diff --git a/sql/catalyst/src/main/antlr4/org/apache/spark/sql/catalyst/parser/SqlBase.g4 b/sql/catalyst/src/main/antlr4/org/apache/spark/sql/catalyst/parser/SqlBase.g4
index 94560317365282397f8121de6afe3b9286e7dcaf..7ffa150096333f7269608a9548ce8b5192667cc6 100644
--- a/sql/catalyst/src/main/antlr4/org/apache/spark/sql/catalyst/parser/SqlBase.g4
+++ b/sql/catalyst/src/main/antlr4/org/apache/spark/sql/catalyst/parser/SqlBase.g4
@@ -561,6 +561,7 @@ primaryExpression
| CASE whenClause+ (ELSE elseExpression=expression)? END #searchedCase
| CASE value=expression whenClause+ (ELSE elseExpression=expression)? END #simpleCase
| CAST '(' expression AS dataType ')' #cast
+ | STRUCT '(' (argument+=namedExpression (',' argument+=namedExpression)*)? ')' #struct
| FIRST '(' expression (IGNORE NULLS)? ')' #first
| LAST '(' expression (IGNORE NULLS)? ')' #last
| POSITION '(' substr=valueExpression IN str=valueExpression ')' #position
@@ -569,7 +570,7 @@ primaryExpression
| qualifiedName '.' ASTERISK #star
| '(' namedExpression (',' namedExpression)+ ')' #rowConstructor
| '(' query ')' #subqueryExpression
- | qualifiedName '(' (setQuantifier? namedExpression (',' namedExpression)*)? ')'
+ | qualifiedName '(' (setQuantifier? argument+=expression (',' argument+=expression)*)? ')'
(OVER windowSpec)? #functionCall
| value=primaryExpression '[' index=valueExpression ']' #subscript
| identifier #columnReference
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/dsl/package.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/dsl/package.scala
index f6792569b704ecf426fb15a8bc38c0324b9bcb2a..7c100afcd738fe19031e6894fbfe757fd6c25a27 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/dsl/package.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/dsl/package.scala
@@ -170,6 +170,7 @@ package object dsl {
case Seq() => UnresolvedStar(None)
case target => UnresolvedStar(Option(target))
}
+ def namedStruct(e: Expression*): Expression = CreateNamedStruct(e)
def callFunction[T, U](
func: T => U,
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala
index ef79cbcaa0ce65539658a8f31f02192bf2e0760a..8eac3ef2d35682f73ae965724f094850db783eb1 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala
@@ -1061,6 +1061,13 @@ class AstBuilder(conf: SQLConf) extends SqlBaseBaseVisitor[AnyRef] with Logging
Cast(expression(ctx.expression), visitSparkDataType(ctx.dataType))
}
+ /**
+ * Create a [[CreateStruct]] expression.
+ */
+ override def visitStruct(ctx: StructContext): Expression = withOrigin(ctx) {
+ CreateStruct(ctx.argument.asScala.map(expression))
+ }
+
/**
* Create a [[First]] expression.
*/
@@ -1091,7 +1098,7 @@ class AstBuilder(conf: SQLConf) extends SqlBaseBaseVisitor[AnyRef] with Logging
// Create the function call.
val name = ctx.qualifiedName.getText
val isDistinct = Option(ctx.setQuantifier()).exists(_.DISTINCT != null)
- val arguments = ctx.namedExpression().asScala.map(expression) match {
+ val arguments = ctx.argument.asScala.map(expression) match {
case Seq(UnresolvedStar(None))
if name.toLowerCase(Locale.ROOT) == "count" && !isDistinct =>
// Transform COUNT(*) into COUNT(1).
diff --git a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/ExpressionParserSuite.scala b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/ExpressionParserSuite.scala
index 4d08f016a4a1685924b7759bd1fc665858ba3d27..45f9f72dccc4517b92d538a060f58462386d0f30 100644
--- a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/ExpressionParserSuite.scala
+++ b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/ExpressionParserSuite.scala
@@ -231,7 +231,7 @@ class ExpressionParserSuite extends PlanTest {
assertEqual("foo(distinct a, b)", 'foo.distinctFunction('a, 'b))
assertEqual("grouping(distinct a, b)", 'grouping.distinctFunction('a, 'b))
assertEqual("`select`(all a, b)", 'select.function('a, 'b))
- assertEqual("foo(a as x, b as e)", 'foo.function('a as 'x, 'b as 'e))
+ intercept("foo(a x)", "extraneous input 'x'")
}
test("window function expressions") {
@@ -330,7 +330,9 @@ class ExpressionParserSuite extends PlanTest {
assertEqual("a.b", UnresolvedAttribute("a.b"))
assertEqual("`select`.b", UnresolvedAttribute("select.b"))
assertEqual("(a + b).b", ('a + 'b).getField("b")) // This will fail analysis.
- assertEqual("struct(a, b).b", 'struct.function('a, 'b).getField("b"))
+ assertEqual(
+ "struct(a, b).b",
+ namedStruct(NamePlaceholder, 'a, NamePlaceholder, 'b).getField("b"))
}
test("reference") {
diff --git a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/PlanParserSuite.scala b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/PlanParserSuite.scala
index bf15b85d5b510247adc2ada4ac1406dcbf3eedc8..5b2573fa4d60108274d300b5effb6e03eed56ebb 100644
--- a/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/PlanParserSuite.scala
+++ b/sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/PlanParserSuite.scala
@@ -223,6 +223,12 @@ class PlanParserSuite extends AnalysisTest {
assertEqual(s"$sql grouping sets((a, b), (a), ())",
GroupingSets(Seq(Seq('a, 'b), Seq('a), Seq()), Seq('a, 'b), table("d"),
Seq('a, 'b, 'sum.function('c).as("c"))))
+
+ val m = intercept[ParseException] {
+ parsePlan("SELECT a, b, count(distinct a, distinct b) as c FROM d GROUP BY a, b")
+ }.getMessage
+ assert(m.contains("extraneous input 'b'"))
+
}
test("limit") {
diff --git a/sql/core/src/test/resources/sql-tests/inputs/struct.sql b/sql/core/src/test/resources/sql-tests/inputs/struct.sql
index e56344dc4de80178abedbf2cdeabf60b603ec77d..93a1238ab18c280a7afc273d0da1d479aa21f6f4 100644
--- a/sql/core/src/test/resources/sql-tests/inputs/struct.sql
+++ b/sql/core/src/test/resources/sql-tests/inputs/struct.sql
@@ -18,3 +18,10 @@ SELECT ID, STRUCT(ST.*,CAST(ID AS STRING) AS E) NST FROM tbl_x;
-- Prepend a column to a struct
SELECT ID, STRUCT(CAST(ID AS STRING) AS AA, ST.*) NST FROM tbl_x;
+
+-- Select a column from a struct
+SELECT ID, STRUCT(ST.*).C NST FROM tbl_x;
+SELECT ID, STRUCT(ST.C, ST.D).D NST FROM tbl_x;
+
+-- Select an alias from a struct
+SELECT ID, STRUCT(ST.C as STC, ST.D as STD).STD FROM tbl_x;
\ No newline at end of file
diff --git a/sql/core/src/test/resources/sql-tests/results/struct.sql.out b/sql/core/src/test/resources/sql-tests/results/struct.sql.out
index 3e32f46195464d584885b7ac8e7ad1342cff476e..1da33bc736f0b55e75ac55cfefc505b582cbbf50 100644
--- a/sql/core/src/test/resources/sql-tests/results/struct.sql.out
+++ b/sql/core/src/test/resources/sql-tests/results/struct.sql.out
@@ -1,5 +1,5 @@
-- Automatically generated by SQLQueryTestSuite
--- Number of queries: 6
+-- Number of queries: 9
-- !query 0
@@ -58,3 +58,33 @@ struct<ID:int,NST:struct<AA:string,C:string,D:string>>
1 {"AA":"1","C":"gamma","D":"delta"}
2 {"AA":"2","C":"epsilon","D":"eta"}
3 {"AA":"3","C":"theta","D":"iota"}
+
+
+-- !query 6
+SELECT ID, STRUCT(ST.*).C NST FROM tbl_x
+-- !query 6 schema
+struct<ID:int,NST:string>
+-- !query 6 output
+1 gamma
+2 epsilon
+3 theta
+
+
+-- !query 7
+SELECT ID, STRUCT(ST.C, ST.D).D NST FROM tbl_x
+-- !query 7 schema
+struct<ID:int,NST:string>
+-- !query 7 output
+1 delta
+2 eta
+3 iota
+
+
+-- !query 8
+SELECT ID, STRUCT(ST.C as STC, ST.D as STD).STD FROM tbl_x
+-- !query 8 schema
+struct<ID:int,named_struct(STC, ST.C AS `C` AS `STC`, STD, ST.D AS `D` AS `STD`).STD:string>
+-- !query 8 output
+1 delta
+2 eta
+3 iota