diff options
author | gatorsmile <gatorsmile@gmail.com> | 2016-12-02 22:12:19 +0800 |
---|---|---|
committer | Wenchen Fan <wenchen@databricks.com> | 2016-12-02 22:12:19 +0800 |
commit | 2f8776ccad532fbed17381ff97d302007918b8d8 (patch) | |
tree | 9f233b8e1990508dc038bda385dc379a0f79c4ff | |
parent | 7935c8470c5c162ef7213e394fe8588e5dd42ca2 (diff) | |
download | spark-2f8776ccad532fbed17381ff97d302007918b8d8.tar.gz spark-2f8776ccad532fbed17381ff97d302007918b8d8.tar.bz2 spark-2f8776ccad532fbed17381ff97d302007918b8d8.zip |
[SPARK-18674][SQL][FOLLOW-UP] improve the error message of using join
### What changes were proposed in this pull request?
Added a test case for using joins with nested fields.
### How was this patch tested?
N/A
Author: gatorsmile <gatorsmile@gmail.com>
Closes #16110 from gatorsmile/followup-18674.
2 files changed, 17 insertions, 7 deletions
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala index 931e6599f8..8faf0eda54 100644 --- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala +++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala @@ -1974,14 +1974,14 @@ class Analyzer( condition: Option[Expression]) = { val leftKeys = joinNames.map { keyName => left.output.find(attr => resolver(attr.name, keyName)).getOrElse { - throw new AnalysisException(s"USING column `$keyName` can not be resolved with the " + - s"left join side, the left output is: [${left.output.map(_.name).mkString(", ")}]") + throw new AnalysisException(s"USING column `$keyName` cannot be resolved on the left " + + s"side of the join. The left-side columns: [${left.output.map(_.name).mkString(", ")}]") } } val rightKeys = joinNames.map { keyName => right.output.find(attr => resolver(attr.name, keyName)).getOrElse { - throw new AnalysisException(s"USING column `$keyName` can not be resolved with the " + - s"right join side, the right output is: [${right.output.map(_.name).mkString(", ")}]") + throw new AnalysisException(s"USING column `$keyName` cannot be resolved on the right " + + s"side of the join. The right-side columns: [${right.output.map(_.name).mkString(", ")}]") } } val joinPairs = leftKeys.zip(rightKeys) diff --git a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/analysis/ResolveNaturalJoinSuite.scala b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/analysis/ResolveNaturalJoinSuite.scala index 1421d36fdb..e449b9669c 100644 --- a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/analysis/ResolveNaturalJoinSuite.scala +++ b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/analysis/ResolveNaturalJoinSuite.scala @@ -28,6 +28,7 @@ class ResolveNaturalJoinSuite extends AnalysisTest { lazy val a = 'a.string lazy val b = 'b.string lazy val c = 'c.string + lazy val d = 'd.struct('f1.int, 'f2.long) lazy val aNotNull = a.notNull lazy val bNotNull = b.notNull lazy val cNotNull = c.notNull @@ -35,6 +36,8 @@ class ResolveNaturalJoinSuite extends AnalysisTest { lazy val r2 = LocalRelation(c, a) lazy val r3 = LocalRelation(aNotNull, bNotNull) lazy val r4 = LocalRelation(cNotNull, bNotNull) + lazy val r5 = LocalRelation(d) + lazy val r6 = LocalRelation(d) test("natural/using inner join") { val naturalPlan = r1.join(r2, NaturalJoin(Inner), None) @@ -108,10 +111,10 @@ class ResolveNaturalJoinSuite extends AnalysisTest { test("using unresolved attribute") { assertAnalysisError( r1.join(r2, UsingJoin(Inner, Seq("d"))), - "USING column `d` can not be resolved with the left join side" :: Nil) + "USING column `d` cannot be resolved on the left side of the join" :: Nil) assertAnalysisError( r1.join(r2, UsingJoin(Inner, Seq("b"))), - "USING column `b` can not be resolved with the right join side" :: Nil) + "USING column `b` cannot be resolved on the right side of the join" :: Nil) } test("using join with a case sensitive analyzer") { @@ -122,7 +125,14 @@ class ResolveNaturalJoinSuite extends AnalysisTest { assertAnalysisError( r1.join(r2, UsingJoin(Inner, Seq("A"))), - "USING column `A` can not be resolved with the left join side" :: Nil) + "USING column `A` cannot be resolved on the left side of the join" :: Nil) + } + + test("using join on nested fields") { + assertAnalysisError( + r5.join(r6, UsingJoin(Inner, Seq("d.f1"))), + "USING column `d.f1` cannot be resolved on the left side of the join. " + + "The left-side columns: [d]" :: Nil) } test("using join with a case insensitive analyzer") { |