[Nepomuk] Query for resource for exactly one known file

Dean Perry happy.heyoka at gmail.com
Fri Aug 10 15:55:56 UTC 2012


Hi all, more noob questions:

I'm trying to write something that will retrieve the resource for exactly one file for which I have the URL 
from a drag and drop operation...  I'm using the Nepomuk::Query stuff because, to be honest, I can't 
make head nor tails of SPARQL yet.

So for starters:
 - I know that I have a valid URL for a file
 - I know its is in the database and
 - using NepSAK I know that the resource has both NFO::fileName and NIE::url associated.

I was thinking that NIE::url would be the thing to compare against:

	QString urlstring = url.toString();
	Nepomuk::Query::ComparisonTerm term (
		Nepomuk::Vocabulary::NIE::url(),
		Nepomuk::Query::LiteralTerm ( urlstring ),
		Nepomuk::Query::ComparisonTerm::Equal
	); 

But this doesn't return any results, even when I know for certain that the file has a resource with the 
NIE::url property that has the value I'm looking for.  Changing 'Equal' to 'Contains' doesn't work either.

Using NFO::fileName works fine, but returns several results if there's a duplicate in the file system:

	QString filename = urlstring.right ( ( urlstring.length() - urlstring.lastIndexOf ( '/' ) )-1 );  // yeah, ugly...
	Nepomuk::Query::ComparisonTerm term(
		Nepomuk::Vocabulary::NFO::fileName(),
		Nepomuk::Query::LiteralTerm(filename),
		Nepomuk::Query::ComparisonTerm::Contains
	);

Using the FileQuery class does not help (but adds some constraints to the SPARQL).

I've also tried to use just the filename and 'Contains' against NIE::url, and that doesn't work either.

As I said, I'm pretty clueless with SPARQL, but when I take the Query generated SPARQL and paste 
into NepSAK the results are the same (obvious) but when I replace the nfo::fileName with nie::url it also 
doesn't work which is not very obvious to me eg:

select distinct ?r , ?v1 as ?_n_f_t_m_ex_ where { ?r nie:url ?v1 . FILTER(bif:contains(?v1, 
"'095262r00ZB_zigbee_rf4ce_sc-ZigBee_RF4CE_Specification_public.pdf'")) . FILTER EXISTS { ?r a 
[ nao:userVisible "true"^^XMLSchema:boolean ] . } . } LIMIT 20


I'm guessing I'm missing something obvious here...

cheers,

dean

Oh and secondary questions:

(1) the generated query has the stuff between ?r and ?v1 which seems extraneous and seems to return 
a result with HTML markup (<b></b> tags) that disappear if I remove them:

 	select distinct ?r (bif:concat(bif:search_excerpt(bif:vector('095262r00ZB_zigbee_rf4ce_sc-
ZigBee_RF4CE_Specification_public.pdf'), ?v1))) ....

vs
	select distinct ?r, ?v1 ....

  produces

	<b>095262r00ZB</b> zigbee rf4ce sc ZigBee RF4CE Specification public.
vs
	095262r00ZB_zigbee_rf4ce_sc-ZigBee_RF4CE_Specification_public.pdf


(2) is it just me or does the 'bool' parameter for QueryServiceClient::syncQuery always return 'false' ?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/nepomuk/attachments/20120811/b870f1c6/attachment.html>


More information about the Nepomuk mailing list