miland-db commented on code in PR #47423:
URL: https://github.com/apache/spark/pull/47423#discussion_r1711354710
##########
sql/core/src/main/scala/org/apache/spark/sql/scripting/SqlScriptingExecutionNode.scala:
##########
@@ -144,46 +265,84 @@ abstract class
CompoundNestedStatementIteratorExec(collection: Seq[CompoundState
case _ => throw SparkException.internalError(
"Unknown statement type encountered during SQL script
interpretation.")
}
- localIterator.hasNext || childHasNext
+ (localIterator.hasNext || childHasNext || returnHere.isDefined) &&
!stopIteration
Review Comment:
`returnHere` is added here to support correct iteration when handling an
error with `continue` handler. We need this check because when we encounter
statement that raises an error, we move `curr` to the next statement to be
ready to continue execution after handler body. In this case:
```
BEGIN
DECLARE flag INT = -1;
SELECT 2;
BEGIN
SELECT 3;
BEGIN
DECLARE zero_division CONDITION FOR '22012';
DECLARE CONTINUE HANDLER FOR zero_division
BEGIN
SELECT flag;
SET VAR flag = 1;
END;
SELECT 4;
SELECT 1/0;
SELECT 5;
END;
SELECT 6;
END;
SELECT 7;
SELECT flag;
END
```
when we return to the execution after handler body is executed `hasNext`
will return false because `curr` is already pointing at `SELECT 5` statement
and there are no more elements. By checking if there is something to return to
using `returnHere`, we solve this case.
If there were some statement after `SELECT 5` in the same block, everything
would work normally.
--
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]
