[
https://issues.apache.org/jira/browse/ORC-145?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17264438#comment-17264438
]
Juho Salmio commented on ORC-145:
---------------------------------
I managed to make some initial progress, but bumped into a small problem. I use
printJsonData from PrintData class when reading the content from the example
files.
It outputs the rows like this:
{{{"int1":-1155869325,"string1":"bb2c72394b1ab9f8"}}}
The expected files has the data formatted like this:
{{{"int1": -1155869325, "string1": "bb2c72394b1ab9f8"}}}
So there are extra spaces after delimiters.
printJsonData uses {{org.codehaus.jettison.json.JSONWriter}} which doesn't seem
to have any formatting options available.
I'm now not sure whether we should change the format of the {{expected }}files
and then change the tests in the c++ side or make some java tests specific json
printing logic and not use the PrintData class at all or something else. Any
pointers?
> Add unit tests for Java to read the example ORC files
> -----------------------------------------------------
>
> Key: ORC-145
> URL: https://issues.apache.org/jira/browse/ORC-145
> Project: ORC
> Issue Type: Improvement
> Affects Versions: 1.7.0
> Reporter: Owen O'Malley
> Priority: Major
> Labels: starter
>
> The C++ reader is tested by reading all of the ORC example files. We need
> similar tests for the Java reader to ensure that we don't break backwards
> compatibilityl
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
