W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > January to March 2011

Re: Draft response to HK-2

From: Andy Seaborne <andy.seaborne@epimorphics.com>
Date: Tue, 08 Feb 2011 09:19:24 +0000
Message-ID: <4D510A9C.8010502@epimorphics.com>
To: Paul Gearon <gearon@ieee.org>
CC: SPARQL Working Group <public-rdf-dawg@w3.org>
The draft is OK - I don't mind having multiple sections about one form. 
  Given we have DELETE WHERE it makes some sense.

The grammar extract is not right - see below.

On 08/02/11 02:30, Paul Gearon wrote:
> Since drafting this response, I recall another reason for keeping
> DELETE/INSERT, INSERT and DELETE all separate.
> The reason was to allow either the DELETE clause or the INSERT clause
> to be optional, but not both. I'm not sure if it matters if we want to
> allow for both to be omitted. I suggest we allow it, and make it a
> no-op, regardless of the contents of the WHERE clause.

The grammar has one rule for the delete-insert-where action.


[42] ... ( DeleteClause InsertClause? | InsertClause ) ...

which allows for

but not missing both.

The document has (3.1.3)

[ WITH <uri> ]
( DELETE { modify_template } )*
( INSERT { modify_template } )*

which is not the same.

1/ "*" (which is "zero of more of") is a change to the language
2/ Omiting both is now possible.

if you want to go with the style of the doc, it should be

[ WITH <uri> ]
DELETE { modify_template }
INSERT { modify_template }

because the subcases are dealt with later.

It might be better to starts with the subcases.

Given the grammar only accepts these forms, the formal description for 
SPARQL Update can work with that or the more general case of 
DeleteClause? InsertClause?  i.e. it only needs DeleteInsertOperation


> Paul
> On Mon, Feb 7, 2011 at 12:48 PM, Paul Gearon<gearon@ieee.org>  wrote:
>> Hi,
>> I have a draft response to Holger Knublauch for comment HK-2 at:
>>   http://www.w3.org/2009/sparql/wiki/CommentResponse:HK-2
>> Regards,
>> Paul Gearon
Received on Tuesday, 8 February 2011 09:20:02 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:01:03 UTC