[SPARK-28748] 0 as decimal (n , n) in Hive tables shows as NULL in Spark

• Type: Bug
• Status: Closed
• Priority: Minor
• Resolution: Duplicate
• Affects Version/s: 2.1.0, 2.2.1, 2.3.1, 2.3.2, 2.4.3
• Component/s: SQL
• Assignee: Unassigned
• Reporter: Rohit Sindhu
• Created: 15/Aug/19 20:36
• Updated: 16/Aug/19 10:59
• Resolved: 16/Aug/19 10:59

Description

Zeros(0) inserted as decimal (n , n) in hive tables shows as null in spark sql.

Repro Steps

Hive Shell

create table test_dec (name string , id decimal(3,3));
insert into test_dec values ('c1' , 0) , ('c2' , 0.0) , ('c3' , 0.1);
select * from test_dec;

c1 0.000
c2 0.000
c3 0.100

 

Spark Shell

spark.sqlContext.sql("select * from test_dec").show;

+----+-----+                                                                    
|name|   id|
+----+-----+
|  c1| null|
|  c2| null|
|  c3|0.100|
+----+-----+

Issue Links

• is superceded by: SPARK-23710 Upgrade the built-in Hive to 2.3.5 for hadoop-3.2