Skip to content

[WIP][Spark-SQL] Optimize the Constant Folding for Expression #482

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
chenghao-intel wants to merge 13 commits into from
Closed

[WIP][Spark-SQL] Optimize the Constant Folding for Expression #482

Changes from 1 commit
Commits
Show all changes
13 commits
Select commit Hold shift + click to select a range
2645d4f
Constant Folding(null propagation)
chenghao-intel Apr 28, 2014
3c045c7
Optimize the Constant Folding by adding more rules
chenghao-intel Apr 28, 2014
536c005
Add Exceptional case for constant folding
chenghao-intel Apr 28, 2014
9cf0396
update code according to the code review comment
chenghao-intel Apr 28, 2014
543ef9d
fix code style issues
chenghao-intel Apr 29, 2014
9ccefdb
Add tests for optimized expression evaluation.
marmbrus Apr 29, 2014
b28e03a
Merge pull request #1 from marmbrus/pr/482
chenghao-intel Apr 30, 2014
27ea3d7
Fix Constant Folding Bugs & Add More Unittests
chenghao-intel Apr 30, 2014
80f9f18
Update the UnitTest for aggregation constant folding
chenghao-intel Apr 30, 2014
50444cc
Remove the unnecessary null checking
chenghao-intel Apr 30, 2014
29c8166
Update the code for feedback of code review
chenghao-intel May 5, 2014
68b9fad
Remove the Literal pattern matching for NullPropagation
chenghao-intel May 5, 2014
2f14b50
Fix code style issues
chenghao-intel May 6, 2014
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Remove the Literal pattern matching for NullPropagation
  • Loading branch information
@chenghao-intel
chenghao-intel committed May 5, 2014
commit 68b9fad66b499e667c827a1a0bb1ef8dccaa813e
2 changes: 0 additions & 2 deletions sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,8 +94,6 @@ object ColumnPruning extends Rule[LogicalPlan] {
object NullPropagation extends Rule[LogicalPlan] {
def apply(plan: LogicalPlan): LogicalPlan = plan transform {
case q: LogicalPlan => q transformExpressionsUp {
Copy link
Contributor

@marmbrus marmbrus Apr 28, 2014

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we want Up? I believe this means we are going to call evaluate on each foldable node working up instead of just calling it once at the top.

Copy link
Contributor Author

@chenghao-intel chenghao-intel Apr 28, 2014

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

expression.foldable is ok by traveling from top to bottom, while null propagation is opposite. I've put them into different rule objects (ConstantFolding & NullPropagation).

// Skip redundant folding of literals.
case l: Literal => l
case e @ Count(Literal(null, _)) => Literal(0, e.dataType)
case e @ Sum(Literal(c, _)) if(c == 0) => Literal(0, e.dataType)
Copy link
Contributor Author

@chenghao-intel chenghao-intel Apr 30, 2014

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not like the previous implementation, I didn't add the null value as Child expression for both Sum and Average in the rule, cause Hive will throw semantic exception.

case e @ Average(Literal(c, _)) if(c == 0) => Literal(0.0, e.dataType)
Expand Down