[
https://issues.apache.org/jira/browse/HIVE-9907?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15058772#comment-15058772
]
Alan Gates commented on HIVE-9907:
----------------------------------
I'm confused on what is being reported here. The initial descriptions seems to
say that this doesn't work with insert values, but there's a test
(insert_values_nonascii.q) that explicitly tests this case and passes. Does
this just not work on some OS or JVM configurations? A later comment implies
this doesn't work with update.
In general the patch looks good, but I'd like to know exactly what we're fixing
first. The patch should address insert values, but won't help update.
Niklaus, I've assigned the JIRA to you. The next step is for you to mark it
patch available so it can be run through the tests.
> insert into table values() when UTF-8 character is not correct
> ----------------------------------------------------------------
>
> Key: HIVE-9907
> URL: https://issues.apache.org/jira/browse/HIVE-9907
> Project: Hive
> Issue Type: Bug
> Components: CLI, Clients, JDBC
> Affects Versions: 0.14.0, 0.13.1, 1.0.0
> Environment: centos 6 LANG=zh_CN.UTF-8
> hadoop 2.6
> hive 1.1.0
> Reporter: Fanhong Li
> Assignee: niklaus xiao
> Priority: Critical
> Attachments: HIVE-9907.1.patch
>
>
> insert into table test_acid partition(pt='pt_2')
> values( 2, '中文_2' , 'city_2' )
> ;
> hive> select *
> > from test_acid
> > ;
> OK
> 2 -�_2 city_2 pt_2
> Time taken: 0.237 seconds, Fetched: 1 row(s)
> hive>
> CREATE TABLE test_acid(id INT,
> name STRING,
> city STRING)
> PARTITIONED BY (pt STRING)
> clustered by (id) into 1 buckets
> stored as ORCFILE
> TBLPROPERTIES('transactional'='true')
> ;
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)