W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > January to March 2012

Re: ACTION-575: Propose tests for ADD, MOVE, and COPY

From: Andy Seaborne <andy.seaborne@epimorphics.com>
Date: Thu, 26 Jan 2012 17:43:06 +0000
Message-ID: <4F2190AA.8040003@epimorphics.com>
To: public-rdf-dawg@w3.org


On 26/01/12 15:46, Gregory Williams wrote:
> On Jan 20, 2012, at 4:12 PM, Matthew Perry wrote:
>
>> Hi,
>>
>> I added some SPARQL Update shortcut tests in /add /move and /copy.
>
> Here's another issue. In copy-07, the update is:
>
> COPY :g1 TO :g1
>
> which the spec says is equivalent to:
>
> DROP SILENT GRAPH :g1;
> INSERT { GRAPH :g1 { ?s ?p ?o } } WHERE { GRAPH :g1 { ?s ?p ?o } }
>
> But I think the natural understanding of copying :g1 to :g1 is what the test case expects (that :g1 is unchanged), not that the graph is dropped, and then the copy is left with an empty graph to copy from/to.

I agree that dropping :g1 is not nice and not natural. In the text's 
defense it does say "Data from the input graph is not affected"

The definition also not distinguish

COPY :g1 TO :xyz
COPY SILENT :g1 TO :xyz

but on a store that needs graph created first, these are different.

Similarly:

MOVE :g1 TO :g1

which is defined like COPY-DROP, should not erase :g1

	Andy

>
> .greg
>
>
Received on Thursday, 26 January 2012 17:43:34 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:15:47 GMT