[
https://issues.apache.org/jira/browse/HCATALOG-249?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13213915#comment-13213915
]
Sushanth Sowmyan commented on HCATALOG-249:
-------------------------------------------
line 94 and similar changes of HCatRecordSerDe - done
line 157 - I dug a bit, and looks like Collections.nCopies(size, null) is a
good way of doing so. Using that.
line 170 - hmm.. I'm a bit conflicted about this. I think that if there's a
parse problem, that is a case for a SerDe exception for being able to decode
that row. Sending back nulls is essentially then sending back incorrect data.
It should be up to the record reader to ignore a serde exception it sees, if it
thinks that a row of malformed json shouldn't block reading a file, and to then
cause it to read the next row. I've noticed that the RCFile Reader has similar
logic, whereupon, if it detects malformed data, and it is configured to ignore
malformed data, it goes on to read the next line.
line 399 - whoops, yes. Removing binary altogether, and changing it to throw an
exception if it detects that it's being used for binary.
> Rework JSON StorageDriver into a JSON SerDe
> -------------------------------------------
>
> Key: HCATALOG-249
> URL: https://issues.apache.org/jira/browse/HCATALOG-249
> Project: HCatalog
> Issue Type: Sub-task
> Affects Versions: 0.4
> Reporter: Alan Gates
> Assignee: Sushanth Sowmyan
> Fix For: 0.4
>
> Attachments: HCATALOG-249.patch, HCATALOG-249.patch.2
>
>
> We have a JSON storage driver. Given the switch to SerDes we need to switch
> it to be a SerDe instead.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
