W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > July to September 2010

Re: Group by expressions

From: Polleres, Axel <axel.polleres@deri.org>
Date: Tue, 31 Aug 2010 13:10:08 +0100
Message-ID: <316ADBDBFE4F4D4AA4FEEF7496ECAEF9035DF843@EVS1.ac.nuigalway.ie>
To: <andy.seaborne@epimorphics.com>
Cc: <public-rdf-dawg@w3.org>
That looks fine to me (and seems no problem to copy/adapt the agg test cases accordingly) Will wait for what you guys decide in today's call!


----- Original Message -----
From: Andy Seaborne <andy.seaborne@epimorphics.com>
To: Polleres, Axel
Cc: SPARQL Working Group <public-rdf-dawg@w3.org>
Sent: Tue Aug 31 11:51:24 2010
Subject: Group by expressions

Agenda for 2010-08-31:
 > Projecting expressions used for grouping - see recent discussion of 
ISSUE-41 on the mailing list

Input: The current grammar has:

[20]  GroupCondition  ::=
    ( BuiltInCall
    |  FunctionCall
    |  '(' Expression ( 'AS' Var )? ')'
    | Var )

and the name introduced can then be used in SELECT expressions and HAVING.

SELECT (?o3+128 AS ?o4)
GROUP BY (?o1+?o2 AS ?o3)
HAVING ( ?o3 > 45)

Currently, the name is optional, reflecting that it may not be used 
again if it's not mentioned in HAVING or SELECT.  This avoids needing to 
require sub-expression determination.

Received on Tuesday, 31 August 2010 12:10:46 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:01:01 UTC