Re: [SPARQL-Update] WITH, GRAPH, and USING [NAMED]

Apologies, this email was meant to be sent to the list.


---------- Forwarded message ----------
From: Pavel Klinov <pavel@complexible.com>
Date: Fri, Aug 5, 2016 at 4:37 PM
Subject: Re: [SPARQL-Update] WITH, GRAPH, and USING [NAMED]
To: Andy Seaborne <andy@apache.org>


Hi Andy,

On Fri, Aug 5, 2016 at 3:32 PM, Andy Seaborne <andy@apache.org> wrote:

> On 05/08/16 09:05, Pavel Klinov wrote:
>
>>
>>
>> On Thu, Aug 4, 2016 at 11:04 PM, Pavel Klinov <pavel@complexible.com
>> <mailto:pavel@complexible.com>> wrote:
>>
>>
>>     According to my, perhaps incorrect, reading of 3.1.3 this should be
>>     equivalent to:
>>
>>     DELETE
>>     {
>>       GRAPH <http://example.org/g1> { ?s ?p ?o }
>>     }
>>     USING <http://example.org/g2> WHERE
>>     {
>>       GRAPH <http://example.org/g1> { :a foaf:knows ?s . ?s ?p ?o }
>>     }
>>
>>     that is, WITH is ignored for DELETE because there's the explicit
>>     GRAPH clause but applies to WHERE because there's no USING [NAMED].
>>     Then this query should not delete any triples because the dataset
>>     for WHERE is composed of the empty default graph and :g2 as the only
>>     named graph, while WHERE matches triples in :g2.
>>
>>
>> A correction here: the dataset for WHERE should be :g2 as the default
>> graph and no named graphs (or unspecified named graphs, as per 13.2 in
>> [3]). But the rest of the email holds, the query can still return no
>> results contarry to the test.
>>
>
> Hi Pavel,
> (not an official reply)
>
> I believe the intent is that WITH sets the default graph but otherwise the
> named graph are in the dataset visible.
>

OK, that's a possible interpretation but it's different from wrapping WHERE
in a GRAPH. Either it does something to the dataset or changes the pattern
in WHERE, those are different things.


> Unlike USING, WITH is not describing the whole dataset to use.  Is there
> text that suggested to it did? Or entangles WITH and USING? If so, an
> errata would be in order.
>

First, there's the snippet you quoted below: "If a graph name is specified
in a WITH clause, then - for the purposes of evaluating the WHERE clause -
this will define an RDF Dataset..."

Second, there's the example in 3.1.3:

[[

To illustrate the use of the WITH clause, an operation of the general form:

WITH <g1> DELETE { a b c } INSERT { x y z } WHERE { ... }
is considered equivalent to:

DELETE { GRAPH <g1> { a b c } } INSERT { GRAPH <g1> { x y z } } USING <g1>
WHERE { ... }

]]

This gives the impression that WITH, when not ignored, is a syntactic sugar
for USING. And since USING has basically the same meaning as FROM, it
should follow that WITH defines the dataset to the extent that FROM does.

Now, regarding your point that it may leave the set of named graphs
unchanged. Actually 13.2 of the SPARQL spec seems to leave such possibility:

[[
If there is no FROM clause, but there is one or more FROM NAMED clauses,
then the dataset includes an empty graph for the default graph.
]]

See, it doesn't say what the set of named graphs is if there's FROM but no
FROM NAMED (so in our case USING but not USING NAMED). So it's indeed
possible for WITH, when interpreted as USING, to describe only the default
graph. But even then the test with :g1 and :g2 doesn't seem right because
in the absense of defined named graphs, an implementation doesn't have to
include :g1. The dataset <:g2, {}> seems like a valid implementation choice.


> The later text:
> [[
> If a graph name is specified in a WITH clause, then - for the purposes of
> evaluating the WHERE clause - this will define an RDF Dataset containing a
> default graph with the specified name
> ]]
>
> look to me to be at odds with:
>
> [[
> The WITH clause defines the graph that will be modified or matched against
> for any of the subsequent elements (in DELETE, INSERT, or WHERE clauses) if
> they do not specify a graph explicitly.
> ]]
>
> [[
> That is, a WITH clause may be viewed as syntactic sugar for wrapping both
> the QuadPatterns in subsequent DELETE and INSERT clauses, and likewise the
> GroupGraphPattern in the subsequent WHERE clause into GRAPH patterns.
> ]]
>
> which both suggest your example of:
>
> >     {
> >       GRAPH <http://example.org/g2> {
> >         GRAPH <http://example.org/g1> { :a foaf:knows ?s . ?s ?p ?o }
> >       }
> >     }
>
> is intended.
>

Yes, I agree that the latter two sentences do suggest that. I don't object
against such an interpretation, just trying to confirm what should be the
right behavior.

Thanks again,
Pavel


>
>         Andy
>
>
>> Thanks,
>> Pavel
>>
>> [3] https://www.w3.org/TR/sparql11-query/#specDataset
>>
>>
>>
>>     However, this is not what DAWG tests [2] expect and it is not how
>>     ARQ or Sesame implement it (they pass the test). What I can tell
>>     from ARQ's source code is that it treats the query as follows:
>>
>>     DELETE
>>     {
>>       GRAPH <http://example.org/g1> { ?s ?p ?o }
>>     }
>>     {
>>       GRAPH <http://example.org/g2> {
>>         GRAPH <http://example.org/g1> { :a foaf:knows ?s . ?s ?p ?o }
>>       }
>>     }
>>
>>     In this case, of course, WHERE will match triples because the inner
>>     GRAPH will set :g1 as the active graph for the BGP. However 3.1.3
>>     says explicitly that WITH, when not ignored, defines an RDF Dataset
>>     -- this isn't what happens here.
>>
>>     Another interpretation could be as follows: WITH is in effect for
>>     WHERE if there's no USING [NAMED] *and* there is no GRAPH clause in
>>     either DELETE or INSERT. According to that interpretation the test
>>     again should pass because WITH is simply ignored for the whole
>>     query, including WHERE, even though there's no USING [NAMED].
>>
>>     So the question is: which interpretation is the intended one?
>>
>>     Thanks,
>>     Pavel
>>
>>     [1] https://www.w3.org/TR/sparql11-update/#deleteInsert
>>     <https://www.w3.org/TR/sparql11-update/#deleteInsert>
>>     [2] https://www.w3.org/2009/sparql/docs/tests/data-sparql11/dele
>> te/manifest#dawg-delete-with-02
>>     <https://www.w3.org/2009/sparql/docs/tests/data-sparql11/del
>> ete/manifest#dawg-delete-with-02>
>>
>>
>>
>
>