On 12/16/15 4:33 AM, Efimov, Alexander wrote: > > Hi Kingsley, > > > > Do you mean that if I modify the query in a following way it should > work without any problem? > > > > PREFIX: sq:<http://our_schema> <http://our_schema> > > > > SELECT * > > FROM <http://this_is_the_graph_name > > > WHERE > > { > > ?customer sq:hasOrder ?order . > > ?order sq:hasGoods ?goods . > > ?goods sq:Name ?goodsName . FILTER (?goodsName in (‘Name1’, > ‘Name2’)) . > > ?goods sq:hasSubCategory ?subCategory . #note that subcategory is IRI > > ?subCategory sq:hasParentCategory ?category #note that category is IRI > > }; >
It will work better (fewer costly joins re. instanceOf handling). Ultimately, for massive amounts of data, you really want to implement the virtual->physical sync option. Then you have the ability to explore data with more flexibility (different rdf views for different purposes) and scalability (instanceOf relations, inference and reasoning etc..). Kingsley > > Thank you. > > Best regards, > > Alexander Efimov. > > > > *From:*Kingsley Idehen [mailto:[email protected]] > *Sent:* Monday, December 14, 2015 11:49 PM > *To:* [email protected] > *Subject:* Re: [Virtuoso-users] Find out which SQL query is generated > from my SPARQL > > > > On 12/14/15 11:34 AM, Efimov, Alexander wrote: > > Hi Kingsley, > > > > The idea is that our concept model I map to with R2RML is not the > same as source tables. > > So just using output from Generate RDF Views would not work. > > I write custom R2RML code which maps from flat DB model to graph > concept model > > > > For example: > > > > Let it be the following tables: > > > > 1. Goods > > a. ID Integer > > b. Name String > > c. Category String > > d. SubCategory String > > > > 2. Customer > > a. ID Integer > > b. Name String > > > > 3. Order > > a. ID Integer > > b. Goods ID Integer > > c. Customer ID Integer > > > > These three tables map to our Concept model which is a bit > different from table structure and I’m executing SPARQL Query like > this: > > > > PREFIX: sq:<http://our_schema> <http://our_schema> > > > > SELECT * > > FROM <http://this_is_the_graph_name > > > WHERE > > { > > ?customer sq:hasOrder ?order . > > ?order sq:hasGoods ?goods . > > ?goods sq:Name ?goodsName . FILTER (?goodsName in (‘Name1’, > ‘Name2’)) . > > ?goods sq:hasSubCategory ?subCategory . #note that subcategory > is IRI > > ?subCategory a sq:SubCategory . > > ?subCategory sq:hasParentCategory ?category #note that > category is IRI > > ?subCategory a sq:Category . > > }; > > > > In my particular case I map data from ~10 views to ~13 classes in > our concept model and WHERE statement is a bit longer. > > > > > > You have a number of rdf:type relations in your query which pose > problems unless you leverage the use of physical triples sync'd with > transient triples (the product of loading your r2rml mappings). > > Command: > > RDF_VIEW_SYNC_TO_PHYSICAL ('{virtual-triples-named-graph-iri}', 1, > '{physical-triples-named-graph-iri}'); > > Example: > > Given a named-graph containing virtual triples (result of loading your > r2rml declarations) identified by the IRI > <http://demo.openlinksw.com/SQLServer#> > <http://demo.openlinksw.com/SQLServer>, you could make the physical > sync replica using a named-graph that holds physical triples > identified by the IRI: <urn:demo.openlinksw.com:SQLServer> using the > command: > > RDF_VIEW_SYNC_TO_PHYSICAL ('http://demo.openlinksw.com/SQLServer# > <http://demo.openlinksw.com/SQLServer>', 1, > 'urn:demo.openlinksw.com:SQLServer'); > > Virtuoso will keep <urn:demo.openlinksw.com:SQLServer> in sync with > <http://demo.openlinksw.com/SQLServer#> > <http://demo.openlinksw.com/SQLServer> and that named graph should be > the target of queries that are scoped to rdf:type relations. > > Links: > > [1] > http://kidehen.blogspot.com/2015/07/conceptual-data-virtualization-across.html > > [2] > http://docs.openlinksw.com/virtuoso/fn_rdf_view_sync_to_physical.html . > > > > -- > Regards, > > Kingsley Idehen > Founder & CEO > OpenLink Software > Company Web: http://www.openlinksw.com > Personal Weblog 1: http://kidehen.blogspot.com > Personal Weblog 2: http://www.openlinksw.com/blog/~kidehen > <http://www.openlinksw.com/blog/%7Ekidehen> > Twitter Profile: https://twitter.com/kidehen > Google+ Profile: https://plus.google.com/+KingsleyIdehen/about > LinkedIn Profile: http://www.linkedin.com/in/kidehen > Personal WebID: http://kingsley.idehen.net/dataspace/person/kidehen#this > > > ------------------------------------------------------------------------------ > > > _______________________________________________ > Virtuoso-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/virtuoso-users -- Regards, Kingsley Idehen Founder & CEO OpenLink Software Company Web: http://www.openlinksw.com Personal Weblog 1: http://kidehen.blogspot.com Personal Weblog 2: http://www.openlinksw.com/blog/~kidehen Twitter Profile: https://twitter.com/kidehen Google+ Profile: https://plus.google.com/+KingsleyIdehen/about LinkedIn Profile: http://www.linkedin.com/in/kidehen Personal WebID: http://kingsley.idehen.net/dataspace/person/kidehen#this
smime.p7s
Description: S/MIME Cryptographic Signature
------------------------------------------------------------------------------
_______________________________________________ Virtuoso-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/virtuoso-users
