N Campbell created SPARK-10503:
----------------------------------
Summary: incorrect predicate evaluation involving NULL value
Key: SPARK-10503
URL: https://issues.apache.org/jira/browse/SPARK-10503
Project: Spark
Issue Type: Bug
Components: SQL
Reporter: N Campbell
Query an ORC table in Hive using the following SQL statement via the SPARKSQL
thrift-server. The row were rnum=0 has a c1 value of null. The resultset
returned by SPARK includes a row where rnum=0 and c1=0 which is incorrect
select tint.rnum, tint.rnum from tint where tint.cint in ( tint.cint )
table in Hive
create table if not exists TINT ( RNUM int , CINT smallint )
TERMINATED BY '\n'
STORED AS orc ;
data loaded into ORC table is
0|\N
1|-1
2|0
3|1
4|10
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
