Nicolas Paris
2018-11-12 20:01:00 UTC
Hi
The 'org.apache.hadoop.hive.serde2.OpenCSVSerde' is a simple and fast
way to handle csv tables in hive.
However its behavior is to QUOTE ALL columns of the table while there
only is a need for VARCHAR/STRING columns containing a separator in it.
The problem I am facing is full quoted CSV are not well handled by
PostgreSQL. The ideal CSV serializer would only quote character based
columns.
Is there any workaround ?
Otherwise would a pull request on the serde would be accepted ?
Thanks,
The 'org.apache.hadoop.hive.serde2.OpenCSVSerde' is a simple and fast
way to handle csv tables in hive.
However its behavior is to QUOTE ALL columns of the table while there
only is a need for VARCHAR/STRING columns containing a separator in it.
The problem I am facing is full quoted CSV are not well handled by
PostgreSQL. The ideal CSV serializer would only quote character based
columns.
Is there any workaround ?
Otherwise would a pull request on the serde would be accepted ?
Thanks,
--
nicolas
nicolas