viirya commented on a change in pull request #32338:
URL: https://github.com/apache/spark/pull/32338#discussion_r619865944
##########
File path:
sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeWithFieldsSuite.scala
##########
@@ -99,7 +99,7 @@ class OptimizeWithFieldsSuite extends PlanTest {
val correctAnswer = testRelation
.select(
Alias(UpdateFields('a, WithField("b1", Literal(5)) :: Nil), "out1")(),
- Alias(UpdateFields('a, WithField("B1", Literal(5)) :: Nil), "out2")())
+ Alias(UpdateFields('a, WithField("b1", Literal(5)) :: Nil), "out2")())
Review comment:
As this is for case-insensitive, seems no big deal. Although for the
semantics, the "B1" is specified later, so I guess it is more reasonable to
keep later one.
##########
File path:
sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeWithFieldsSuite.scala
##########
@@ -126,4 +126,25 @@ class OptimizeWithFieldsSuite extends PlanTest {
comparePlans(optimized, correctAnswer)
}
}
+
+ test("SPARK-35213: ensure optimize WithFields maintains correct struct
ordering") {
+ val originalQuery = testRelation
+ .select(
+ Alias(UpdateFields('a,
+ WithField("a1", Literal(3)) ::
+ WithField("b1", Literal(4)) ::
+ WithField("a1", Literal(5)) ::
+ Nil), "out")())
+
+ val optimized = Optimize.execute(originalQuery.analyze)
+ val correctAnswer = testRelation
+ .select(
+ Alias(UpdateFields('a,
+ WithField("a1", Literal(5)) ::
+ WithField("b1", Literal(4)) ::
+ Nil), "out")())
+ .analyze
+
+ comparePlans(optimized, correctAnswer)
+ }
Review comment:
If you check the output data type, you can see the struct type is not
different:
```
optimized: ArrayBuffer(StructType(StructField(a1,IntegerType,false),
StructField(b1,IntegerType,false)))
correctAnswer: ArrayBuffer(StructType(StructField(a1,IntegerType,false),
StructField(b1,IntegerType,false)))
```
By design, `UpdateFields` will keep the order of fields in struct expression.
##########
File path:
sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeWithFieldsSuite.scala
##########
@@ -126,4 +126,25 @@ class OptimizeWithFieldsSuite extends PlanTest {
comparePlans(optimized, correctAnswer)
}
}
+
+ test("SPARK-35213: ensure optimize WithFields maintains correct struct
ordering") {
+ val originalQuery = testRelation
+ .select(
+ Alias(UpdateFields('a,
+ WithField("a1", Literal(3)) ::
+ WithField("b1", Literal(4)) ::
+ WithField("a1", Literal(5)) ::
+ Nil), "out")())
+
+ val optimized = Optimize.execute(originalQuery.analyze)
+ val correctAnswer = testRelation
+ .select(
+ Alias(UpdateFields('a,
+ WithField("a1", Literal(5)) ::
+ WithField("b1", Literal(4)) ::
+ Nil), "out")())
+ .analyze
+
+ comparePlans(optimized, correctAnswer)
+ }
Review comment:
But yea, it looks better to keep original `WithField` order.
##########
File path:
sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/OptimizeWithFieldsSuite.scala
##########
@@ -126,4 +126,25 @@ class OptimizeWithFieldsSuite extends PlanTest {
comparePlans(optimized, correctAnswer)
}
}
+
+ test("SPARK-35213: ensure optimize WithFields maintains correct struct
ordering") {
Review comment:
struct ordering -> withfield ordering
##########
File path:
sql/core/src/test/scala/org/apache/spark/sql/ColumnExpressionSuite.scala
##########
@@ -1686,6 +1686,61 @@ class ColumnExpressionSuite extends QueryTest with
SharedSparkSession {
StructType(Seq(StructField("a", IntegerType, nullable = true))))
}
+ test("SPARK-35213: chained withField operations should have correct schema
for new columns") {
+ val df = spark.createDataFrame(
+ sparkContext.parallelize(Row(null) :: Nil),
+ StructType(Seq(StructField("data", NullType))))
+
+ checkAnswer(
+ df.withColumn("data", struct()
+ .withField("a", struct())
+ .withField("b", struct())
+ .withField("a.aa", lit("aa1"))
+ .withField("b.ba", lit("ba1"))
+ .withField("a.ab", lit("ab1"))),
+ Row(Row(Row("aa1", "ab1"), Row("ba1"))) :: Nil,
+ StructType(Seq(
+ StructField("data", StructType(Seq(
+ StructField("a", StructType(Seq(
+ StructField("aa", StringType, nullable = false),
+ StructField("ab", StringType, nullable = false)
+ )), nullable = false),
+ StructField("b", StructType(Seq(
+ StructField("ba", StringType, nullable = false)
+ )), nullable = false)
+ )), nullable = false)
Review comment:
nit: Using ddl might be more readable?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
