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

Re: [TF-LIB] first text for new library built-in functions

From: Lee Feigenbaum <lee@thefigtrees.net>
Date: Thu, 03 Jun 2010 16:08:36 -0400
Message-ID: <4C080BC4.5060702@thefigtrees.net>
To: Andy Seaborne <andy.seaborne@talis.com>
CC: SPARQL Working Group WG <public-rdf-dawg@w3.org>
On 6/3/2010 4:05 PM, Andy Seaborne wrote:
>
>
> On 03/06/2010 5:17 PM, Lee Feigenbaum wrote:
>> On 6/3/2010 8:47 AM, Andy Seaborne wrote:
>>> I've added some text for the new SPARQL 1.1 library functions:
>>>
>>> http://www.w3.org/2009/sparql/docs/query-1.1/rq25.xml#func-coalesce
>>>
>>> The text is not finished; for example, the functions aren't yet in the
>>> functions/operator table, and the details need to be checked.
>>>
>>> The styling of all the functions seems to not working as it did for the
>>> SPARQL 1.0 doc.
>>>
>>> COALESCE, IF, IN, NOT IN, IRI, URI, BNODE, STRDT, STRLANG
>>>
>>> Any I have forgotten or any new ones that people wish to propose?
>>
>> I thought we had discussed explicitly importing more XPath F&O
>> functions, right? Things like string concatenation and some math
>> operations?
>>
>> I'm wary of polluting the language with too many unnecessary keywords,
>> but it would be good to make clear that implementations ought to share
>> URIs for these common functions.
>
> Lee,
>
> We have indeed discussed importing more F&O functions. Last I can find
> for that is:
>
> http://lists.w3.org/Archives/Public/public-rdf-dawg/2009JulSep/0244.html

Ah, great, this is exactly what I was looking for. The list in that 
email looks good to me and I'd love to see if we can reach consensus 
around including those functions (with URIs but no keywords) as required 
parts of SPARQL 1.1.

Lee

>
> and Axel proposed:
>
> http://www.w3.org/2009/sparql/wiki/Feature:FunctionLibrary#Starting_Points
>
> I thought we resolved to use F&O, not the other possible starting points
> but I can't find the actual resolution on that.
>
> http://www.w3.org/2009/sparql/track/actions/97
>
>
>
> This is a different strand: these are the keywords that we have agreed
> (or so I thought) and are in the grammar. None overlap with F&O
> functions [*].
>
> [[
> Revision 1.37 2010/03/30 08:29:52 aseaborne
> Add placeholders for SPARQL 1.1 extra functions
> ]]
>
> Some are fundamental (constructing terms: IRI, URI, STRDT, STRLANG,
> BNODE), some are not just evaluations (COALESCE, IF), and some rely on
> SPARQL operations and have specific syntax in the grammar (IN and NOT
> rely on SPARQL && = and || != respectively)
>
> I believe I am following:
>
> http://www.w3.org/2009/sparql/wiki/Feature:FunctionLibrary#SPARQL_specific
> and
> http://lists.w3.org/Archives/Public/public-rdf-dawg/2010JanMar/0374.html
> and other discussions.
>
>
> It would be good if someone would progress defining a function library.
>
> Andy
>
>
> [*] except arguable STRDT but you can write STRDT(?lex, ?dt) whereas you
> have to write a concrete URI to do the casting form:
>
> xsd:integer("123")
>
>>
>> Lee
>>
>>> We need to allocate IRIs for all the built-ins, nwe and existing.
>>>
>>> Andy
>>>
>>>
>
Received on Thursday, 3 June 2010 20:09:14 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:15:42 GMT