W3C home > Mailing lists > Public > public-rdf-dawg-comments@w3.org > November 2011

Re: SPARQL 1.1 - LeftJoin definition

From: David McNeil <dmcneil@revelytix.com>
Date: Mon, 28 Nov 2011 15:51:52 -0600
Message-ID: <CA+8VvdxzRR5q2Qb4WKbFmLPveqhJaKV21PQAqoW7URfvYtYe2w@mail.gmail.com>
To: "Eric Prud'hommeaux" <eric@w3.org>
Cc: public-rdf-dawg-comments@w3.org
On Mon, Nov 28, 2011 at 2:35 PM, Eric Prud'hommeaux <eric@w3.org> wrote:

> yep, but the wierd bit is that the filter in the OPTIONAL is scoped to
> both the left and right sides of the OPTIONAL:
> [[
>  SELECT ?label
>  {
>      {
>          {SELECT*{BIND(1 AS?match) BIND(10 AS?x)}}
>      } OPTIONAL {
>          {SELECT*{BIND("COMPAT PASS" AS?label) BIND(1 AS?match)BIND(10
> AS?y)}} UNION
>          {SELECT*{BIND("COMPAT FAIL" AS?label) BIND(1 AS?match)BIND(11
> AS?y)}} UNION
>          {SELECT*{BIND("INCOMPAT"    AS?label) BIND(2 AS?match)BIND(10
> AS?y)}}
>          FILTER (?x = ?y)
>      }
>  }
> ]]
>

Ah yes, that makes sense. So if the filter expression spans the left &
right then it cannot be pushed down. Whereas if the filter expression only
involves the right side, then it can be pushed down. Thanks for the help.

btw, will you be coming to http://www.w3.org/2011/09/LinkedData/ ?
>

I am not planning on attending.

-David
Received on Monday, 28 November 2011 21:52:25 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 28 November 2011 21:52:26 GMT