[jira] [Commented] (HIVE-4100) Improve regex_replace UDF to allow non-ascii characters

["Teddy Choi (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HIVE-4100?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13737987#comment-13737987 ] 

Teddy Choi commented on HIVE-4100:
----------------------------------

If we allow "\uffff" form, then "\UDFBa" in "hive\ql\udf\UDFBack.java" may be parsed as a unicode character. It is ambiguous.

How about this way?
{code}REGEXP_REPLACE(some_column, "[^\\u0000-\\uffff]", UNESCAPE_UNICODE("\ufffd")){code}
In this example, UNESCAPE_UNICODE is just a random name.
                
> Improve regex_replace UDF to allow non-ascii characters
> -------------------------------------------------------
>
>                 Key: HIVE-4100
>                 URL: https://issues.apache.org/jira/browse/HIVE-4100
>             Project: Hive
>          Issue Type: Improvement
>          Components: UDF
>    Affects Versions: 0.10.0
>            Reporter: Mark Grover
>            Assignee: Mark Grover
>
> There have a been a few email threads on the user mailing list regarding regex_replace UDF not supporting non-ASCII characters. We should validate that and improve the UDF to allow it. Translate UDF will be a good reference since it does that by using code points instead of characters

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira