[
https://issues.apache.org/jira/browse/HIVE-2283?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13065677#comment-13065677
]
Navis commented on HIVE-2283:
-----------------------------
FYI, test query was..
explain select distinct t2.key_int1, Q1.something, t3.key_string2 from t2 join
(select key_int1, upper(key_int1) as something from t1) Q1 on Q1.key_int1 =
t2.key_int1 AND lower(Q1.something)='value' join t3 on
Q1.something=t3.key_string2
executed by 3MR(2MapJoin)
> Backtracking real column names for EXPLAIN output
> -------------------------------------------------
>
> Key: HIVE-2283
> URL: https://issues.apache.org/jira/browse/HIVE-2283
> Project: Hive
> Issue Type: Improvement
> Components: Query Processor
> Affects Versions: 0.8.0
> Reporter: Navis
> Priority: Minor
> Attachments: HIVE-2283.1.patch
>
>
> GUI people suggested that showing real column names for result of EXPLAIN
> statement would make customers feel more comfortable with HIVE. I agreed and
> working on it.
> {code}
> a. current EXPLAIN
> Select Operator
> expressions:
> expr: _col10
> type: int
> expr: _col17
> type: string
> Group By Operator
> keys:
> expr: _col0
> type: int
> expr: _col17
> type: int
> b. suggested EXPLAIN
> Select Operator
> expressions: _col10=t2.key_int1, _col17=upper(t1.key_int1),
> _col22=t3.key_string2
> Group By Operator
> keys: _col10=t2.key_int1, _col17=upper(t1.key_int1)
> {code}
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
