HyukjinKwon commented on a change in pull request #32355:
URL: https://github.com/apache/spark/pull/32355#discussion_r680435570
##########
File path: sql/core/src/test/scala/org/apache/spark/sql/JoinHintSuite.scala
##########
@@ -597,4 +597,68 @@ class JoinHintSuite extends PlanTest with
SharedSparkSession with AdaptiveSparkP
assert(df9.collect().size == df10.collect().size)
}
}
+
+ test("SPARK-35221: Add join hint build side check") {
+ withSQLConf(SQLConf.AUTO_BROADCASTJOIN_THRESHOLD.key -> "-1",
+ SQLConf.PREFER_SORTMERGEJOIN.key -> "true") {
+ Seq("left_outer", "left_semi", "left_anti").foreach { joinType =>
+ val hintAppender = new LogAppender(s"join hint build side check for
$joinType")
+ withLogAppender(hintAppender, level = Some(Level.WARN)) {
+ assertShuffleMergeJoin(
+ df1.hint("BROADCAST").join(df2, $"a1" === $"b1", joinType))
+ assertBroadcastHashJoin(
+ df1.join(df2.hint("BROADCAST"), $"a1" === $"b1", joinType),
BuildRight)
+
+ assertShuffleMergeJoin(
+ df1.hint("SHUFFLE_HASH").join(df2, $"a1" === $"b1", joinType))
+ assertShuffleHashJoin(
+ df1.join(df2.hint("SHUFFLE_HASH"), $"a1" === $"b1", joinType),
BuildRight)
+ }
+
+ val logs = hintAppender.loggingEvents.map(_.getRenderedMessage)
+ .filter(_.startsWith("A join hint"))
Review comment:
nit but I think the test here can be flaky (e.g., the test will be
broken if there's another log with "A join hint" added). What about just
checking if there's a log that contains the message instead of checking the
length of these logs?
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
