comment.
Anyway, not a big deal, it's faster to skip some cases that will never hit missing attr issue.
cloud-fan
reviewed
Oct 4, 2023
| case _: Command => None | ||
| case _: MultiInstanceRelation => None | ||
| case n if canGetOutputAttrs(n) => | ||
| if ( n.missingInput.nonEmpty) { |
There was a problem hiding this comment.
Suggested change
| if ( n.missingInput.nonEmpty) { | |
| if (n.missingInput.nonEmpty) { |
cloud-fan
reviewed
Oct 4, 2023
| None | ||
| } catch { | ||
| case _: AnalysisException => | ||
| Some(s"Aggregate: ${a.toString} is not a valid aggregate expression") |
There was a problem hiding this comment.
shall we take the actual error message? case e: AnalysisException => Some(e.getMessage)
cloud-fan
reviewed
Oct 4, 2023
| .orElse(LogicalPlanIntegrity.validateNoDanglingReferences(currentPlan)) | ||
| .orElse(LogicalPlanIntegrity.validateGroupByTypes(currentPlan)) | ||
| .orElse(LogicalPlanIntegrity.validateAggregateExpressions(currentPlan)) | ||
| .map( err => s"${err}\nPrevious schema:${previousPlan.output.mkString(", ")}" + |
There was a problem hiding this comment.
Suggested change
| .map( err => s"${err}\nPrevious schema:${previousPlan.output.mkString(", ")}" + | |
| .map(err => s"${err}\nPrevious schema:${previousPlan.output.mkString(", ")}" + |
cloud-fan
approved these changes
Oct 4, 2023
|
thanks, merging to master! |
LuciferYang
pushed a commit
to LuciferYang/spark
that referenced
this pull request
Oct 7, 2023
…ewrites ### What changes were proposed in this pull request? Adds per-rule validation checks for the following: 1. aggregate expressions in Aggregate plans are valid. 2. Grouping key types in Aggregate plans cannot by of type Map. 3. No dangling references have been generated. This validation is by default enabled for all tests or selectively using the spark.sql.planChangeValidation=true flag. ### Why are the changes needed? Extra validation for optimizer rewrites. ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? Unit tests Closes apache#41763 from YannisSismanis/SC-130139_followup. Authored-by: Yannis Sismanis <yannis.sismanis@databricks.com> Signed-off-by: Wenchen Fan <wenchen@databricks.com>
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.
What changes were proposed in this pull request?
Adds per-rule validation checks for the following:
This validation is by default enabled for all tests or selectively using the spark.sql.planChangeValidation=true flag.
Why are the changes needed?
Extra validation for optimizer rewrites.
Does this PR introduce any user-facing change?
No
How was this patch tested?
Unit tests