[HIVE-581] improve group by syntax

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.11.0
• Component/s: Clients, Query Processor
• Labels: features
• Hadoop Flags: Reviewed
• Release Note: Allow the syntax of position alias in groupby expression.
• Assignee: Zhenxiao Luo Assign to me
• Reporter: Larry Ogrodnek
• Created: 25/Jun/09 02:29
• Updated: 31/Mar/15 08:38
• Resolved: 27/Dec/12 15:23

Description

It would be nice if group by allowed either column aliases or column position (like mysql).

It can be burdensome to have to repeat UDFs both in the select and in the group by.

e.g. instead of:
select f1(col1), f2(col2), f3(col3), count(1) group by f1(col1), f2(col2), f3(col3);

it would allow:
select f1(col1), f2(col2), f3(col3), count(1) group by 1, 2, 3;

Issue Links

• is duplicated by: HIVE-8388 Support GROUP/ORDER BY column references by ordinal position
• is related to: HIVE-1947 ORDER BY should support column position expressions
• relates to: HIVE-494 Select columns by index instead of name