[ 
https://issues.apache.org/jira/browse/HIVE-10720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14550908#comment-14550908
 ] 

Aihua Xu commented on HIVE-10720:
---------------------------------

The issue is due to ColumnProjectionUtils class is only supported for one table 
column projection. If the query involves multiple tables, then the later one 
will overwrite the previous and causes retrieving incorrect results or not 
retrieving any values. 

I will revert HIVE-5193 so that HCatalog will not use that class for now and do 
the correct column projection enhancement later.

> Pig using HCatLoader to access RCFile and perform join but get incorrect 
> result.
> --------------------------------------------------------------------------------
>
>                 Key: HIVE-10720
>                 URL: https://issues.apache.org/jira/browse/HIVE-10720
>             Project: Hive
>          Issue Type: Bug
>          Components: HCatalog
>    Affects Versions: 1.3.0
>            Reporter: Aihua Xu
>            Assignee: Aihua Xu
>
> {noformat}
> Create table tbl1 (key string, value string) stored as rcfile;
> Create table tbl2 (key string, value string);
> insert into tbl1 values('1', 'value1');
> insert into tbl2 values('1', 'value2');
> {noformat}
> Pig script:
> {noformat}
> tbl1 = LOAD 'tbl1' USING org.apache.hive.hcatalog.pig.HCatLoader();
> tbl2 = LOAD 'tbl2' USING org.apache.hive.hcatalog.pig.HCatLoader();
> src_tbl1 = FILTER tbl1 BY (key == '1');
> prj_tbl1 = FOREACH src_tbl1 GENERATE
>            key as tbl1_key,
>            value as tbl1_value,
>            '333' as tbl1_v1;
>            
> src_tbl2 = FILTER tbl2 BY (key == '1');
> prj_tbl2 = FOREACH src_tbl2 GENERATE
>            key as tbl2_key,
>            value as tbl2_value;
>            
> result = JOIN prj_tbl1 BY (tbl1_key), prj_tbl2 BY (tbl2_key);
> prj_result = FOREACH result 
>       GENERATE  prj_tbl1::tbl1_key AS key1,
>                 prj_tbl1::tbl1_value AS value1,
>                 prj_tbl1::tbl1_v1 AS v1,
>                 prj_tbl2::tbl2_key AS key2,
>                 prj_tbl2::tbl2_value AS value2;
>                
> dump prj_result;
> {noformat}
> Based on the pig script, we could see different invalid results or even no 
> result which should return.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to