[
https://issues.apache.org/jira/browse/PHOENIX-3187?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15429145#comment-15429145
]
James Taylor commented on PHOENIX-3187:
---------------------------------------
Can you elaborate more on why you need multi-byte characters for CHAR datatype?
An easy workaround it to just use VARCHAR as it supports multi-byte characters.
The reason CHAR doesn't is that we take advantage of knowing the byte length of
a value given the declaration of the length in the schema. This gives us the
opportunity of performing certain operations when otherwise we couldn't.
> Support multi-byte characters for CHAR datatype
> -----------------------------------------------
>
> Key: PHOENIX-3187
> URL: https://issues.apache.org/jira/browse/PHOENIX-3187
> Project: Phoenix
> Issue Type: Improvement
> Affects Versions: 4.7.0
> Reporter: Vikash Talanki
> Priority: Minor
>
> Currently, Phoenix CHAR data type supports only single byte characters. We
> need support for multi byte characters in CHAR datatype.
> Case 1: Support multi byte characters in CHAR data type like VARCHAR -
> CHAR(1) can take any multi-byte character.
> Case 2: Support multi byte characters in CHAR data type only if it fits in
> defined length - CHAR(1) takes only single byte character but CHAR( n ) can
> take one n-byte character. In other words, the total number of bytes of all
> characters should not exceed the defined length.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
