ODIn/spark-instrumented-optimizer
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

[SPARK-14001][SQL] support multi-children Union in SQLBuilder

Browse source 
## What changes were proposed in this pull request?

The fix is simple, use the existing `CombineUnions` rule to combine adjacent Unions before build SQL string.

## How was this patch tested?

The re-enabled test

Author: Wenchen Fan <wenchen@databricks.com>

Closes #11818 from cloud-fan/bug-fix.
This commit is contained in:
Wenchen Fan 2016-03-18 19:42:33 +08:00 committed by Cheng Lian
parent 7783b6f38f
commit 0f1015ffdd
2 changed files with 11 additions and 10 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

14
sql/hive/src/main/scala/org/apache/spark/sql/hive/SQLBuilder.scala
View file

@ -24,9 +24,8 @@ import scala.util.control.NonFatal
import org.apache.spark.internal.Logging
import org.apache.spark.sql.{DataFrame, SQLContext}
import org.apache.spark.sql.catalyst.TableIdentifier
import org.apache.spark.sql.catalyst.analysis.EliminateSubqueryAliases
import org.apache.spark.sql.catalyst.expressions._
import org.apache.spark.sql.catalyst.optimizer.CollapseProject
import org.apache.spark.sql.catalyst.optimizer.{CollapseProject, CombineUnions}
import org.apache.spark.sql.catalyst.plans.logical._
import org.apache.spark.sql.catalyst.rules.{Rule, RuleExecutor}
import org.apache.spark.sql.catalyst.util.quoteIdentifier
@ -384,11 +383,18 @@ class SQLBuilder(logicalPlan: LogicalPlan, sqlContext: SQLContext) extends Loggi
object Canonicalizer extends RuleExecutor[LogicalPlan] {
override protected def batches: Seq[Batch] = Seq(
Batch("Collapse Project", FixedPoint(100),
Batch("Prepare", FixedPoint(100),
// The `WidenSetOperationTypes` analysis rule may introduce extra `Project`s over
// `Aggregate`s to perform type casting. This rule merges these `Project`s into
// `Aggregate`s.
CollapseProject),
CollapseProject,
// Parser is unable to parse the following query:
// SELECT `u_1`.`id`
// FROM (((SELECT `t0`.`id` FROM `default`.`t0`)
// UNION ALL (SELECT `t0`.`id` FROM `default`.`t0`))
// UNION ALL (SELECT `t0`.`id` FROM `default`.`t0`)) AS u_1
// This rule combine adjacent Unions together so we can generate flat UNION ALL SQL string.
CombineUnions),
Batch("Recover Scoping Info", Once,
// A logical plan is allowed to have same-name outputs with different qualifiers(e.g. the
// `Join` operator). However, this kind of plan can't be put under a sub query as we will

7
sql/hive/src/test/scala/org/apache/spark/sql/hive/LogicalPlanToSQLSuite.scala
View file

@ -141,12 +141,7 @@ class LogicalPlanToSQLSuite extends SQLBuilderTest with SQLTestUtils {
checkHiveQl("SELECT * FROM t0 UNION SELECT * FROM t0")
}
// Parser is unable to parse the following query:
// SELECT `u_1`.`id`
// FROM (((SELECT `t0`.`id` FROM `default`.`t0`)
// UNION ALL (SELECT `t0`.`id` FROM `default`.`t0`))
// UNION ALL (SELECT `t0`.`id` FROM `default`.`t0`)) AS u_1
ignore("three-child union") {
test("three-child union") {
checkHiveQl(
"""
|SELECT id FROM parquet_t0