Suresh Thalamati created SPARK-15315:
----------------------------------------
Summary: CSV datasource writes garbage for complex types instead
of rasing error.
Key: SPARK-15315
URL: https://issues.apache.org/jira/browse/SPARK-15315
Project: Spark
Issue Type: Bug
Components: SQL
Affects Versions: 2.0.0
Reporter: Suresh Thalamati
Priority: Minor
Complex types support does not seem to exist yet for csv writer. Error should
be raised instead of writing garbage into the files.
Read throw error java.lang.RuntimeException: Unsupported type: map
Repro:
{code}
Seq((1, Map("Tesla" -> 3))).toDF("id", "cars").write.csv("/tmp/output")
cat part-r-00007-f684166f-ac28-48a1-99dc-1bb2c2ef1164.csv
1,org.apache.spark.sql.catalyst.expressions.UnsafeMapData@6d10824e
Seq((1, Array("Tesla"))).toDF("id",
"cars").write.mode("overwrite").csv("/tmp/output")
cat part-r-00007-5bc39b53-25b6-4802-a98d-ec0cbd0e189a.csv
1,org.apache.spark.sql.catalyst.expressions.UnsafeArrayData@a7e180b5
scala> Seq((1, "Tesla")).toDF("id", "name").select(struct("id",
"name")).write.mode("overwrite").csv("/tmp/output")
cat part-r-00007-e0c688e1-e04d-4fa0-baf9-31183871ef6d.csv
"[0,1,1800000005,616c736554]"
{code}
--
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
