[
https://issues.apache.org/jira/browse/HIVE-8004?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14123776#comment-14123776
]
David Beveridge commented on HIVE-8004:
---------------------------------------
I've been thinking over the same thing myself (where we could, as part of
SerDe, have a Transform function that would encode the data in a
hive-friendly way (such as the hex value of the string)
Of course, that means that queries couldn't use the data directly, which
would be a pain--querying for
parser.imageimportnames.736f6d654261645472616974 = "true", rather than
parser.imageimportnames.someBadTrait = "true", (or some equivalent
mechanism ).
Perhaps it would be worthwhile to measure the interest in this in the
field? We certainly might be an edge case . . . .
> Add Case-Sensitivity to Column Names
> ------------------------------------
>
> Key: HIVE-8004
> URL: https://issues.apache.org/jira/browse/HIVE-8004
> Project: Hive
> Issue Type: Improvement
> Components: Metastore
> Affects Versions: 0.11.0
> Environment: n/a
> Reporter: David Beveridge
>
> Adding an option to adopt case-sensitivity to column names should be
> considered.
> I will use our own case for consideration: we are using Hive for both
> Natural Language Processing queries and hosting data for Genetic Algorithms
> (and querying traits).
> In both cases, we need to dynamically generate schema to host the data;
> meaning that the Column Name in Hive is actually data (as well as its value)
> and, thus, case-sensitivity matters.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
