[
https://issues.apache.org/jira/browse/HIVE-337?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12681064#action_12681064
]
Zheng Shao commented on HIVE-337:
---------------------------------
For the migration path, we can easily create new tables with new SerDes, and
everything will work transparently. If you are suggesting letting LazySerDe
automatically figure out the old/new format, I don't think that's even
possible, and if it is, users will be easily confused by that.
Delimited format is meant to be simple and human-readable, and it is only good
for simple data. If the structure really gets complicated, we should store the
data in binary format instead of delimited format. For example, we can use
Thrift etc.
If we really want to write a new SerDe that shares a lot with LazySimpleSerDe
(with an extended delimited format), we can easily do that by reusing a lot of
the classes introduced by LazySimpleSerDe. There is not much to copy - and if
there is, it's better to factor the common code out, instead of pushing all
logics (new format/old format) into the same class.
Let's open another jira for discussions on new features like this.
So the question here is that we have to make a choice between the two: whether
treat "" to be an empty array or an array with an empty string as the only
element (and the same question for NULL).
> LazySimpleSerDe should support array and map types
> --------------------------------------------------
>
> Key: HIVE-337
> URL: https://issues.apache.org/jira/browse/HIVE-337
> Project: Hadoop Hive
> Issue Type: Bug
> Components: Serializers/Deserializers
> Affects Versions: 0.2.0
> Reporter: Zheng Shao
> Assignee: Zheng Shao
> Priority: Blocker
>
> Once we do that, we can completely deprecate
> DynamicSerDe/TCTLSeparatedProtocol, and close any bugs that
> DynamicSerDe/TCTLSeparatedProtocol has.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.