www.openlinksw.com
docs.openlinksw.com

Book Home

Contents
Preface

Overview

What is Virtuoso?
Why Do I Need Virtuoso?
Key Features of Virtuoso
Virtuoso 6 FAQ
Tips and Tricks
How Can I execute SPARQL queries containing '$' character using ISQL? How can I find on which table deadlocks occur? How Can I configure parameters to avoid out of memory error? What are "Generate RDB2RDF triggers" and "Enable Data Syncs with Physical Quad Store" RDF Views options? How to Manage Date Range SPARQL queries? How can I see which quad storages exist and in which quad storage a graph resides? Can I drop and re-create the DefaultQuadStorage? How to display only some information from RDF graph? Is it possible to have the SPARQL endpoint on a different port than the Conductor? How to enable the Virtuoso Entity Framework 3.5 ADO.Net Provider in Visual Studio 2010? How Can I Control the normalization of UNICODE3 accented chars in free-text index? How Can I define graph with virt:rdf_sponger option set to "on"? How do I use SPARUL to change a selection of property values from URI References to Literals? How is a Checkpoint performed against a Virtuoso Clustered Server? How can I use CONSTRUCT with PreparedStatements? How can perform SPARQL Updates without transactional log size getting exceed? How can I write custom crawler using PL? How Can I Get an exact mapping for a date? How Can I Get certificate attributes using SPARQL? How can I make Multi Thread Virtuoso connection using JDBC?? How Do I Perform Bulk Loading of RDF Source Files into one or more Graph IRIs? How to exploit RDF Schema and OWL Inference Rules with minimal effort? How can I dump arbitrary query result as N-Triples? How do I bind named graph parameter in prepared statement? How can I insert binary data to Virtuoso RDF storage in plain queries and with parameter binding via ADO.NET calls? How can I insert RDF data from Visual Studio to Virtuoso? How does default describe mode work? What should I do if the Virtuoso Server is not responding to HTTP requests? What CXML params are supported for the SPARQL URL pattern? How can I replicate all graphs? What is best method to get a random sample of all triples for a subset of all the resources of a SPARQL endpoint? How can I replicate all graphs? How can I use SPARQL to make Meshups? How can I use the net_meter utility before starting the ingestion to a cluster? How can I use the LOAD command to import RDF data? How can I delete graphs using stored procedure? How can I use SPARUL to add missing triples to a Named Graph? How can I use the SPARQL IF operator for SPARQL-BI endpoint? How can I handle checkpoint condition? How can I incorporate Content Negotiation into RDF bulk loaders? Virtuoso Linked Data Deployment In 3 Simple Steps? What are the differences between create, drop, clear and delete Graph? How can I perform search for predicate values? How can I use INSERT via CONSTRUCT Statements? How to clear graphs which are related to empty graphs? How can I use sub-queries to enable literal values based joins? How can I execute query with labels preference order? How can I get object datatype? How Can I Backup and Restore individual table(s) and individual index(s)? What bif:contains free-text options can I use? What SPARQL Endpoint Protection Methods can I use? How do I assign SPARQL role to SQL user?

1.5. Tips and Tricks

1.5.1. How Can I execute SPARQL queries containing '$' character using ISQL?

Assuming a SPARQL query should filter on the length of labels:

SELECT ?label
FROM <http://mygraph.com>
WHERE 
  { 
    ?s ?p ?label
    FILTER(regex(str(?label), "^.{1,256}$") )
  } 

ISQL uses '$' character as a prefix for macro names of its preprocessor. When '$' character is used in SPARQL query to be executed in ISQL, the character should be replaced with '$$' notation or an escape char + numeric code:

SQL> SPARQL 
SELECT ?label
FROM <http://mygraph.com>
WHERE 
  { 
    ?s ?p ?label
    FILTER(REGEX(str(?label), "^.{1,256}$$") )
  } 	

Note also that the FILTER written in this way, finds ?label-s with length less than 256.

To achieve fast results, REGEX should be replaced with the bif:length function:

SQL> SPARQL 
SELECT ?label
FROM <http://mygraph.com>
WHERE 
  { 
    ?s ?p ?label
    FILTER (bif:length(str(?label))<= 256)
  } 	

In this way the SPARQL query execution can work much faster if the interoperability is not required and ?label-s are numerous.


1.5.2. How can I find on which table deadlocks occur?

One possible way to find on which table deadlocks occur is to execute the following statement:

SELECT TOP 10 * 
FROM SYS_L_STAT 
ORDER BY deadlocks DESC	

1.5.3. How Can I configure parameters to avoid out of memory error?

In order to avoid out of memory error, you should make sure the values for the paramaters NumberOfBuffers and MaxCheckpointRemap are not set with the same values.

For example, the following configuration will cause an error out of memory:

# virtuoso.ini

...
[Parameters]
NumberOfBuffers = 246837
MaxDirtyBuffers = 18517
MaxCheckpointRemap = 246837
...

Changing the value of the parameter MaxCheckpointRemap with let's say 1/4 of the DB size will resolve the issue.


1.5.4. What are "Generate RDB2RDF triggers" and "Enable Data Syncs with Physical Quad Store" RDF Views options?

These RDF Views options basically persist the triples in the transient View Graph in the Native Quad Store. The Data Sync is how you keep the transient views in sync with the persisted triples.

Without this capability you cannot exploit faceted browsing without severe performance overhead when using Linked Data based conceptual views over ODBC or JDBC accessible data sources.

Note: Using these options when the RFViews have already been created is not currently possible via the Conductor UI. Instead you should be able to add them manually from isql:

  1. Drop the RDF View graph and Quad Map.
  2. Create it again with the RDB2RDF Triggers options enabled.
See Also:

RDB2RDF Triggers


1.5.5. How to Manage Date Range SPARQL queries?

The following examples demonstrate how to manage date range in a SPARQL query:

Example with date range

SELECT ?s ?date 
FROM <http://dbpedia.org> 
WHERE 
  { 
    ?s ?p ?date . FILTER ( ?date >= "19450101"^^xsd:date && ?date <= "19451231"^^xsd:date )  
  } 
LIMIT 100

Example with bif:contains

Suppose there is the following query using bif:contains for date:

If ?date is of type xsd:date or xsd:dateTime and of valid syntax then bif:contains(?date, '"1945*"' ) will not found it, because it will be parsed at load/create and stored as SQL DATE value.

So if data are all accurate and typed properly then the filter is:

(?date >= xsd:date("1945-01-01") && ?date < xsd:date("1946-01-01"))	

i.e. the query should be:

SELECT DISTINCT ?s ?date
FROM <http://dbpedia.org>
WHERE
  {
    ?s ?p ?date . FILTER( ?date >= xsd:date("1945-01-01") && ?date < xsd:date("1946-01-01")  && (str(?p) != str(rdfs:label)) )
  }
LIMIT 10	

If data falls, then the free-text will be OK for tiny examples but not for "big" cases because bif:contains(?date, '"1945*"') would require that less than 200 words in the table begins with 1945. Still, some data can be of accurate type and syntax so range comparison should be used for them and results aggregated via UNION.

If dates mention timezones then the application can chose the beginning and the end of the year in some timezones other than the default.


1.5.6. How can I see which quad storages exist and in which quad storage a graph resides?

Let's take for example a created RDF view from relational data in Virtuoso. The RDF output therefor should have two graphs which reside in a quad storage named for ex.:

http://localhost:8890/rdfv_demo/quad_storage/default	

Also the RDF is accessible over the SPARQL endpoint with the following query:

define input:storage <http://localhost:8890/rdfv_demo/quad_storage/default>
SELECT * 
WHERE 
  { 
    ?s ?p ?o 
  }	

Now one could ask is there a way to define internally (once) that the quad storage should be included in queries to the SPARQL endpoint? So that the user does not have to define the input:storage explicitly in each query, like this:

http://localhost:8890/sparql?query=select * where { ?s ?p ?o }&default-graph-uri=NULL&named-graph-uri=NULL	

All metadata about all RDF storages are kept in "system" graph <http://www.openlinksw.com/schemas/virtrdf#> ( namespace prefix virtrdf: ). Subjects of type virtrdf:QuadStorage are RDF storages. There are three of them by default:

SQL> SPARQL SELECT * FROM virtrdf: WHERE { ?s a virtrdf:QuadStorage };
s
VARCHAR
_______________________________________________________________________________

http://www.openlinksw.com/schemas/virtrdf#DefaultQuadStorage
http://www.openlinksw.com/schemas/virtrdf#DefaultServiceStorage
http://www.openlinksw.com/schemas/virtrdf#SyncToQuads

3 Rows. -- 3 msec.	

There are two ways of using the RDF View from above in SPARQL endpoint without define input:storage:

  1. Create RDF View right in virtrdf:DefaultQuadStorage or add the view in other storage and then copy it from there to virtrdf:DefaultQuadStorage.
    • In any of these two variants, use:
      SPARQL ALTER QUAD STORAGE virtrdf:DefaultQuadStorage . . .	
      
  2. Use SYS_SPARQL_HOST table as described here and set SH_DEFINES so it contains your favorite define input:storage.

1.5.7. Can I drop and re-create the DefaultQuadStorage?

Currently system metadata consist of three "levels":

  1. QuadMapFormats are used to describe transformations of individual SQL values (or types of SQL values),
  2. QuadMaps refers to QuadMapFormats (via QuadMapValues) and describe some "minimal" independent RDB2RDF transformations,
  3. QuadStorages organizes QuadMaps.

QuadStorages contains only "symlinks" to maps, if you drop a storage you don't drop all mappings inside. If you drop the DefaultQuadStorage or some other built-in thing, it can be safely recovered by DB.DBA.RDF_AUDIT_METADATA, with first parameter set to 1. This will keep your own data intact. However we recommend to write a script that declares all your formats, Linked Data Views and storages, to be able to reproduce the configuration after any failures.


1.5.8. How to display only some information from RDF graph?

Virtuoso supports graph-level security, as described here but not subject-level or predicate-level. When exposing data that needs protected access, triples should be confined to private name graphs which are protected by ACLs using WebID.

Note, how you can use WebID to protect Virtuoso SPARQL endpoints.

See Also:

RDF Graphs Security


1.5.9. Is it possible to have the SPARQL endpoint on a different port than the Conductor?

Virtuoso Web Server has the capability to create extra listeners using the Conductor interface.

  1. At install time you have your HTTP Server port in your virtuoso.ini set to 8890, which you want to keep in your local network as this contains ALL the endpoints that you have registered in Virtuoso. So as long as you do not open this port in your firewall, you can only get at it from the local machine.
  2. Next you should create a new vhost entry using the EXTERNAL name of your machine and use port 80 (or a higher port if you do not want to run as root) for ex.:
    Interface: 0.0.0.0
    Port: 8080
    Http Host:  my.example.com	
    
  3. Next you add a "New directory to this line", click on "Type" radio button and choose "Sparql access point" from the drop-down list and press Next button. Set "Path" to /sparql and press the "Save Changes" button to store.
  4. At this point you have created: http://my.example.com:8080/sparql which functions exactly the same as your internal http://localhost:8890/sparql. You can now open your firewall and allow outside machines to connect to port 8080 so people can use your SPARQL endpoint without access to any other endpoint on your Virtuoso installation.
  5. You should probably also change your virtuoso.ini so:
    [URIQA]
    DefaultHost = my.example.com:8080	
    
  6. If you use port 80, you do not have to add :80 at the end of this setting, although it should not make any difference.
  7. You can now add other directories / endpoints to the new my.example.com interface you just created e.g. a nice / directory that points to a index.html which describes your site etc.
See Also:

Internet Domains


1.5.10. How to enable the Virtuoso Entity Framework 3.5 ADO.Net Provider in Visual Studio 2010?

The Virtuoso Entity Framework 3.5 ADO.Net Provider is current only list as a Visible control in the Visual Studio 2008 IDE as the current installers only create the necessary registry entries for Visual Studio 2008. To make it visible in the Visual Studio 2010's IDE the following registry settings need to be manually updated and manual addtions to some of the VS 2010 XML configuration files:

  1. Export following registry keys to .reg files and using a text editor, such as Wordpad, edit the Visual Studio version numbers from 8.0 or 9.0 to 10.0:
    HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\VisualStudio\8.0\DataProviders\{EE00F82B-C5A4-4073-8FF1-33F815C9801D}
    - and -
    HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\VisualStudio\8.0\DataSources\{90FBCAF2-8F42-47CD-BF1A-88FF41173060}
    
  2. Once edited, save and double click them to create the new registry entries under:
    HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\VisualStudio\10.0....	
    
  3. In addition, locate the file C:\WINDOWS\Microsoft.NET\Framework\v2.0.50727\CONFIG\machine.config, locate the node, then locate the <add name="VirtuosoClient3? Data Provider" ... node within it:
    <add name="VirtuosoClient3 Data Provider" invariant="OpenLink.Data.Virtuoso"
        description=".NET Framework Data Provider for Virtuoso" type="OpenLink.Data.Virtuoso.VirtuosoClientFactory, virtado3, Version=6.2.3128.2, Culture=neutral, PublicKeyToken=391bf132017ae989" />	
    

and copy is to the equivalent C:\WINDOWS\Microsoft.NET\Frameworks\v4.0.30128\CONFIG\machine.config location.

Visual Studio 2010 will then have the necessary information to locate and load the Virtuoso ADO.Net provider in its IDE.

The registry should typically contain the following entries for Visual Studio 2010 as a result:

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}]
@=".NET Framework Data Provider for Virtuoso"
"AssociatedSource"="{4D90D7C5-69A6-43EE-83ED-59A0E442D260}"
"CodeBase"="C:\\Windows\\assembly\\GAC_MSIL\\virtado3\\6.2.3128.1__391bf132017ae989\\virtado3.dll"
"Description"="Provider_Description, OpenLink.Data.Virtuoso.DDEX.Net3.DDEXResources, virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989"
"DisplayName"="Provider_DisplayName, OpenLink.Data.Virtuoso.DDEX.Net3.DDEXResources, virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989"
"InvariantName"="OpenLink.Data.Virtuoso"
"PlatformVersion"="2.0"
"ShortDisplayName"="Provider_ShortDisplayName, OpenLink.Data.Virtuoso.DDEX.Net3.DDEXResources, virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989"
"Technology"="{77AB9A9D-78B9-4ba7-91AC-873F5338F1D2}"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects]

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IDSRefBuilder]
@="Microsoft.VisualStudio.Data.Framework.DSRefBuilder"
"Assembly"="Microsoft.VisualStudio.Data.Framework, Version=10.0.0.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataAsyncCommand]

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataCommand]

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataConnectionProperties]
@="OpenLink.Data.Virtuoso.DDEX.Net3.VirtuosoDataConnectionProperties"
"Assembly"="virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataConnectionSupport]
@="Microsoft.VisualStudio.Data.Framework.AdoDotNet.AdoDotNetConnectionSupport"
"Assembly"="Microsoft.VisualStudio.Data.Framework, Version=10.0.0.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataConnectionUIControl]
@="OpenLink.Data.Virtuoso.DDEX.Net3.VirtuosoDataConnectionUIControl"
"Assembly"="virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataConnectionUIProperties]
@="OpenLink.Data.Virtuoso.DDEX.Net3.VirtuosoDataConnectionProperties"
"Assembly"="virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataMappedObjectConverter]

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataObjectIdentifierResolver]
@="OpenLink.Data.Virtuoso.DDEX.Net3.VirtuosoDataObjectIdentifierResolver"
"Assembly"="virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataObjectSupport]
@="Microsoft.VisualStudio.Data.Framework.DataObjectSupport"
"Assembly"="Microsoft.VisualStudio.Data.Framework, Version=10.0.0.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a"
"XmlResource"="OpenLink.Data.Virtuoso.DDEX.Net3.VirtuosoObjectSupport"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataSourceInformation]
@="OpenLink.Data.Virtuoso.DDEX.Net3.VirtuosoDataSourceInformation"
"Assembly"="virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataTransaction]

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}\SupportedObjects\IVsDataViewSupport]
@="Microsoft.VisualStudio.Data.Framework.DataViewSupport"
"Assembly"="Microsoft.VisualStudio.Data.Framework, Version=10.0.0.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a"
"XmlResource"="OpenLink.Data.Virtuoso.DDEX.Net3.VirtuosoViewSupport"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataSources\{4D90D7C5-69A6-43EE-83ED-59A0E442D260}]
@="OpenLink Virtuoso Data Source"
"DefaultProvider"="{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}"

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataSources\{4D90D7C5-69A6-43EE-83ED-59A0E442D260}\SupportingProviders]

[HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\VisualStudio\10.0\DataSources\{4D90D7C5-69A6-43EE-83ED-59A0E442D260}\SupportingProviders\{0886A2BB-03D0-4E00-8A3D-F235A5DC0F6D}]
"Description"="DataSource_Description, OpenLink.Data.Virtuoso.DDEX.Net3.DDEXResources, virtado3, Version=6.2.3128.1, Culture=neutral, PublicKeyToken=391bf132017ae989" 	

The next Virtuoso releases, 6.3+ will support this new Visual Studio 2010 release out of the box.


1.5.11. How Can I Control the normalization of UNICODE3 accented chars in free-text index?

The normalization of UNICODE3 accented chars in free-text index can be controlled by setting up the configuration parameter XAnyNormalization in the virtuoso.ini config file, section [I18N]. This parameter controls whether accented UNICODE characters should be converted to their non-accented base variants at the very beginning of free-text indexing or parsing a free-text query string. The parameter's value is an integer that is bitmask with only 2 bits in use atm:

  1. 0: the default behavior, do not normalize anything, so "Jose" and "José" are two distinct words.
  2. 2: Any combined char is converted to its (smallest known) base. So "é" will lose its accent and become plain old ASCII "e".
  3. 3: This is equl to 1|2 and when set then performs both conversions. As a result, pair of base char and combinig char loses its second char and chars with accents will lose accents.

If the parameter is required at all, the needed value is probably 3. So the fragment of virtuoso.ini is:

[I18N]
XAnyNormalization=3

In some seldom case the value of 1 can be appropriate. The parameter should be set once before creating the database. If changed on the existing database, all free-text indexes that may contain non-ASCII data should be re-created. On a typical system, the parameter affects all text columns, XML columns, RDF literals and queries.

Strictly speaking, it affects not all of them but only items that use default "x-any" language or language derived from x-any such as "en" and "en-US" but if you haven't tried writing new C plugins for custom languages you should not look so deep.

As an example, with XAnyNormalization=3 once can get the following:

SQL>SPARQL 

INSERT IN <http://InternationalNSMs/>
   { <s> <sp> "Índio João Macapá Júnior Tôrres Luís Araújo José" ; 
     <ru> "Он добавил картошки, посолил и поставил аквариум на огонь" . }

INSERT INTO <http://InternationalNSMs/>, 2 (or less) triples -- done


SQL> DB.DBA.RDF_OBJ_FT_RULE_ADD (null, null, 'InternationalNSMs.wb');

Done. -- 0 msec.

SQL>vt_inc_index_db_dba_rdf_obj();

Done. -- 26 msec.

SQL>SPARQL 
SELECT * 
FROM <http://InternationalNSMs/> 
WHERE 
  {
    ?s ?p ?o 
  }
ORDER BY ASC (str(?o))

s  sp  Índio João Macapá Júnior Tôrres Luís Araújo José
s  ru  Он добавил картошки, посолил и поставил аквариум на огонь

2 Rows. -- 2 msec.

SQL> SPARQL 
SELECT * 
FROM <http://InternationalNSMs/> 
WHERE 
  { 
    ?s ?p ?o . ?o bif:contains "'Índio João Macapá Júnior Tôrres Luís Araújo José'" . 
  }

s  sp  Índio João Macapá Júnior Tôrres Luís Araújo José

1 Rows. -- 2 msec.

SQL>SPARQL 
SELECT * 
FROM <http://InternationalNSMs/> 
WHERE
  { 
    ?s ?p ?o . ?o bif:contains "'Indio Joao Macapa Junior Torres Luis Araujo Jose'" . }

s  sp  Índio João Macapá Júnior Tôrres Luís Araújo José

1 Rows. -- 1 msec.

SQL> SPARQL 
SELECT * 
FROM <http://InternationalNSMs/> 
WHERE 
  { 
    ?s ?p ?o . ?o bif:contains "'поставил аквариум на огонь'" . }

s  ru  Он добавил картошки, посолил и поставил аквариум на огонь

There was also request for function that normalizes characters in strings as free-text engine will do with XAnyNormalization=3 , the function will be provided as a separate patch and depends on this specific patch.

See Also:

Virtuoso Configuration File


1.5.12. How Can I define graph with virt:rdf_sponger option set to "on"?

Suppose we have the following scenario:

  1. Create Virtuoso user using Conductor.
  2. Create for the user a RDF Sink folder.
  3. In the properties of the RDF sink folder add the following virt:rdf_graph option:
    http://localhost:8080/DAV/home/dba/rdf_sink/	
    
  4. Set virt:rdf_sponger to "on".
  5. Upload RDF files to the RDF Sink folder.
  6. As result the RDF data should be stored in graph for ex. (depending on your folder name etc.):
    http://local.virt/DAV/home/wa_sink/rdf_sink/wa_address_agents.rdf	
    
  7. In order to define any graph you want with the option from above, you should execute:
    SQL>DAV_PROP_SET (davLocation,  'virt:rdf_graph', iri, _uid, _pwd);	
    
    • Calling this function uses the given IRI as the graph IRI when sponging stuff put in the rdf_sink/ collection davLocation.
  8. Finally you should execute the following command to get the RDF data from the new graph:
    SQL>SELECT DAV_PPROP_GET ('/DAV/home/dba/rdf_sink/', 'virt:rdf_graph','dba', 'dba');	
    

1.5.13. How do I use SPARUL to change a selection of property values from URI References to Literals?

Assume a given graph where triples are comprised of property values that are mixed across URI References and Typed Literals as exemplified by the results of the query below:

SELECT DISTINCT ?sa ?oa 
FROM <http://ucb.com/nbeabase>
WHERE 
  { 
    ?sa a <http://ucb.com/nbeabase/resource/Batch> .
    ?sa <http://ucb.com/nbeabase/resource/chemAbsNo> ?oa . FILTER regex(?oa, '-','i')
  }

You can use the following SPARUL pattern to harmonize the property values across relevant triples in a specific graph, as shown below:

SQL> SPARQL 
INSERT INTO GRAPH <http://ucb.com/nbeabase> 
  { 
    ?sa <http://ucb.com/nbeabase/resource/sampleId> `str (?oa)` 
  }  
WHERE 
  { 
    ?sa <http://ucb.com/nbeabase/resource/chemAbsNo> ?oa . FILTER regex(?oa, '-','i')   
  }	

1.5.14. How is a Checkpoint performed against a Virtuoso Clustered Server?

The cluster cl_exec() function is used to perform a checkpoint across all node of a Virtuoso cluster as follows:

SQL>cl_exec ('checkpoint'); 	

This typically needs to be run after loading RDF datasets into a Virtuoso cluster to prevent lose of data when the cluster is restarted.


1.5.15. How can I use CONSTRUCT with PreparedStatements?

Assume a given query which uses pragma output:format '_JAVA_' with CONSTRUCT:

SPARQL DEFINE output:format '_JAVA_'
   CONSTRUCT { ?s ?p ?o }
WHERE
  { 
    ?s ?p ?o . 
    FILTER (?s = iri(?::0)) 
  }
LIMIT 1

In order to work correctly, the query should be modified to:

SPARQL DEFINE output:format '_JAVA_'
   CONSTRUCT { `iri(?::0)` ?p ?o }
WHERE
  { 
    `iri(?::0)` ?p ?o 
  }
LIMIT 1	

Equivalent variant of the query is also:

SPARQL DEFINE output:format '_JAVA_'
  CONSTRUCT { ?s ?p ?o }
WHERE
  { 
    ?s ?p ?o . 
    FILTER (?s = iri(?::0)) 
  }
LIMIT 1	

1.5.16. How can perform SPARQL Updates without transactional log size getting exceed?

Since SPARUL updates are generally not meant to be transactional, it is best to run these in:

SQL> log_enable (2);	

mode, which commits every operation as it is done. This prevents one from running out of rollback space. Also for bulk updates, transaction logging can be turned off. If so, one should do a manual checkpoint after the operation to ensure persistence across server restart since there is no roll forward log.

Alternatively, the "TransactionAfterImageLimit" parameter can be set in the virtuoso.ini config file to a higher value than its 50MB default:

#virtuoso.ini
...
[Parameters]
...
TransactionAfterImageLimit = N bytes default 50000000
...	
See Also:

Using SPARUL

Virtuoso INI Parameters


1.5.17. How can I write custom crawler using PL?

The following code is an example of loading data via crawler with special function to generate link for downloading:

create procedure EUROPEANA_STORE ( 
  in _host varchar, 
  in _url varchar, 
  in _root varchar, 
  inout _content varchar, 
  in _s_etag varchar, 
  in _c_type varchar, 
  in store_flag int := 1, 
  in udata any := null, 
  in lev int := 0)
{
   declare url varchar;
   declare xt, xp any;
   declare old_mode int;
   xt := xtree_doc (_content, 2);
   xp := xpath_eval ('//table//tr/td/a[@href]/text()', xt, 0);
   commit work;
   old_mode := log_enable (3,1);
   foreach (any u in xp) do
     {
       u := cast (u as varchar);
       url := sprintf ('http://semanticweb.cs.vu.nl/europeana/api/export_graph?graph=%U&mimetype=default&format=turtle', u);
       dbg_printf ('%s', u);
	 {
	   declare continue handler for sqlstate '*' { 
	     dbg_printf ('ERROR: %s', __SQL_MESSAGE);
	   };
	   SPARQL LOAD ?:url into GRAPH ?:u;
	 }
     }
   log_enable (old_mode, 1);
   return WS.WS.LOCAL_STORE (_host, _url, _root, _content, _s_etag, _c_type, store_flag, 0);
}	
See Also:

Set Up the Content Crawler to Gather RDF


1.5.18. How Can I Get an exact mapping for a date?

Assume a given attempts to get an exact mapping for the literal "1950" using bif:contains:

SPARQL
SELECT DISTINCT ?s ?o 
FROM <http://dbpedia.org> 
WHERE 
  {
    ?s ?p ?o . 
   FILTER( bif:contains (?o, '"1950"') 
           && isLiteral(?o) 
           && ( str(?p) ! = rdfs:label  || str(?p) !=  foaf:name 
           && ( ?o='1950')
         )
  }	

As an integer 1950 or date 1950-01-01 are not texts, they are not in free-text index and thus invisible for CONTAINS free-text predicate.

A possible way to make them visible that way is to introduce an additional RDF predicate that will contain objects of the triples in question, converted to strings via str() function.

Thus better results will be approached: if searches about dates are frequent then a new predicate can have date/datetime values extracted from texts, eliminating the need for bif:contains.

Therefor, the query from above should be changed to:

SPARQL
SELECT DISTINCT ?s ?o 
FROM <http://dbpedia.org> 
WHERE 
  {
    ?s ?p ?o . 
    FILTER (  isLiteral(?o) 
              && (  str(?p) != str(rdfs:label) || str(?p) !=  foaf:name ) 
              && str(?o) in ("1950", "1950-01-01"))
  }

1.5.19. How Can I Get certificate attributes using SPARQL?

The SPARQL query should use the cert:hex and cert:decimal in order to get to the values, so for ex:

PREFIX cert: <http://www.w3.org/ns/auth/cert#> 
PREFIX rsa: <http://www.w3.org/ns/auth/rsa#> 

SELECT ?webid 
FROM <http://webid.myxwiki.org/xwiki/bin/view/XWiki/homepw4>
WHERE 
  {
    [] cert:identity ?webid ;
             rsa:modulus ?m ;
     rsa:public_exponent ?e .
     ?m cert:hex "b520f38479f5803a7ab33233155eeef8ad4e1f575b603f7780f3f60ceab1\n34618fbe117539109c015c5f959b497e67c1a3b2c96e5f098bb0bf2a6597\n779d26f55fe8d320de7af0562fd2cd067dbc9d775b22fc06e63422717d00\na6801dedafd7b54a93c3f4e59538475673972e524f4ec2a3667d0e1ac856\nd532e32bf30cef8c1adc41718920568fbe9f793daeeaeeaa7e8367b7228a\n895a6cf94545a6f6286693277a1bc7750425ce6c35d570e89453117b88ce\n24206afd216a705ad08b7c59\n"^^xsd:string .
     ?e cert:decimal "65537"^^xsd:string
  }	

1.5.20. How can I make Multi Thread Virtuoso connection using JDBC??

See details here.


1.5.21. How Do I Perform Bulk Loading of RDF Source Files into one or more Graph IRIs?

See details here.


1.5.22. How to exploit RDF Schema and OWL Inference Rules with minimal effort?

When you install Virtuoso, it's reasoner and highly scalable inference capabilities may not be obvious. Typical cases involve using rdfs:subClassOf predicates in queries and wondering why reasoning hasn't occurred in line with the semantics defined in RDF Schema.

The experience applies when using more sophisticated predicates from OWL such as owl:equivalentProperty, owl:equivalentClass, owl:sameAs, owl:SymmetricalProperty, owl:inverseOf etc ...

Virtuoso implemented inference rules processing in a loosely coupled manner that allow users to conditionally apply inference context (via rules) to SPARQL queries. Typically, you have to create these rules following steps outlined here.

This tips and tricks note provides a shortcut for setting up and exploring RDF Schema and OWL reasoning once you've installed the Virtuoso Faceted Browser VAD package.

See Also:

Inference Rules and Reasoning

Virtuoso Faceted Browser Installation and configuration

Virtuoso Faceted Web Service


1.5.23. How can I dump arbitrary query result as N-Triples?

Assume the following arbitrary query:

SPARQL define output:format "NT" 
CONSTRUCT { ?s a ?t } 
FROM virtrdf:
WHERE { ?s a ?t };	

For iteration over result-set of an arbitrary query, use exec_next() in a loop that begins with exec() with cursor output variable as an argument and ends with exec_close() after it is out of data.


1.5.24. How do I bind named graph parameter in prepared statement?

Assume the following SPARQL query:

CONSTRUCT 
  { 
    ?s ?p ?o
  } 
FROM ?context 
WHERE 
  { 
    ?s ?p ?o 
  }	

To bind the named graph context of the query from above, the best solution due to performance implications, is to change the syntax of the query as:

CONSTRUCT 
  { 
    ?s ?p ?o
  } 
WHERE 
  { 
    graph `iri(??)` { ?s ?p ?o } 
  }	

Note: In case of using "FROM clause", it needs a constant in order to check at the compile time whether the IRI refers to a graph or a graph group:


1.5.25. How can I insert binary data to Virtuoso RDF storage in plain queries and with parameter binding via ADO.NET calls?

The following example shows different methods for insert binary data to Virtuoso RDF storage in plain queries and with parameter binding via ADO.NET calls:

# Test_Bin.cs


using System;
using System.Runtime.InteropServices;
using System.Text;
using System.Data;
using OpenLink.Data.Virtuoso;

#if ODBC_CLIENT
namespace OpenLink.Data.VirtuosoOdbcClient
#elif CLIENT
namespace OpenLink.Data.VirtuosoClient
#else
namespace OpenLink.Data.VirtuosoTest
#endif
{
    class Test_Bin
    {
        [STAThread]
        static void Main(string[] args)
        {
            IDataReader myread = null;
            IDbConnection c;

            c = new VirtuosoConnection("HOST=localhost:1111;UID=dba;PWD=dba;");

            IDbCommand cmd = c.CreateCommand();
            int ros;

            try
            {
                c.Open();

                cmd.CommandText = "sparql clear graph <ado.bin>";
                cmd.ExecuteNonQuery();

//insert binary as base64Binary
                cmd.CommandText = "sparql insert into graph <ado.bin> { <res1> <attr> \"GpM7\"^^<http://www.w3.org/2001/XMLSchema#base64Binary> }";
                cmd.ExecuteNonQuery();

//insert binary as hexBinary
                cmd.CommandText = "sparql insert into graph <ado.bin> { <res2> <attr> \"0FB7\"^^<http://www.w3.org/2001/XMLSchema#hexBinary> }";
                cmd.ExecuteNonQuery();


//prepare for insert with parameter binding
                cmd.CommandText = "sparql define output:format '_JAVA_' insert into graph <ado.bin> { `iri($?)` <attr> `bif:__rdf_long_from_batch_params($?,$?,$?)` }";

//bind parameters for insert binary as base64Binary
                IDbDataParameter param = cmd.CreateParameter();
                param.ParameterName = "p1";
                param.DbType = DbType.AnsiString;
                param.Value = "res3";
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p2";
                param.DbType = DbType.Int32;
                param.Value = 4;
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p3";
                param.DbType = DbType.AnsiString;
                param.Value = "GpM7";
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p4";
                param.DbType = DbType.AnsiString;
                param.Value = "http://www.w3.org/2001/XMLSchema#base64Binary";
                cmd.Parameters.Add(param);

                cmd.ExecuteNonQuery();

                cmd.Parameters.Clear();

//bind parameters for insert binary as hexBinary
                param = cmd.CreateParameter();
                param.ParameterName = "p1";
                param.DbType = DbType.AnsiString;
                param.Value = "res4";
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p2";
                param.DbType = DbType.Int32;
                param.Value = 4;
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p3";
                param.DbType = DbType.AnsiString;
                param.Value = "0FB7";
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p4";
                param.DbType = DbType.AnsiString;
                param.Value = "http://www.w3.org/2001/XMLSchema#hexBinary";
                cmd.Parameters.Add(param);

                cmd.ExecuteNonQuery();

                cmd.Parameters.Clear();

//bind parameters for insert binary as byte[]
                param = cmd.CreateParameter();
                param.ParameterName = "p1";
                param.DbType = DbType.AnsiString;
                param.Value = "res5";
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p2";
                param.DbType = DbType.Int32;
                param.Value = 3;
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p3";
                param.DbType = DbType.Binary;
                byte[] bin_val = {0x01, 0x02, 0x03, 0x04, 0x05};
                param.Value = bin_val;
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p4";
                param.DbType = DbType.AnsiString;
                param.Value = System.DBNull.Value; 
                cmd.Parameters.Add(param);

                cmd.ExecuteNonQuery();

                cmd.Parameters.Clear();

//execute select and check the results
                cmd.CommandText = "sparql SELECT ?s ?o FROM <ado.bin> WHERE {?s ?p ?o}"; ;
                myread = cmd.ExecuteReader();
                int r = 0;

                while (myread.Read())
                {
                    Console.WriteLine("=== ROW === "+r);
                    for (int i = 0; i < myread.FieldCount; i++)
                    {
                        string s;
                        if (myread.IsDBNull(i))
                            Console.Write("N/A|\n");
                        else
                        {
                            object o = myread.GetValue(i);
                            Type t = myread.GetFieldType(i);

                            s = myread.GetString(i);
                            Console.Write(s + "[");
                            if (o is SqlExtendedString)
                            {
                                SqlExtendedString se = (SqlExtendedString)o;
                                Console.Write("IriType=" + se.IriType + ";StrType=" + se.StrType + ";Value=" + se.ToString());
                                Console.Write(";ObjectType=" + o.GetType() + "]|\n");
                            }
                            else if (o is SqlRdfBox)
                            {
                                SqlRdfBox se = (SqlRdfBox)o;
                                Console.Write("Lang=" + se.StrLang + ";Type=" + se.StrType + ";Value=" + se.Value);
                                Console.Write(";ObjectType=" + o.GetType() + "]|\n");
                                object v = se.Value;
                                if (v is System.Byte[])
                                {
                                    byte[] vb = (byte[])v;
                                    for (int z = 0; z < vb.Length; z++)
                                    {
                                        Console.WriteLine(""+z+"="+vb[z]);
                                    }
                                }
                            }
                            else
                                Console.Write(o.GetType() + "]|\n");
                        }
                    }
                    r++;
                }
            }
            catch (Exception e)
            {
                Console.WriteLine("{0} Exception caught.", e);
            }
            finally
            {
                //		  if (myread != null)
                //		    myread.Close();

                if (c.State == ConnectionState.Open)
                    c.Close();

            }

        }
    }
}

Output log for example is in the log.txt:

# log.txt
=== ROW === 0
res1[IriType=IRI;StrType=IRI;Value=res1;ObjectType=OpenLink.Data.Virtuoso.SqlExtendedString]|
GpM7[Lang=;Type=http://www.w3.org/2001/XMLSchema#base64Binary;Value=GpM7;ObjectType=OpenLink.Data.Virtuoso.SqlRdfBox]|
=== ROW === 1
res2[IriType=IRI;StrType=IRI;Value=res2;ObjectType=OpenLink.Data.Virtuoso.SqlExtendedString]|
0FB7[Lang=;Type=http://www.w3.org/2001/XMLSchema#hexBinary;Value=0FB7;ObjectType=OpenLink.Data.Virtuoso.SqlRdfBox]|
=== ROW === 2
res3[IriType=IRI;StrType=IRI;Value=res3;ObjectType=OpenLink.Data.Virtuoso.SqlExtendedString]|
GpM7[Lang=;Type=http://www.w3.org/2001/XMLSchema#base64Binary;Value=GpM7;ObjectType=OpenLink.Data.Virtuoso.SqlRdfBox]|
=== ROW === 3
res4[IriType=IRI;StrType=IRI;Value=res4;ObjectType=OpenLink.Data.Virtuoso.SqlExtendedString]|
0FB7[Lang=;Type=http://www.w3.org/2001/XMLSchema#hexBinary;Value=0FB7;ObjectType=OpenLink.Data.Virtuoso.SqlRdfBox]|
=== ROW === 4
res5[IriType=IRI;StrType=IRI;Value=res5;ObjectType=OpenLink.Data.Virtuoso.SqlExtendedString]|
0102030405[System.Byte[]]|	

1.5.26. How can I insert RDF data from Visual Studio to Virtuoso?

The following example shows how to insert RDF Data from Visual Studio to Virtuoso:

    
using System;
using System.Runtime.InteropServices;
using System.Text;
using System.Data;
using OpenLink.Data.Virtuoso;

#if ODBC_CLIENT
namespace OpenLink.Data.VirtuosoOdbcClient
#elif CLIENT
namespace OpenLink.Data.VirtuosoClient
#else
namespace OpenLink.Data.VirtuosoTest
#endif
{
    class Test_Insert
    {
        [STAThread]
        static void Main(string[] args)
        {
            IDataReader myread = null;
            IDbConnection c;

            c = new VirtuosoConnection("HOST=localhost:1111;UID=dba;PWD=dba;Charset=UTF-8");

            IDbCommand cmd = c.CreateCommand();
            int ros;

            try
            {
                c.Open();

                cmd.CommandText = "sparql clear graph <ado.net>";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P01> \"131\"^^<http://www.w3.org/2001/XMLSchema#short> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P02> \"1234\"^^<http://www.w3.org/2001/XMLSchema#integer> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P03> \"12345.12\"^^<http://www.w3.org/2001/XMLSchema#float> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P04> \"123456.12\"^^<http://www.w3.org/2001/XMLSchema#double> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P05> \"123456.12\"^^<http://www.w3.org/2001/XMLSchema#decimal> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P06> \"01020304\"^^<http://www.w3.org/2001/XMLSchema#hexBinary> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P07> \"01.20.1980T04:51:13\"^^<http://www.w3.org/2001/XMLSchema#dateTime> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P08> \"01.20.1980\"^^<http://www.w3.org/2001/XMLSchema#date> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P09> \"01:20:19\"^^<http://www.w3.org/2001/XMLSchema#time> }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql insert into graph <ado.net> { <a> <P10> \"test\" }";
                cmd.ExecuteNonQuery();

                cmd.CommandText = "sparql define output:format '_JAVA_' insert into graph <ado.net> { <b> `iri($?)` `bif:__rdf_long_from_batch_params($?,$?,$?)` }";

//add Object URI
                add_triple(cmd, "S01", 1, "test1", null);

//add Object BNode
                add_triple(cmd, "S02", 1, "_:test2", null);


//add Literal
                add_triple(cmd, "S03", 3, "test3", null);

//add Literal with Datatype
                add_triple(cmd, "S04", 4, "1234", "http://www.w3.org/2001/XMLSchema#integer");

//add Literal with Lang
                add_triple(cmd, "S05", 5, "test5", "en");


                add_triple(cmd, "S06", 3, (short)123, null);
                add_triple(cmd, "S07", 3, 1234, null);
                add_triple(cmd, "S08", 3, (float)12345.12, null);
                add_triple(cmd, "S09", 3, 123456.12, null);
                add_triple(cmd, "S10", 3, new DateTime(2001, 02, 23, 13, 44, 51, 234), null);
                add_triple(cmd, "S11", 3, new DateTime(2001, 02, 24), null);
                add_triple(cmd, "S12", 3, new TimeSpan(19, 41, 23), null);
 

                add_triple(cmd, "S13", 4, "GpM7", "http://www.w3.org/2001/XMLSchema#base64Binary");
                add_triple(cmd, "S14", 4, "0FB7", "http://www.w3.org/2001/XMLSchema#hexBinary");
                byte[] bin_val = { 0x01, 0x02, 0x03, 0x04, 0x05 };
                add_triple(cmd, "S15", 3, bin_val, null);

            }
            catch (Exception e)
            {
                Console.WriteLine("{0} Exception caught.", e);
            }
            finally
            {

                if (c.State == ConnectionState.Open)
                    c.Close();

            }

        }


        static void add_triple(IDbCommand cmd, string sub, int ptype, object val, string val_add)
        {
                cmd.Parameters.Clear();

                IDbDataParameter param = cmd.CreateParameter();
                param.ParameterName = "p1";
                param.DbType = DbType.AnsiString;
                param.Value = sub;
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p2";
                param.DbType = DbType.Int32;
                param.Value = ptype;
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p3";
                if (val != null && val.GetType() == typeof (System.String))
                    param.DbType = DbType.AnsiString;
                param.Value = val;
                cmd.Parameters.Add(param);

                param = cmd.CreateParameter();
                param.ParameterName = "p4";
                param.DbType = DbType.AnsiString;
                param.Value = val_add;
                cmd.Parameters.Add(param);

                cmd.ExecuteNonQuery();
        }



    }
}	
See Also:

RDF Insert Methods in Virtuoso


1.5.27. How does default describe mode work?

The default describe mode works only if subject has a type/class. To get any anonymous subject described CBD mode should be used:

 
$ curl -i -L -H "Accept: application/rdf+xml" http://lod.openlinksw.com/describe/?url=http%3A%2F%2Fwww.mpii.de%2Fyago%2Fresource%2FEddie%255FMurphy
HTTP/1.1 303 See Other
Date: Mon, 21 Mar 2011 14:24:36 GMT
Server: Virtuoso/06.02.3129 (Linux) x86_64-generic-linux-glibc25-64  VDB
Content-Type: text/html; charset=UTF-8
Accept-Ranges: bytes
TCN: choice
Vary: negotiate,accept,Accept-Encoding
Location: http://lod.openlinksw.com/sparql?query=%20DESCRIBE%20%3Chttp%3A%2F%2Fwww.mpii.de%2Fyago%2Fresource%2FEddie%255FMurphy%3E&format=application%
2Frdf%2Bxml
Content-Length: 0

HTTP/1.1 200 OK
Date: Mon, 21 Mar 2011 14:24:37 GMT
Server: Virtuoso/06.02.3129 (Linux) x86_64-generic-linux-glibc25-64  VDB
Accept-Ranges: bytes
Content-Type: application/rdf+xml; charset=UTF-8
Content-Length: 467967

<?xml version="1.0" encoding="utf-8" ?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#">
<rdf:Description rdf:about="http://www.mpii.de/yago/resource/MTV%5FMovie%5FAward%5Ffor%5FBest%5FComedic%5FPerformance"><n0pred:hasInternalWikipediaLinkTo xmlns:n0pred="http://www.mpii.de/yago/resource/" rdf:resource="http://www.mpii.de/yago/resource/Eddie%5FMurphy"/></rdf:Description>
<rdf:Description rdf:about="http://www.mpii.de/yago/resource/#fact_23536547421"><rdf:subject rdf:resource="http://www.mpii.de/yago/resource/Eddie%5FMurphy"/></rdf:Description>
<rdf:Description rdf:about="http://www.mpii.de/yago/resource/#fact_23536896725"><rdf:object rdf:resource="http://www.mpii.de/yago/resource/Eddie%5FMurphy"/></rdf:Description>
...

1.5.28. What should I do if the Virtuoso Server is not responding to HTTP requests?

Assume the Virtuoso server is not responding to HTTP requests although SQL connection is working. In order to determine what activity is being performed that might account for this:

  1. Check the status:
     
    SQL> status('');
    REPORT
    VARCHAR
    _______________________________________________________________________________
    
    
    OpenLink Virtuoso VDB Server
    Version 06.02.3129-pthreads for Linux as of Mar 16 2011 
    Registered to Uriburner (Personal Edition, unlimited connections)
    Started on: 2011/03/17 10:49 GMT+60
     
    Database Status:
      File size 0, 37598208 pages, 7313125 free.
      1000000 buffers, 993399 used, 76771 dirty 0 wired down, repl age 25548714 0 w. io 0 w/crsr.
      Disk Usage: 2642884 reads avg 4 msec, 30% r 0% w last  1389 s, 1557572 writes,
        15331 read ahead, batch = 79.  Autocompact 308508 in 219226 out, 28% saved.
    Gate:  71130 2nd in reads, 0 gate write waits, 0 in while read 0 busy scrap. 
    Log = virtuoso.trx, 14922248 bytes
    VDB: 0 exec 0 fetch 0 transact 0 error
    1757362 pages have been changed since last backup (in checkpoint state)
    Current backup timestamp: 0x0000-0x00-0x00
    Last backup date: unknown
    Clients: 5 connects, max 2 concurrent
    RPC: 116 calls, -1 pending, 1 max until now, 0 queued, 2 burst reads (1%), 0 second brk=9521074176
    Checkpoint Remap 331113 pages, 0 mapped back. 1180 s atomic time.
        DB master 37598208 total 7313125 free 331113 remap 40593 mapped back
       temp  569856 total 569851 free
     
    Lock Status: 52 deadlocks of which 0 2r1w, 86078 waits,
       Currently 1 threads running 0 threads waiting 0 threads in vdb.
    Pending:
    
    
    25 Rows. -- 1274 msec.
    SQL> 
    	
    
  2. Connect with the PL debugger and see what is running currently using the info threads call:
     
    $ isql 1111 dba <password> -D
    DEBUG> info threads	
    
  3. This should return the current code being executed by the Sever.
  4. Run txn_killall() to kill any pending transactions which may enable the server to start responding to HTTP requests again:
     
    
    SQL> txn_killall();
    
    Done. -- 866 msec.	
    

1.5.29. What CXML params are supported for the SPARQL URL pattern?

The following options are supported for CXML link behavior in the SPARQL URL Pattern:

  1. Local faceted navigation links:
     
    # CXML_redir_for_subjs=&CXML_redir_for_hrefs=&: 
    
    http://lod.openlinksw.com/sparql?default-graph-uri=&should-sponge=&query=SELECT+DISTINCT+%3Fcafe+%3Flat+%3Flong+%3Fcafename+%3Fchurchname++%28+bif%3Around%28bif%3Ast_distance+%28%3Fchurchgeo%2C+%3Fcafegeo%29%29+%29++WHERE++{++++%3Fchurch+a+lgv%3Aplace_of_worship+.++++%3Fchurch+geo%3Ageometry+%3Fchurchgeo+.++++%3Fchurch+lgv%3Aname+%3Fchurchname+.++++%3Fcafe+a+lgv%3Acafe+.++++%3Fcafe+lgv%3Aname+%3Fcafename+.++++%3Fcafe+geo%3Ageometry+%3Fcafegeo+.++++%3Fcafe+geo%3Alat+%3Flat+.++++%3Fcafe+geo%3Along+%3Flong+.++++FILTER+%28+bif%3Ast_intersects+%28%3Fchurchgeo%2C+bif%3Ast_point+%282.3498%2C48.853%29%2C5%29++%26%26+bif%3Ast_intersects+%28%3Fcafegeo%2C+%3Fchurchgeo%2C+0.2%29+%29+}+LIMIT+50&debug=on&timeout=&format=text%2Fhtml&CXML_redir_for_subjs=&CXML_redir_for_hrefs=&save=display&fname=
    
  2. External Resource Links:
     
    # CXML_redir_for_subjs=&CXML_redir_for_hrefs=121: 
    
    http://lod.openlinksw.com/sparql?default-graph-uri=&should-sponge=&query=SELECT+DISTINCT+%3Fcafe+%3Flat+%3Flong+%3Fcafename+%3Fchurchname++%28+bif%3Around%28bif%3Ast_distance+%28%3Fchurchgeo%2C+%3Fcafegeo%29%29+%29++WHERE++{++++%3Fchurch+a+lgv%3Aplace_of_worship+.++++%3Fchurch+geo%3Ageometry+%3Fchurchgeo+.++++%3Fchurch+lgv%3Aname+%3Fchurchname+.++++%3Fcafe+a+lgv%3Acafe+.++++%3Fcafe+lgv%3Aname+%3Fcafename+.++++%3Fcafe+geo%3Ageometry+%3Fcafegeo+.++++%3Fcafe+geo%3Alat+%3Flat+.++++%3Fcafe+geo%3Along+%3Flong+.++++FILTER+%28+bif%3Ast_intersects+%28%3Fchurchgeo%2C+bif%3Ast_point+%282.3498%2C48.853%29%2C5%29++%26%26+bif%3Ast_intersects+%28%3Fcafegeo%2C+%3Fchurchgeo%2C+0.2%29+%29+}+LIMIT+50&debug=on&timeout=&format=text%2Fhtml&CXML_redir_for_subjs=&CXML_redir_for_hrefs=121&save=display&fname=
    
    
  3. External faceted navigation links:
     
    # CXML_redir_for_subjs=&CXML_redir_for_hrefs=LOCAL_PIVOT: 
    
    http://lod.openlinksw.com/sparql?default-graph-uri=&should-sponge=&query=SELECT+DISTINCT+%3Fcafe+%3Flat+%3Flong+%3Fcafename+%3Fchurchname++%28+bif%3Around%28bif%3Ast_distance+%28%3Fchurchgeo%2C+%3Fcafegeo%29%29+%29++WHERE++{++++%3Fchurch+a+lgv%3Aplace_of_worship+.++++%3Fchurch+geo%3Ageometry+%3Fchurchgeo+.++++%3Fchurch+lgv%3Aname+%3Fchurchname+.++++%3Fcafe+a+lgv%3Acafe+.++++%3Fcafe+lgv%3Aname+%3Fcafename+.++++%3Fcafe+geo%3Ageometry+%3Fcafegeo+.++++%3Fcafe+geo%3Alat+%3Flat+.++++%3Fcafe+geo%3Along+%3Flong+.++++FILTER+%28+bif%3Ast_intersects+%28%3Fchurchgeo%2C+bif%3Ast_point+%282.3498%2C48.853%29%2C5%29++%26%26+bif%3Ast_intersects+%28%3Fcafegeo%2C+%3Fchurchgeo%2C+0.2%29+%29+}+LIMIT+50&debug=on&timeout=&format=text%2Fhtml&CXML_redir_for_subjs=&CXML_redir_for_hrefs=LOCAL_PIVOT&save=display&fname=
    
  4. External description resource(TTL):
     
    # CXML_redir_for_subjs=&CXML_redir_for_hrefs=LOCAL_TTL:
    
    http://lod.openlinksw.com/sparql?default-graph-uri=&should-sponge=&query=SELECT+DISTINCT+%3Fcafe+%3Flat+%3Flong+%3Fcafename+%3Fchurchname++%28+bif%3Around%28bif%3Ast_distance+%28%3Fchurchgeo%2C+%3Fcafegeo%29%29+%29++WHERE++{++++%3Fchurch+a+lgv%3Aplace_of_worship+.++++%3Fchurch+geo%3Ageometry+%3Fchurchgeo+.++++%3Fchurch+lgv%3Aname+%3Fchurchname+.++++%3Fcafe+a+lgv%3Acafe+.++++%3Fcafe+lgv%3Aname+%3Fcafename+.++++%3Fcafe+geo%3Ageometry+%3Fcafegeo+.++++%3Fcafe+geo%3Alat+%3Flat+.++++%3Fcafe+geo%3Along+%3Flong+.++++FILTER+%28+bif%3Ast_intersects+%28%3Fchurchgeo%2C+bif%3Ast_point+%282.3498%2C48.853%29%2C5%29++%26%26+bif%3Ast_intersects+%28%3Fcafegeo%2C+%3Fchurchgeo%2C+0.2%29+%29+}+LIMIT+50&debug=on&timeout=&format=text%2Fhtml&CXML_redir_for_subjs=&CXML_redir_for_hrefs=LOCAL_TTL&save=display&fname=
    
  5. External description resource(CXML):
     
    # CXML_redir_for_subjs=&CXML_redir_for_hrefs=LOCAL_CXML: 
    
    http://lod.openlinksw.com/sparql?default-graph-uri=&should-sponge=&query=SELECT+DISTINCT+%3Fcafe+%3Flat+%3Flong+%3Fcafename+%3Fchurchname++%28+bif%3Around%28bif%3Ast_distance+%28%3Fchurchgeo%2C+%3Fcafegeo%29%29+%29++WHERE++{++++%3Fchurch+a+lgv%3Aplace_of_worship+.++++%3Fchurch+geo%3Ageometry+%3Fchurchgeo+.++++%3Fchurch+lgv%3Aname+%3Fchurchname+.++++%3Fcafe+a+lgv%3Acafe+.++++%3Fcafe+lgv%3Aname+%3Fcafename+.++++%3Fcafe+geo%3Ageometry+%3Fcafegeo+.++++%3Fcafe+geo%3Alat+%3Flat+.++++%3Fcafe+geo%3Along+%3Flong+.++++FILTER+%28+bif%3Ast_intersects+%28%3Fchurchgeo%2C+bif%3Ast_point+%282.3498%2C48.853%29%2C5%29++%26%26+bif%3Ast_intersects+%28%3Fcafegeo%2C+%3Fchurchgeo%2C+0.2%29+%29+}+LIMIT+50&debug=on&timeout=&format=text%2Fhtml&CXML_redir_for_subjs=&CXML_redir_for_hrefs=LOCAL_CXML&save=display&fname=
    

1.5.30. How can I replicate all graphs?

To replicate all graphs ( except the system graph http://www.openlinksw.com/schemas/virtrdf# ), one should use http://www.openlinksw.com/schemas/virtrdf#rdf_repl_all as graph IRI:

 
SQL> DB.DBA.RDF_RDF_REPL_GRAPH_INS ('http://www.openlinksw.com/schemas/virtrdf#rdf_repl_all');

1.5.31. What is best method to get a random sample of all triples for a subset of all the resources of a SPARQL endpoint?

The best method to get a random sample of all triples for a subset of all the resources of a SPARQL endpoint, is decimation in its original style:

 
SELECT ?s ?p ?o 
FROM <some-graph>
WHERE 
  { 
    ?s ?p ?o .
    FILTER ( 1 > bif:rnd (10, ?s, ?p, ?o) )
  }

By tweaking first argument of bif:rnd() and the left side of the inequality you can tweak decimation ratio from 1/10 to the desired value. What's important is to know that the SQL optimizer has a right to execute bif:rnd (10) only once at the beginning of the query, so we had to pass additional three arguments that can be known only when a table row is fetched so bif:rnd (10, ?s, ?p, ?o) is calculated for every row and thus any given row is either returned or ignored independently from others.

However, bif:rnd (10, ?s, ?p, ?o) contains a subtle inefficiency. In RDF store, graph nodes are stored as numeric IRI IDs and literal objects can be stored in a separate table. The call of an SQL function needs arguments of traditional SQL datatypes, so the query processor will extract the text of IRI for each node and the full value for each literal object. That is significant waste of time. The workaround is:

 
SPARQL 
SELECT ?s ?p ?o 
FROM <some-graph> 
WHERE 
  { 
    ?s ?p ?o .
    FILTER ( 1>  <SHORT_OR_LONG::bif:rnd>  (10, ?s, ?p, ?o))  
  }

This tells the SPARQL front-end to omit redundant conversions of values.


1.5.32. How can I replicate all graphs?

To replicate all graphs ( except the system graph http://www.openlinksw.com/schemas/virtrdf# ), one should use http://www.openlinksw.com/schemas/virtrdf#rdf_repl_all as graph IRI:

 
SQL> DB.DBA.RDF_RDF_REPL_GRAPH_INS ('http://www.openlinksw.com/schemas/virtrdf#rdf_repl_all');

1.5.33. How can I use SPARQL to make Meshups?

The following example demonstrates how to use SPARQL in order to make Meshups:

 
PREFIX dbo: <http://dbpedia.org/ontology/>
PREFIX rtb: <http://www.openlinksw.com/schemas/oat/rdftabs#>

CONSTRUCT 
  {
    ?museum geo:geometry ?museumgeo ;
             rtb:useMarker 'star' ;
             foaf:name ?musname;
             rdfs:comment ?muscomment.
    ?edu geo:geometry ?edugeo ;
          rtb:useMarker 'book' ;
          foaf:name ?eduname;
          rdfs:comment ?educomment.
    ?wh geo:geometry ?whgeo;
            rtb:useMarker '03';
            foaf:name ?whname;
            rdfs:comment ?whcomment.
  }
WHERE 
  { 
    {
      ?museum a dbo:Museum;
              geo:geometry ?museumgeo;
              foaf:name ?musname;
              rdfs:comment ?muscomment.
      filter (lang(?musname)='en' && lang(?muscomment)='en')
    } 
    UNION 
    {
      ?edu a dbo:University;
           geo:geometry ?edugeo;
           foaf:name ?eduname;
           rdfs:comment ?educomment.
      filter (lang(?eduname)='en' && lang(?educomment)='en')
    } 
    UNION 
    {
      ?wh a dbo:WorldHeritageSite;
            geo:geometry ?whgeo;
            rdfs:label ?whname;
            rdfs:comment ?whcomment. 
      filter (lang(?whname)='en' && lang(?whcomment)='en')
    } 
  }

1.5.34. How can I use the net_meter utility before starting the ingestion to a cluster?

The net_meter utility is a SQL procedure that runs a network test procedure on every host of the cluster. The network test procedure sends a message to every other host of the cluster and waits for the replies from each host. After the last reply is received the action is repeated. This results in a symmetrical load of the network, all points acting as both clients and servers to all other points.

net_meter ( 
  in n_threads int, 
  in n_batches int, 
  in bytes int, 
  in ops_per_batch int)

The parameters have the following meaning:

1.5.34.1. Example

Assume anuser has run 2 sets of tests on a cluster:

The first one was 1 thread, 1000 batches, 1000 bytes per exchange, 1 operation per batch:

SQL> net_meter (1, 1000, 1000, 1);

round_trips     MBps
REAL            REAL
    
_______________________________________

1245.393315542000254  2.489418401123078

1 Rows. -- 39345 msec.	

resulted in a measured throughput of 2.5 MBps

The second one was 1 thread, 5000 batches, 10000 bytes per exchange, 1 operation per batch:

SQL> net_meter (1, 5000, 10000, 1);

round_trips     MBps
REAL            REAL
    
___________________________________________

15915.291672080031181  305.017186586494738

1 Rows. -- 15394 msec.	

resulted in a measured throughput of 305 MBps.

This indicates that the user's network is slow when sending multiple short messages.

When using the ingestion you should check the:

status('cluster');	

command a few times and check the XX KB/s amount which should be around or above the 2500 mark.



1.5.35. How can I use the LOAD command to import RDF data?

SPARQL INSERT can be done using the LOAD command:

  
SPARQL INSERT INTO <..> { .... } [[FROM ...] { ... }]

SPARQL LOAD <x> [INTO <y>]

-- <ResourceURL> will be the Graph IRI of the loaded data:
SPARQL LOAD <ResourceURL>	

1.5.35.1. Examples

1.5.35.1.1. Load from Resource URL

In order to load data from resource URL for ex: http://www.w3.org/People/Berners-Lee/card#i , execute from isql the following command:

  
SQL> SPARQL LOAD <http://www.w3.org/People/Berners-Lee/card#i>;
callret-0
VARCHAR
_______________________________________________________________________________

Load <http://www.w3.org/People/Berners-Lee/card#i> into graph <http://www.w3.org/People/Berners-Lee/card#i> -- done

1 Rows. -- 703 msec.
SQL>	

1.5.35.1.2. Load from file
  1. Create DAV collection which is visible to public, for ex: http://localhost:8890/DAV/tmp
  2. Upload to the DAV collection a file for ex. with name listall.rq and with the following content:
      
    SPARQL
    PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
    PREFIX sioc: <http://rdfs.org/sioc/ns#>
    SELECT ?x ?p ?o
    FROM <http://mygraph.com>
    WHERE
      {
        ?x rdf:type sioc:User .
        ?x ?p ?o.
        ?x sioc:id ?id .
        FILTER REGEX(str(?id), "^King")
      }
    ORDER BY ?x	
    
  3. Execute from ISQL the following command:
      
    SQL>SPARQL LOAD bif:concat ("http://", bif:registry_get("URIQADefaultHost"), "/DAV/tmp/listall.rq") into graph <http://myNewGraph.com>;
    callret-0
    VARCHAR
    _______________________________________________________________________________
    
    Load <http://localhost:8890/DAV/tmp/listall.rq> into graph <http://myNewGraph.com> -- done
    
    1 Rows. -- 321 msec.	
    

1.5.35.1.3. Directly LOAD triples using ISQL
SQL>SPARQL INSERT INTO graph <http://mygraph.com>
{
  <http://myopenlink.net/dataspace/Kingsley#this>
  <http://www.w3.org/1999/02/22-rdf-syntax-ns#type>
  <http://rdfs.org/sioc/ns#User> .

  <http://myopenlink.net/dataspace/Kingsley#this>
  <http://rdfs.org/sioc/ns#id>
  <Kingsley> .

  <http://myopenlink.net/dataspace/Caroline#this>
  <http://www.w3.org/1999/02/22-rdf-syntax-ns#type>
  <http://rdfs.org/sioc/ns#User> .

  <http://myopenlink.net/dataspace/Caroline#this>
  <http://rdfs.org/sioc/ns#id>
  <Caroline> .
};	



1.5.36. How can I delete graphs using stored procedure?

The following script demonstrates the use of custom stored procedures for deleting graph(s). It first creates a table GRAPHS_TO_DELETE, into which the names of the graphs to be deleted should be inserted, as follows:

 
use MYUSR;

create procedure GRAPHS_TO_DELETE_SP (in gd_iris any)
{
  declare gd_iri iri_id;
  declare dp, row any;
  result_names (gd_iri);
  dp := dpipe (0, '__I2IDN');
  foreach (varchar iri in GD_IRIS) do
    {
      dpipe_input (dp, iri);
    }
  while (0 <> (row := dpipe_next (dp, 0)))
    {
      result (row[0]);
    }
}
;

drop view GRAPHS_TO_DELETE_VIEW;
create procedure view GRAPHS_TO_DELETE_VIEW as MYUSR.DBA.GRAPHS_TO_DELETE_SP (gd_iris) (gd_iri any);

create procedure DELETE_GRAPHS (in g_iris any)
{
  declare g_iids any;
  if (not isvector (g_iris) and g_iris is not null)
    signal ('22023', '.....', 'The input argument must be an array of strings or null');
  if (not length (g_iris))
    return 0;
  delete from DB.DBA.RDF_QUAD where G in (select * from GRAPHS_TO_DELETE_VIEW where gd_iris = g_iris) option (loop exists);
  return row_count ();
}
;

Finally call the procedure DELETE_GRAPHS to perform the deletion of the specified graphs. Note it does not return a result set and can be called as follows:

 
SQL> select MYUSR.DBA.DELETE_GRAPHS (vector ('g1', 'g2', 'g3'));	

This will return the number of triples removed from the specified graphs.

Note: the procedure only applies to the cluster so to get IRI IDs via partitioned pipe (DAQ). It is not usable on single.


1.5.37. How can I use SPARUL to add missing triples to a Named Graph?

1.5.37.1. What?

Use of SPARUL to add missing triples to a Named Graph. For example, an ontology/vocabulary extension.


1.5.37.2. Why?

A lot of ontologies and vocabularies started life prior to emergence of the Linked Data meme. As a result, many do not include rdfs:isDefinedBy relations (via triples) that associate Classes and Properties in an ontology with the ontology itself, using de-referencable URIs. The downside of this pattern is that Linked Data's follow-your-nose pattern isn't exploitable when viewing these ontologies e.g., when using contemporary Linked Data aware browsers.


1.5.37.3. How?

If SPARUL privileges are assigned to SPARQL or other accounts associated with SPARQL Endpoint. Or via WebID? protected SPARQL endpoint with SPARUL is granted to SPARQL or specific accounts or WebIDs in a group.

INSERT INTO <LocalNamedGraphIRI>
  { ?s rdfs:DefinedBy <LocalOntologyEntityURI>.
    ?o rdfs:isDefinedBy <http://www.w3.org/ns/auth/acl>. }
FROM <ExtSourceNamedGraphIRI> 
WHERE 
  { 
    ?s a ?o 
  }
1.5.37.3.1. Example
  1. Load Quad Named Graph via Sponger based query:
    DEFINE get:soft "replace"
    SELECT DISTINCT *
    FROM <http://www.w3.org/ns/auth/acl#>
    WHERE 
      { 
        ?s ?p ?o 
      }
    
  2. Added Triples via SPARUL to Ontology Named Graph:
    INSERT INTO <http://www.w3.org/ns/auth/acl#>
      {  ?s rdfs:DefinedBy <http://www.w3.org/ns/auth/acl>.
         ?o rdfs:DefinedBy <http://www.w3.org/ns/auth/acl>. }
    FROM <http://www.w3.org/ns/auth/acl#> 
    WHERE 
      {
        ?s a ?o
      }
    
  3. Via Conductor or Command-line iSQL courtesy of SPASQL execute the following statements:
    1. Remove an existing graph:
      SPARQL
      CLEAR GRAPH <http://www.w3.org/ns/auth/acl/> ;
      
      SPARQL
      CLEAR GRAPH <http://www.w3.org/ns/auth/acl> ;
      
      SPARQL
      CLEAR GRAPH <http://www.w3.org/ns/auth/acl#> ;
      
    2. Load a new graph:
      SPARQL
      LOAD <http://www.w3.org/ns/auth/acl> ;
      
    3. Add missing rdfs:isDefinedBy triples via SPARUL:
      SPARQL
      INSERT INTO <http://www.w3.org/ns/auth/acl>
        {  ?s rdfs:DefinedBy <http://www.w3.org/ns/auth/acl>.
           ?o rdfs:DefinedBy <http://www.w3.org/ns/auth/acl>. }
      FROM <http://www.w3.org/ns/auth/acl> 
      WHERE 
        {
          ?s a ?o
        } ;
      
  4. Verification: Access the following url: http://<cname>/describe/?uri=http://www.w3.org/ns/auth/acl>
See Also:

SPARUL -- an Update Language For RDF Graphs

Virtuoso Sponger




1.5.38. How can I use the SPARQL IF operator for SPARQL-BI endpoint?

Assume a SPARQL query is to be executed against the Virtuoso DBpedia SPARQL endpoint (http://dbpedia.org/sparql) to retrieve the decimal longitude of the "NJ Devils' hometown" with cardinal direction, which determines whether the decimal is positive (in the case of East) or negative (in the case of West).

Virtuoso supports SPARQL-BI, an extended SPARQL 1.0 implementation from before SPARQL 1.1 was ratified, thus the "IF" operator is not currently supported, but an equivalent bif:either built-in SQL function does exist enabling an equivalent query to be executed:

PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX dbo: <http://dbpedia.org/ontology/>
PREFIX dbp: <http://dbpedia.org/property/>
SELECT ?team ( (bif:either( ?ew = 'W', -1, 1)) * (?d + (((?m * 60) + ?s) / 3600.0)) as ?v)
  { 
    ?team a dbo:HockeyTeam . ?team rdfs:label 'New Jersey Devils'@en . 
    ?team dbp:city ?cityname . ?city rdfs:label ?cityname . 
    ?city dbp:longd ?d; dbp:longm ?m; dbp:longs ?s; dbp:longew ?ew . 
  }  	
SPARQL IF operator for SPARQL-BI endpoint
Figure: 1.5.38.1. SPARQL IF operator for SPARQL-BI endpoint
See Also:

Supported SPARQL-BI "define" pragmas

SPARQL-BI


1.5.39. How can I handle checkpoint condition?

In general, to control checkpoint in order to bypass client timeouts during long checkpoints when inserting triples into the Virtuoso server, one must disable automatic checkpoint by:

SQL> checkpoint_interval (0);	

and also to make sure the AutoCheckpointLogSize is off. Then can be performed checkpoint whenever the client wants using the 'checkpoint' command.

However the need of cache check is not needed unless instance shows regular errors. By default the cache check is disabled.

Virtuoso offers a new option in the Virtuoso ini file to enable the check of page maps: PageMapCheck, 1/0 default 0:

-- Virtuoso.ini

...
[Parameters]
...
PageMapCheck  = 0
...	

Also customer can add CheckpointSyncMode = 0 in order to disable sync during checkpoint to speed the operations.


1.5.40. How can I incorporate Content Negotiation into RDF bulk loaders?

The examples from below demonstrate how to incorporate Content Negotiation into RDF bulk loaders:


1.5.41. Virtuoso Linked Data Deployment In 3 Simple Steps?

Injecting Linked Data into the Web has been a major pain point for those who seek personal, service, or organization-specific variants of DBpedia. Basically, the sequence goes something like this:

  1. You encounter DBpedia or the LOD Cloud Pictorial.
  2. You look around (typically following your nose from link to link).
  3. You attempt to publish your own stuff.
  4. You get stuck.

The problems typically take the following form:

  1. Functionality confusion about the complementary Name and Address functionality of a single URI abstraction.
  2. Terminology confusion due to conflation and over-loading of terms such as Resource, URL, Representation, Document, etc.
  3. Inability to find robust tools with which to generate Linked Data from existing data sources such as relational databases, CSV files, XML, Web Services, etc.

To start addressing these problems, here is a simple guide for generating and publishing Linked Data using Virtuoso.

1.5.41.1. Step 1 - RDF Data Generation

Existing RDF data can be added to the Virtuoso RDF Quad Store via a variety of built-in data loader utilities.

Many options allow you to easily and quickly generate RDF data from other data sources:


1.5.41.2. Step 2 - Linked Data Deployment

Install the Faceted Browser VAD package (fct_dav.vad) which delivers the following:

  1. Faceted Browser Engine UI
  2. Dynamic Hypermedia Resource Generator:
    • delivers descriptor resources for every entity (data object) in the Native or Virtual Quad Stores
    • supports a broad array of output formats, including HTML+RDFa, RDF/XML, N3/Turtle, NTriples, RDF-JSON, OData+Atom, and OData+JSON.

1.5.41.3. Step 3 - Linked Data Consumption & Exploitation

Three simple steps allow you, your enterprise, and your customers to consume and exploit your newly deployed Linked Data --

  1. Load a page like this in your browser: http://<cname>[:<port>]/describe/?uri=<entity-uri>
    • <cname>[:<port>] gets replaced by the host and port of your Virtuoso instance
    • <entity-uri> gets replaced by the URI you want to see described -- for instance, the URI of one of the resources you let the Sponger handle.
  2. Follow the links presented in the descriptor page.
  3. If you ever see a blank page with a hyperlink subject name in the About: section at the top of the page, simply add the parameter "&sp=1" to the URL in the browser's Address box, and hit [ENTER]. This will result in an "on the fly" resource retrieval, transformation, and descriptor page generation.
  4. Use the navigator controls to page up and down the data associated with the "in scope" resource descriptor.


1.5.42. What are the differences between create, drop, clear and delete Graph?

In Virtuoso it does not matter whether CREATE GRAPH and DROP GRAPH are called or not. Their purpose is to provide compatibility with the original SPARUL that was designed for Jena. Some Jena triple stores require explicit creation of each graph (like CREATE TABLE in SQL), they report errors if one tries to create a graph twice and so on.

For Virtuoso, a new graph is not an important system event, it has single quad store shared by all graphs. When a graph is made by CREATE GRAPH, its name is placed in an auxiliary table that is used solely to signal appropriate errors on CREATE graph that is CREATE-d already or on DROP of a missing graph; this table is not used in any way in SPARQL or other subsystems. In a perfect world, smart development tools will query that table to give hints to a programmer regarding available graphs, but the reality is not so perfect.

What's more important is a difference between DELETE FROM g { ?s ?p ?o } FROM g WHERE { ?s ?p ?o } and CLEAR GRAPH g, as both will delete all triples from the specified graph <g> with equal speed, but CLEAR GRAPH will also delete free-text index data about occurrences of literals in this specific graph. So CLEAR GRAPH will make the database slightly more compact and the text search slightly faster. (Naturally, DROP GRAPH makes CLEAR GRAPH inside, not just DELETE FROM ... )

See Also:

RDF Insert Methods in Virtuoso


1.5.43. How can I perform search for predicate values?

What?

Creation of a stored procedure that enables users to find properties based on their string based token patterns.

Why?

When working with datasets from disparate datasources in a common named graph, there are times when you seek to scope sparql or spasql queries to specific property IRI/URI patterns without embarking upon inefficient regex heuristics.

What?

Make a stored procedure for querying against the main quad store table (rdf_quad). Surface the procedure as a magic predicate using "bif:" prefix. To find all the properties whose predicates start with "http://www.openlinksw.com/", a Virtuoso/PL procedure can be used as listed below:

SQL> create procedure PREDICATES_OF_IRI_PATH ( 
  in path varchar, 
  in dump_iri_ids integer := 0)
{
  declare PRED_IRI varchar;
  declare PRED_IRI_ID IRI_ID;
  declare path_head_len integer;
  
  if (dump_iri_ids)
    result_names (PRED_IRI_ID);
  else
    result_names (PRED_IRI);
    
  for ( SELECT RP_NAME, RP_ID 
        FROM RDF_PREFIX
        WHERE (RP_NAME >= path) AND (RP_NAME < path || chr(255)) ) do
    {
      declare fourbytes varchar;
      fourbytes := '----';
      fourbytes[0] := bit_shift (RP_ID, -24);
      fourbytes[1] := bit_and (bit_shift (RP_ID, -16), 255);
      fourbytes[2] := bit_and (bit_shift (RP_ID, -8), 255);
      fourbytes[3] := bit_and (RP_ID, 255);
      
      for ( SELECT RI_NAME, RI_ID from RDF_IRI
            WHERE (RI_NAME >= fourbytes) AND (RI_NAME < fourbytes || chr(255)) ) do
        {
          if (exists (SELECT TOP 1 1 FROM RDF_QUAD WHERE P=RI_ID))
            result (case when (dump_iri_ids) then RI_ID else RP_NAME || subseq (RI_NAME, 4) end);
        }
    }
    
  for ( path_head_len := length (path)-1; path_head_len >= 0; path_head_len := path_head_len - 1)
    {
      for ( SELECT RP_NAME, RP_ID from RDF_PREFIX
            WHERE RP_NAME = subseq (path, 0, path_head_len) ) do
        {
          declare fourbytes varchar;
          fourbytes := '----';
          fourbytes[0] := bit_shift (RP_ID, -24);
          fourbytes[1] := bit_and (bit_shift (RP_ID, -16), 255);
          fourbytes[2] := bit_and (bit_shift (RP_ID, -8), 255);
          fourbytes[3] := bit_and (RP_ID, 255);
          
          for ( SELECT RI_NAME, RI_ID from RDF_IRI
                WHERE (RI_NAME >= fourbytes || subseq (path, path_head_len))
                AND (RI_NAME < fourbytes || subseq (path, path_head_len) || chr(255)) ) do
            {
              if (exists (SELECT TOP 1 1 FROM RDF_QUAD WHERE P=RI_ID))
                result (case when (dump_iri_ids) then RI_ID else RP_NAME || subseq (RI_NAME, 4) end);
            }
        }
    }
}
;


Done. -- 16 msec.	

Example Usage

  1. Execute:
    set echo on;	
    
  2. Collect all predicates that start with:
    -- http://www.openlinksw.com/
    SQL>PREDICATES_OF_IRI_PATH ('http://www.openlinksw.com/', 1);
    VARCHAR
    _______________________________________________________________________________
    
    #i351
    #i159
    #i10
    #i8
    ...
    
    -- http://www.openlinksw.com/schemas/virtrdf
    SQL>PREDICATES_OF_IRI_PATH ('http://www.openlinksw.com/schemas/virtrdf', 1);
    PRED_IRI_ID
    VARCHAR
    _______________________________________________________________________________
    
    #i159
    #i10
    #i8
    #i6
    #i160
    #i269
    #i278
    #i275
    
    
    -- http://www.openlinksw.com/schemas/virtrdf#
    SQL>PREDICATES_OF_IRI_PATH ('http://www.openlinksw.com/schemas/virtrdf#',1);
    PRED_IRI_ID
    VARCHAR
    _______________________________________________________________________________
    
    #i159
    #i10
    #i8
    #i6
    #i160
    #i269
    #i278
    ...
    
    -- http://www.openlinksw.com/schemas/virtrdf#i
    SQL>PREDICATES_OF_IRI_PATH ('http://www.openlinksw.com/schemas/virtrdf#i',1);
    PRED_IRI_ID
    VARCHAR
    _______________________________________________________________________________
    
    #i159
    #i10
    #i8
    
    -- other
    SQL>PREDICATES_OF_IRI_PATH ('no://such :)', 1);
    0 Rows. -- 0 msec.
    	
    
  3. Show all predicates that start with:
    -- http://www.openlinksw.com/
    SQL>PREDICATES_OF_IRI_PATH ('http://www.openlinksw.com/');
    PRED_IRI
    VARCHAR
    _______________________________________________________________________________
    
    http://www.openlinksw.com/schemas/DAV#ownerUser
    http://www.openlinksw.com/schemas/virtrdf#inheritFrom
    http://www.openlinksw.com/schemas/virtrdf#isSpecialPredicate
    http://www.openlinksw.com/schemas/virtrdf#item
    http://www.openlinksw.com/schemas/virtrdf#loadAs
    http://www.openlinksw.com/schemas/virtrdf#noInherit
    http://www.openlinksw.com/schemas/virtrdf#qmGraphMap
    http://www.openlinksw.com/schemas/virtrdf#qmMatchingFlags
    http://www.openlinksw.com/schemas/virtrdf#qmObjectMap
    http://www.openlinksw.com/schemas/virtrdf#qmPredicateMap
    http://www.openlinksw.com/schemas/virtrdf#qmSubjectMap
    ...
    
    -- http://www.openlinksw.com/schemas/virtrdf
    SQL>PREDICATES_OF_IRI_PATH ('http://www.openlinksw.com/schemas/virtrdf');
    PRED_IRI
    VARCHAR
    _______________________________________________________________________________
    
    http://www.openlinksw.com/schemas/virtrdf#inheritFrom
    http://www.openlinksw.com/schemas/virtrdf#isSpecialPredicate
    http://www.openlinksw.com/schemas/virtrdf#item
    http://www.openlinksw.com/schemas/virtrdf#loadAs
    http://www.openlinksw.com/schemas/virtrdf#noInherit
    http://www.openlinksw.com/schemas/virtrdf#qmGraphMap
    http://www.openlinksw.com/schemas/virtrdf#qmMatchingFlags
    http://www.openlinksw.com/schemas/virtrdf#qmObjectMap
    http://www.openlinksw.com/schemas/virtrdf#qmPredicateMap
    http://www.openlinksw.com/schemas/virtrdf#qmSubjectMap
    http://www.openlinksw.com/schemas/virtrdf#qmTableName
    ...
    
    -- http://www.openlinksw.com/schemas/virtrdf#
    SQL>PREDICATES_OF_IRI_PATH ('http://www.openlinksw.com/schemas/virtrdf#');
    PRED_IRI
    VARCHAR
    _______________________________________________________________________________
    
    http://www.openlinksw.com/schemas/virtrdf#inheritFrom
    http://www.openlinksw.com/schemas/virtrdf#isSpecialPredicate
    http://www.openlinksw.com/schemas/virtrdf#item
    http://www.openlinksw.com/schemas/virtrdf#loadAs
    http://www.openlinksw.com/schemas/virtrdf#noInherit
    http://www.openlinksw.com/schemas/virtrdf#qmGraphMap
    http://www.openlinksw.com/schemas/virtrdf#qmMatchingFlags
    http://www.openlinksw.com/schemas/virtrdf#qmObjectMap
    http://www.openlinksw.com/schemas/virtrdf#qmPredicateMap
    http://www.openlinksw.com/schemas/virtrdf#qmSubjectMap
    http://www.openlinksw.com/schemas/virtrdf#qmTableName
    http://www.openlinksw.com/schemas/virtrdf#qmf01blankOfShortTmpl
    http://www.openlinksw.com/schemas/virtrdf#qmf01uriOfShortTmpl
    http://www.openlinksw.com/schemas/virtrdf#qmfBoolOfShortTmpl
    http://www.openlinksw.com/schemas/virtrdf#qmfBoolTmpl
    ...
    
    -- http://www.openlinksw.com/schemas/virtrdf#i
    SQL>PREDICATES_OF_IRI_PATH ('http://www.openlinksw.com/schemas/virtrdf#i');
    PRED_IRI
    VARCHAR
    _______________________________________________________________________________
    
    http://www.openlinksw.com/schemas/virtrdf#inheritFrom
    http://www.openlinksw.com/schemas/virtrdf#isSpecialPredicate
    http://www.openlinksw.com/schemas/virtrdf#item
    
    3 Rows. -- 15 msec.
    
    -- other
    SQL>PREDICATES_OF_IRI_PATH ('no://such :)');
    PRED_IRI
    VARCHAR
    _______________________________________________________________________________
    
    
    0 Rows. -- 15 msec.
    

If you want to use the procedure's output inside SPARQL queries, it can be wrapped by a procedure view and it in turn can be used in an RDF View but it is redundant for most applications.

For typical "almost static" data, it is more practical to write a procedure that will store all found predicates in some dedicated "dictionary" graph and then use the graph as usual.


1.5.44. How can I use INSERT via CONSTRUCT Statements?

You can write an ordinary CONSTRUCT statement, ensure that it generates the triples intended to be added, and then replace the leading CONSTRUCT keyword with the INSERT INTO phrase.

Example:

  1. Assume new triples need to be added for the equivalentClass:
    CONSTRUCT 
      {  
        ?s <http://www.w3.org/2002/07/owl#equivalentClass> `iri (bif:replace(?o,'http://schema.rdfs.org/', 'http://schema.org/'))`
       } 
    FROM <http://www.openlinksw.com/schemas/rdfs> 
    WHERE 
      { 
        ?s <http://www.w3.org/2002/07/owl#equivalentClass> ?o
      };	
    
  2. Execute the CONSTRUCT query from the htp://cname/sparql SPARQL endpoint.
  3. View the generated triples to ensure they are correct.
  4. Replace CONSTRUCT with INSERT INTO:
    SPARQL INSERT INTO <http://www.openlinksw.com/schemas/rdfs>  
      {  
        ?s <http://www.w3.org/2002/07/owl#equivalentClass> `iri (bif:replace(?o,'http://schema.rdfs.org/', 'http://schema.org/'))`
       } 
    FROM <http://www.openlinksw.com/schemas/rdfs> 
    WHERE 
      { 
        ?s <http://www.w3.org/2002/07/owl#equivalentClass> ?o
      } ;	
    

1.5.45. How to clear graphs which are related to empty graphs?

The following example demonstrates how to remove graphs which are related to empty graphs:

     
PREFIX nrl:<http://www.semanticdesktop.org/ontologies/2007/08/15/nrl#>
SELECT ( bif:exec(bif:sprintf("SPARQL CLEAR GRAPH<%s>", str(?mg))))
WHERE 
  {
    ?mg nrl:coreGraphMetadataFor ?g .
    FILTER(?g in ( <urn:nepomuk:local:8a9e692a> )) .
    FILTER ( !bif:exists((SELECT (1) WHERE { GRAPH ?g { ?s ?p ?o . } . })) ) .
  }	 

1.5.46. How can I use sub-queries to enable literal values based joins?

1.5.46.1. What?

Use of subqueries to enable literal values based joins.


1.5.46.2. Why?

Sophisticated access to literal values via subqueries provides powerful mechanism for enhancing sparql graph patterns via dynamic literal value generation.


1.5.46.3. How?

Use select list variables from subqueries to produce literal object values in sparql graph patterns.

1.5.46.3.1. Example

Assume in the following query, where should be used a sub-query to replace ?app:

SELECT DISTINCT ?r 
WHERE 
  {
    graph ?g 
      {
        ?r nie:url ?url .
      }  .
      ?g nao:maintainedBy ?app .
      ?app nao:identifier "nepomukindexer" .
  }	

If it is not sure that ?app is the only, for e.g., the triple ?app nao:identifier "nepomukindexer" can appear in more than one graph, then the query should be changed to:

SELECT DISTINCT ?r 
WHERE 
  {
    graph ?g 
      {
        ?r nie:url ?url .
      }  .
      ?g nao:maintainedBy ?app .
      FILTER (?app = (SELECT ?a WHERE { ?a nao:identifier "nepomukindexer" }))
}	

or even shorter:

SELECT DISTINCT ?r 
WHERE 
  {
   graph ?g 
     {
       ?r nie:url ?url .
     }  .
   ?g nao:maintainedBy `(SELECT ?a WHERE { ?a nao:identifier "nepomukindexer" })` .
 }	



1.5.47. How can I execute query with labels preference order?

The way to prefer one label to the other can be done as follows:


1.5.48. How can I get object datatype?

To get object datatype should be used the internal Virtuoso/PL function DB.DBA.RDF_DATATYPE_OF_OBJ, visible in SPARQL as sql:RDF_DATATYPE_OF_OBJ.

Suppose the following scenario:

# Explicit typecast (insert) 
SQL> sparql insert into <test_datatype> { <a> <string> "string 1"^^xsd:string . };
callret-0
VARCHAR
_______________________________________________________________________________

Insert into <test_datatype>, 1 (or less) triples -- done

1 Rows. -- 94 msec.


#Not explicit typecast (insert)
SQL> sparql insert into <test_datatype> { <a> <string> "string 2". };
callret-0
VARCHAR
_______________________________________________________________________________

Insert into <test_datatype>, 1 (or less) triples -- done

1 Rows. -- 16 msec.

SQL> SPARQL 
SELECT ?o (iri(sql:RDF_DATATYPE_OF_OBJ(?o, 'untyped!'))) 
FROM <test_datatype> { <a> <string> ?o} ;
o                       callret-1
VARCHAR                 VARCHAR
_______________________________________________________________________________

string 1                http://www.w3.org/2001/XMLSchema#string 
string 2                untyped!

2 Rows. -- 16 msec.
SQL>	

1.5.49. How Can I Backup and Restore individual table(s) and individual index(s)?

See more details here.


1.5.50. What bif:contains free-text options can I use?

Virtuoso supports the following free-text options for bif:contains:

  1. OFFBAND: See description for this free-text option in this section.
    • Note: it is useful only if data comers via an RDF View and the source relational table uses this trick;
  2. SCORE: This free-text option is calculated as described in this section:
    SQL>SPARQL 
    SELECT ?s1 as ?c1, ?sc, ?rank 
    WHERE 
      {
        {
          { 
            SELECT ?s1, (?sc * 3e-1) as ?sc, ?o1, 
              (sql:rnk_scale (<LONG::IRI_RANK> (?s1))) as ?rank  
            WHERE 
              { 
                ?s1 ?s1textp ?o1 . 
                ?o1 bif:contains  '"CNET"'  option (score ?sc)  . 
              } 
            ORDER BY DESC (?sc * 3e-1 + sql:rnk_scale (<LONG::IRI_RANK> (?s1)))  
            LIMIT 20  
            OFFSET 0 
          }
        }
      };
    
    c1                                              sc      rank
    _________________________________________________________________________
    http://www.mixx.com/stories/45003360/justi...   3 	    5.881291583872891e-14
    http://www.mixx.com/stories/45099313/bing_...   2.7     5.881291583872891e-14
    http://dbpedia.org/resource/CBS_Interactive 	  1.5 	  5.881291583872891e-14
    http://dbpedia.org/resource/CBS_Interactive 	  1.5 	  5.881291583872891e-14
    
    
    4 Rows. -- 1 msec.
    
  3. SCORE_LIMIT: This free-text option works as it is in plain SQL free-text. See more details :
    SQL> SPARQL 
    SELECT ?s ?sc 
    WHERE 
      { 
        ?s ?p ?o . 
        ?o bif:contains "tree" OPTION (score ?sc , score_limit 20)
      };
      
    s                                                                      sc
    VARCHAR                                                                INTEGER
    ________________________________________________________________________________   
    
                                                                          
    http://www.openlinksw.com/virtrdf-data-formats#default                 24
    http://www.openlinksw.com/virtrdf-data-formats#default                 24
    http://www.openlinksw.com/virtrdf-data-formats#sql-longvarbinary       21
    http://www.openlinksw.com/virtrdf-data-formats#sql-varchar-dt          20
    http://www.openlinksw.com/virtrdf-data-formats#sql-nvarchar-dt         20
    http://www.openlinksw.com/virtrdf-data-formats#sql-varchar-lang        20
    http://www.openlinksw.com/virtrdf-data-formats#sql-nvarchar-lang       20
    
    7 Rows. -- 2 msec.
    

1.5.51. What SPARQL Endpoint Protection Methods can I use?

Virtuoso supports the following SPARQL Endpoint protection methods:

  1. Secure SPARQL Endpoint Guide using WebID Protocol
  2. Secure SPARQL Endpoint Guide via OAuth
  3. Secure SPARQL Endpoint Guide via SQL Accounts -- usage path digest authentication

1.5.52. How do I assign SPARQL role to SQL user?

This section a sample scenario how to assign SPARQL ( for ex. SPARQL_SELECT ) role to Virtuoso SQL user:

  1. Go to http://cname/conductor.
  2. Enter dba credentials.
  3. Go to System Admin -> User Accounts:
    Assign SPARQL Role to SQL User
    Figure: 1.5.52.1. Assign SPARQL Role to SQL User
  4. Click "Edit" for a given user from the very last right column:
    Assign SPARQL Role to SQL User
    Figure: 1.5.52.1. Assign SPARQL Role to SQL User
  5. From "Accounts Roles" drop-down list select a SPARQL Role, for ex. SPARQL_SELECT and click the ">>" button:
    Assign SPARQL Role to SQL User
    Figure: 1.5.52.1. Assign SPARQL Role to SQL User
  6. Click "Save".