Github user mgaido91 commented on the issue:
https://github.com/apache/spark/pull/22029
Presto and Hive work in different ways.
**Hive**
Hive behaves as Spark before the PR, but Hive doesn't support IN with
subqueries where the subquery outputs more than one field. So in Hive there is
not the incoherence which is present in Spark (ie. with the same data IN with
literals returns a different result than IN with subquery).
**Presto**
Presto, instead, behaves as Spark before the PR, but when there is a null
both in the value and in the list of literal, it throws an exception and it
behaves in the same way also when IN is used with subqueries.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
