On Wed, Feb 13, 2013 at 3:07 PM, Reto Bachmann-Gmür <[email protected]> wrote:
> do they support another method to access al entities?
no
> As for Sparql I do not really understand what they issue is, why does
> it makes a difference for LIMIT if there are some triple-pattern in
> the WHERE clause.
Because with a LIMIT of 10 you will get 7/7 Triples of the first
Entity and 3/15 Triples of the second Entity. Meaning that simple
Triple selection queries can not be used for paging - and when you
want to request all entities you will need to use paging.
So I suppose that one needs to use a special case for such kind of
queries such as
SELECT DISTINCT ?s WHERE ?s ?p ?o OFFSET x LIMIT y
and than load all outgoing triples for returned subjects. I would need
to have a look at the code and possible solutions
best
Rupert
>
> Cheers,
> Reto
>
> On Wed, Feb 13, 2013 at 2:05 PM, Rupert Westenthaler
> <[email protected]> wrote:
>> Hi
>>
>> On Tue, Feb 12, 2013 at 2:12 PM, Reto Bachmann-Gmür <[email protected]> wrote:
>>> Update: I just found FieldQueryImpl so I don't need to create a new
>>> class or to use a Factory.
>>>
>>> However the following returns no results even though entities are available:
>>>
>>> final FieldQuery query = new FieldQueryImpl();
>>> final QueryResultList<Representation> result = siteManager.find(query);
>>> return "executing query result: "+result.results();
>>>
>>> Is it more complicated to make a query without any restriction?
>>
>> I have never tried to execute a query without constraint. This case is
>> also not considered by the implementations.
>>
>> A quick look at the SolrYard implementation suggests that this would
>> generate a SolrQuery with an empty query. An empty query results in an
>> empty query response by Solr. To iterate over all entities you would
>> need to use *:* as query. You can try this on the dev.Iks-project.eu
>> server by using
>>
>> http://dev.iks-project.eu:8081/solr/default/iptc/select?q=
>>
>> and
>>
>> http://dev.iks-project.eu:8081/solr/default/iptc/select?q=*:*
>>
>> This means adding this feature for the Solr based implementation would
>> be easy. For SPARQL I do not have a clear Idea how to formulate an
>> Query that iterate over all Entities - meaning selecting any subject
>> and providing all outgoing triples - mainly because the LIMIT in
>> SPARQL is based on Triples and not based on selected subjects.
>>
>> best
>> Rupert
>>
>>
>>>
>>> Cheers,
>>> Reto
>>>
>>> On Tue, Feb 12, 2013 at 2:05 PM, Reto Bachmann-Gmür <[email protected]>
>>> wrote:
>>>> Thanks Rupert,
>>>>
>>>> Indeed the SiteManager is what I was looking for.
>>>>
>>>> Now i?m trying to find out how to query against the SiteManager. The
>>>> SiteManager has method that take a FieldQuery but it has no method
>>>> providing a QueryFactory. Can I just pass it an anonymous instance of
>>>> Query? What are the set- and remove methods for, is it safe not to
>>>> implement them?
>>>>
>>>> Will a query with an empty constraint set result in all entities being
>>>> returned?
>>>>
>>>> Cheers,
>>>> Reto
>>>>
>>>> On Sun, Feb 10, 2013 at 7:23 PM, Rupert Westenthaler
>>>> <[email protected]> wrote:
>>>>> Hi
>>>>>
>>>>> Are you talking about the entityhub and not about the dbpedia
>>>>> referenced site? The Entityhub is empty by default.
>>>>>
>>>>> best
>>>>> Rupert
>>>>>
>>>>> On Sun, Feb 10, 2013 at 1:36 PM, Reto Bachmann-Gmür <[email protected]>
>>>>> wrote:
>>>>>> Hi Rupert
>>>>>>
>>>>>> I'm using the full launcher. On
>>>>>> http://dev.iks-project.eu:8081/entityhub I don't see more data either.
>>>>>> It doesn't find dbpedia:Paris and the following query yields to no
>>>>>> result:
>>>>>>
>>>>>> {
>>>>>> "query": {
>>>>>> "selected": [],
>>>>>> "constraints": [{
>>>>>> "type": "text",
>>>>>> "patternType": "wildcard",
>>>>>> "text": "*",
>>>>>> "field":
>>>>>> "http:\/\/stanbol.apache.org\/ontology\/entityhub\/entityhub#label"
>>>>>> }],
>>>>>> "limit": 10,
>>>>>> "offset": 0,
>>>>>> "ldpath": "name = rdfs:label[@en] :: xsd:string;"
>>>>>> },
>>>>>> "results": []
>>>>>> }
>>>>>>
>>>>>> Cheers,
>>>>>> Reto
>>>>>>
>>>>>> On Sat, Feb 9, 2013 at 10:30 PM, Rupert Westenthaler
>>>>>> <[email protected]> wrote:
>>>>>>> Hi Reto
>>>>>>>
>>>>>>> AFAIK the stable, full and full-war launcher should still have the
>>>>>>> dbpedia defaultdata index included. If this would not be the case for
>>>>>>> the full launcher, than the integration-test would fail. What launcher
>>>>>>> are you referring to?
>>>>>>>
>>>>>>> best
>>>>>>> Rupert
>>>>>>>
>>>>>>> On Sat, Feb 9, 2013 at 8:51 PM, Reto Bachmann-Gmür <[email protected]>
>>>>>>> wrote:
>>>>>>>> Hello
>>>>>>>>
>>>>>>>> The released version of Stanbol had this 43k default index for the
>>>>>>>> Entityhub. The current snapshot version seems not to have any entity.
>>>>>>>> Because of this the examples on the example page return no results
>>>>>>>> (e.g.
>>>>>>>> the query for http://dbpedia.org/resource/Paris). I like having some
>>>>>>>> data
>>>>>>>> in Entityhub by default. If the data is removed then I think the
>>>>>>>> example
>>>>>>>> should be changed as to include some example creating an entity so
>>>>>>>> that the
>>>>>>>> user can click through it and have some actual results.
>>>>>>>>
>>>>>>>> Cheers,
>>>>>>>> Reto
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> | Rupert Westenthaler [email protected]
>>>>>>> | Bodenlehenstraße 11 ++43-699-11108907
>>>>>>> | A-5500 Bischofshofen
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> | Rupert Westenthaler [email protected]
>>>>> | Bodenlehenstraße 11 ++43-699-11108907
>>>>> | A-5500 Bischofshofen
>>
>>
>>
>> --
>> | Rupert Westenthaler [email protected]
>> | Bodenlehenstraße 11 ++43-699-11108907
>> | A-5500 Bischofshofen
--
| Rupert Westenthaler [email protected]
| Bodenlehenstraße 11 ++43-699-11108907
| A-5500 Bischofshofen
