Skip Headers
Oracle® Database Semantic Technologies Developer's Guide
11g Release 2 (11.2)

Part Number E25609-03
Go to Documentation Home
Home
Go to Book List
Book List
Go to Table of Contents
Contents
Go to Index
Index
Go to Master Index
Master Index
Go to Feedback page
Contact Us

Go to previous page
Previous
Go to next page
Next
PDF · Mobi · ePub

7 Jena Adapter for Oracle Database

The Jena Adapter for Oracle Database (referred to here as the Jena Adapter) provides a Java-based interface to Oracle Database Semantic Technologies by implementing the well-known Jena Graph, Model, and DatasetGraph APIs. (Jena is an open source framework, and its license and copyright conditions are described in http://jena.sourceforge.net/license.html.)

The DatasetGraph APIs are for managing named graph data, also referred to as quads. In addition, the Jena Adapter provides network analytical functions on top of semantic data through integrating with the Oracle Spatial network data model.

This chapter assumes that you are familiar with major concepts explained in Chapter 1, "Oracle Database Semantic Technologies Overview" and Chapter 2, "OWL Concepts". It also assumes that you are familiar with the overall capabilities and use of the Jena Java framework. For information about the Jena framework, see http://jena.sourceforge.net/, especially the Jena Documentation page. If you use the network analytical function, you should also be familiar with the Oracle Spatial network data model, which is documented in Oracle Spatial Topology and Network Data Models Developer's Guide.

The Jena Adapter extends the semantic data management capabilities of Oracle Database Release 11.2 RDF/OWL.

This chapter includes the following major topics:

Disclaimer:

The current Jena Adapter release has been tested against Jena 2.6.4, ARQ 2.8.8, and Joseki 3.4.4. Because of the nature of open source projects, you should not use this Jena Adapter with later versions of Jena, ARQ, or Joseki.

7.1 Setting Up the Software Environment

To use the Jena Adapter, you must first ensure that the system environment has the necessary software, including Oracle Database 11g Release 2 with the Spatial and Partitioning options and with Semantic Technologies support enabled, Jena version 2.6.4, the Jena Adapter, and JDK 1.6. You can set up the software environment by performing these actions:

  1. Install Oracle Database Release 11.2 Enterprise Edition with the Oracle Spatial and Partitioning Options.

  2. If you have not yet installed Release 11.2.0.3 or later, install the 11.2.0.2 Patch Set for Oracle Database Server (https://updates.oracle.com/ARULink/PatchDetails/process_form?patch_num=10098816).

  3. Enable the support for Semantic Technologies, as explained in Section A.1.

  4. Install Jena (version 2.6.4): download the .zip file from http://sourceforge.net/projects/jena/files/Jena/Jena-2.6.4/jena-2.6.4.zip/download and unzip it. (The directory or folder into which you unzip it will be referred to as <Jena_DIR>.)

    The Java package will be unpacked into <Jena_DIR>.

    Note that Jena 2.6.4 comes with ARQ version 2.8.7 (arq-2.8.7.jar); however, this version of the Jena Adapter actually requires a newer ARQ version (arq-2.8.8.jar). You can download arq-2.8.8.jar from http://sourceforge.net/projects/jena/files/ARQ/ARQ-2.8.8/arq-2.8.8.zip/download and unzip it to a temporary directory. Remove the arq-2.8.7.jar file from <Jena_DIR>/Jena-2.6.4/lib/, and copy arq-2.8.8.jar from the temporary directory into <Jena_DIR>/Jena-2.6.4/lib/.

  5. Download the Jena Adapter (jena_adaptor_for_release11.2.0.3.zip) from the Oracle Database Semantic Technologies page (http://www.oracle.com/technetwork/database/options/semantic-tech/and click Downloads), and unzip it into a temporary directory, such as (on a Linux system) /tmp/jena_adapter. (If this temporary directory does not already exist, create it before the unzip operation.)

    The Jena Adapter directories and files have the following structure:

    examples/
       examples/Test10.java
       examples/Test11.java
       examples/Test12.java
       examples/Test13.java
       examples/Test14.java
       examples/Test15.java
       examples/Test16.java
       examples/Test17.java
       examples/Test18.java
       examples/Test19.java
       examples/Test20.java
       examples/Test6.java
       examples/Test7.java
       examples/Test8.java
       examples/Test9.java
       examples/Test.java
    jar/
       jar/sdordfclient.jar
    javadoc/
      javadoc/javadoc.zip
    joseki/
       joseki/index.html
       joseki/application.xml
       joseki/update.html
       joseki/xml-to-html.xsl
       joseki/joseki-config.ttl
    sparqlgateway/ 
      default.xslt
      noop.xslt
      qb1.sparql
      . . .
      browse.jsp
      index.html
      . . .
      application.xml
      WEB-INF/
        WEB-INF/web.xml
        WEB-INF/weblogic.xml
      WEB-INF/lib/
      StyleSheets/
        StyleSheets/paginator.css
        StyleSheets/sg.css
        StyleSheets/sgmin.css
      Scripts/
        Scripts/load.js
        Scripts/paginator.js
        Scripts/tooltip.js
      admin/
        admin/sparql.jsp
        admin/xslt.jsp
    web/
      web/web.xml
    
  6. Copy ojdbc6.jar into <Jena_DIR>/lib (Linux) or <Jena_DIR>\lib (Windows). (ojdbc6.jar is in $ORACLE_HOME/jdbc/lib or %ORACLE_HOME%\jdbc\lib.)

  7. Copy sdordf.jar into <Jena_DIR>/lib (Linux) or <Jena_DIR>\lib (Windows). (sdordf.jar is in $ORACLE_HOME/md/jlib or %ORACLE_HOME%\md\jlib.)

  8. If JDK 1.6 is not already installed, install it.

  9. If the JAVA_HOME environment variable does not already refer to the JDK 1.6 installation, define it accordingly. For example:

    setenv JAVA_HOME /usr/local/packages/jdk16/
    
  10. If the SPARQL service to support the SPARQL protocol is not set up, set it up as explained in Section 7.2.

After setting up the software environment, ensure that your Semantic Technologies environment can enable you to use the Jena Adapter to perform queries, as explained in Section 7.3.

7.2 Setting Up the SPARQL Service

Setting up a SPARQL endpoint using the Jena Adapter involves downloading Joseki, an open source HTTP engine that supports the SPARQL protocol and SPARQL queries. This section explains how to set up a SPARQL service using a servlet deployed in WebLogic Server. The number and complexity of the steps reflect the fact that Oracle is not permitted to bundle all the dependent third-party libraries in a .war or .ear file.

  1. Download and Install Oracle WebLogic Server 11g Release 1 (10.3.1). For details, see http://www.oracle.com/technology/products/weblogic/ and http://www.oracle.com/technetwork/middleware/ias/downloads/wls-main-097127.html.

  2. Ensure that you have Java 6 installed, because it is required by Joseki 3.4.4.

  3. Download Joseki 3.4.4 (joseki-3.4.4.zip) from http://sourceforge.net/projects/joseki/files/Joseki-SPARQL/.

  4. Unpack joseki-3.4.4.zip into a temporary directory. For example:

    mkdir /tmp/joseki
    cp joseki-3.4.4.zip /tmp/joseki
    cd /tmp/joseki
    unzip joseki-3.4.4.zip
    
  5. Ensure that you have downloaded and unzipped the Jena Adapter for Oracle Database, as explained in Section 7.1.

  6. Create a directory named joseki.war at the same level as the jena_adapter directory, and go to it. For example:

    mkdir /tmp/joseki.war
    cd /tmp/joseki.war
    
  7. Copy necessary files into the directory created in the preceding step:

    cp /tmp/jena_adapter/joseki/*  /tmp/joseki.war
    cp -rf /tmp/joseki/Joseki-3.4.4/webapps/joseki/StyleSheets  /tmp/joseki.war
    
  8. Create directories and copy necessary files into them, as follows:

    mkdir /tmp/joseki.war/WEB-INF
    cp /tmp/jena_adapter/web/*  /tmp/joseki.war/WEB-INF
     
    mkdir /tmp/joseki.war/WEB-INF/lib
    cp /tmp/joseki/Joseki-3.4.4/lib/*.jar   /tmp/joseki.war/WEB-INF/lib
    cp /tmp/jena_adapter/jar/*.jar   /tmp/joseki.war/WEB-INF/lib
    ##
    ## Assume ORACLE_HOME points to the home directory of a Release 11.2.0.3 Oracle Database.
    ##
    cp $ORACLE_HOME/md/jlib/sdordf.jar /tmp/joseki.war/WEB-INF/lib
    cp $ORACLE_HOME/jdbc/lib/ojdbc6.jar /tmp/joseki.war/WEB-INF/lib
    
  9. Using the WebLogic Server Administration console, create a J2EE data source named OracleSemDS. During the data source creation, you can specify a user and password for the database schema that contains the relevant semantic data against which SPARQL queries are to be executed.

    If you need help in creating this data source, see Section 7.2.1, "Creating the Required Data Source Using WebLogic Server".

  10. Go to the autodeploy directory of WebLogic Server and copy files, as follows. (For information about auto-deploying applications in development domains, see: http://download.oracle.com/docs/cd/E11035_01/wls100/deployment/autodeploy.html)

    cd <domain_name>/autodeploy
    cp -rf  /tmp/joseki.war  <domain_name>/autodeploy
    

    In the preceding example, <domain_name> is the name of a WebLogic Server domain.

    Note that while you can run a WebLogic Server domain in two different modes, development and production, only development mode allows you use the auto-deployment feature.

  11. Check the files and the directory structure to see if they reflect the following:

    .
    |-- META-INF
    |   `-- MANIFEST.MF
    |-- StyleSheets
    |   `-- joseki.css
    |-- WEB-INF
    |   |-- lib
    |   |   |-- arq-2.8.8-tests.jar
    |   |   |-- arq-2.8.8.jar
    |   |   |-- icu4j-3.4.4.jar
    |   |   |-- iri-0.8.jar
    |   |   |-- jena-2.6.4-tests.jar
    |   |   |-- jena-2.6.4.jar
    |   |   |-- jetty-6.1.25.jar
    |   |   |-- jetty-util-6.1.25.jar
    |   |   |-- joseki-3.4.4.jar
    |   |   |-- junit-4.5.jar
    |   |   |-- log4j-1.2.14.jar
    |   |   |-- lucene-core-2.3.1.jar
    |   |   |-- ojdbc6.jar
    |   |   |-- sdb-1.3.4.jar
    |   |   |-- sdordf.jar
    |   |   |-- sdordfclient.jar
    |   |   |-- servlet-api-2.5-20081211.jar
    |   |   |-- slf4j-api-1.5.8.jar
    |   |   |-- slf4j-log4j12-1.5.8.jar
    |   |   |-- stax-api-1.0.1.jar
    |   |   |-- tdb-0.8.10.jar
    |   |   |-- wstx-asl-3.2.9.jar
    |   |   `-- xercesImpl-2.7.1.jar
    |   `-- web.xml
    |-- application.xml
    |-- index.html
    |-- joseki-config.ttl
    |-- update.html
    `-- xml-to-html.xsl
    
  12. If you want to build a .war file from the /tmp/joseki.war directory (note that a .war file is required if you want to deploy Joseki to an OC4J container), enter the following commands:

    cd /tmp/joseki.war
    jar cvf /tmp/joseki_app.war *
    
  13. Start or restart WebLogic Server.

  14. Verify your deployment by using your Web browser to connect to a URL in the following format (assume that the Web application is deployed at port 7001): http://<hostname>:7001/joseki

    You should see a page titled Oracle SPARQL Service Endpoint using Joseki, and the first text box should contain an example SPARQL query.

  15. Click Submit Query.

    You should see a page titled Oracle SPARQL Endpoint Query Results. There may or may not be any results, depending on the underlying semantic model against which the query is executed.

By default, the joseki-config.ttl file contains an oracle:Dataset definition using a model named M_NAMED_GRAPHS. The following snippet shows the configuration. The oracle:allGraphs predicate denotes that the SPARQL service endpoint will serve queries using all graphs stored in the M_NAMED_GRAPHS model.

<#oracle> rdf:type oracle:Dataset;
    joseki:poolSize     1 ;         ## Number of concurrent connections allowed to this dataset.
    oracle:connection
    [ a oracle:OracleConnection ;
    ];
    oracle:allGraphs [ oracle:firstModel "M_NAMED_GRAPHS" ] .

The M_NAMED_GRAPHS model will be created automatically (if it does not already exist) upon the first SPARQL query request. You can add a few example triples and quads to test the named graph functions; for example:

SQL> CONNECT username/password
SQL> INSERT INTO m_named_graphs_tpl VALUES(sdo_rdf_triple_s('m_named_graphs','<urn:s>','<urn:p>','<urn:o>'));
SQL> INSERT INTO m_named_graphs_tpl VALUES(sdo_rdf_triple_s('m_named_graphs:<urn:G1>','<urn:g1_s>','<urn:g1_p>','<urn:g1_o>'));
SQL> INSERT INTO m_named_graphs_tpl VALUES(sdo_rdf_triple_s('m_named_graphs:<urn:G2>','<urn:g2_s>','<urn:g2_p>','<urn:g2_o>'));
SQL> COMMIT;

After inserting the rows, go to http://<hostname>:7001/joseki, type the following SPARQL query, and click Submit Query:

SELECT ?g ?s ?p ?o
WHERE 
  { GRAPH ?g { ?s ?p ?o} }

The result should be an HTML table with four columns and two sets of result bindings.

The http://<hostname>:7001/joseki page also contains a JSON Output option. If this option is selected (enabled), the SPARQL query response is converted to JSON format.

7.2.1 Creating the Required Data Source Using WebLogic Server

If you need help creating the required J2EE data source using the WebLogic Server admin console, you can follow these steps:

  1. Login to: http://<hostname>:7001/console

  2. In the Domain Structure panel, click Services.

  3. Click JDBC

  4. Click Data Sources.

  5. In the Summary of JDBC Data Sources panel, click New under the Data Sources table.

  6. In the Create a New JDBC Data Source panel, enter or select the following values.

    Name: OracleSemDS

    JNDI Name: OracleSemDS

    Database Type: Oracle

    Database Driver: Oracle's Driver (Thin) Versions: 9.0.1,9.2.0,10,11

  7. Click Next twice.

  8. In the Connection Properties panel, enter the appropriate values for the Database Name, Host Name, Port, Database User Name (schema that contains semantic data), Password fields.

  9. Click Next.

  10. Select (check) the target server or servers to which you want to deploy this OracleSemDS data source.

  11. Click Finish.

    You should see a message that all changes have been activated and no restart is necessary.

7.2.2 Configuring the SPARQL Service

By default, the SPARQL Service endpoint assumes that the queries are to be executed against a semantic model with a pre-set name. This semantic model is owned by the schema specified in the J2EE data source with JNDI name OracleSemDS. Note that you do not need to create this model explicitly using PL/SQL or Java; if the model does not exist in the network, it will be automatically created, along with the necessary application table and index.

Note:

Effective with the Jena Adapter release in November 2011, the application table index (<model_name>_idx) definition is changed to accommodate named graph data (quads).

For existing models created by an older version of the Jena Adapter, you can migrate the application table index name and definition by using the static OracleUtils.migrateApplicationTableIndex(oracle, graph, dop) method in the oracle.spatial.rdf.client.jena package. (See the Javadoc for more information.) Note that the new index definition is critical to the performance of DML operations against the application table.

However, you must configure the SPARQL service by editing the joseki-config.ttl configuration file, which is in <domain_name>/autodeploy/joseki.war.

The supplied joseki-config.ttl file includes a section similar to the following for the Oracle data set:

#
## Datasets
#
[] ja:loadClass "oracle.spatial.rdf.client.jena.assembler.OracleAssemblerVocab" .
 
oracle:Dataset  rdfs:subClassOf  ja:RDFDataset .
 
<#oracle> rdf:type oracle:Dataset;
    joseki:poolSize     1 ;         ## Number of concurrent connections allowed to this dataset.
    oracle:connection
    [ a oracle:OracleConnection ;
    ];
    oracle:defaultModel [ oracle:firstModel "TEST_MODEL" ] .

In this section of the file, you can:

  • Modify the joseki:poolSize value, which specifies the number of concurrent connections allowed to this Oracle data set (<#oracle> rdf:type oracle:Dataset;), which points to various RDF models in the database.

  • Modify the name (or the object value of oracle:firstModel predicate) of the defaultModel, to use a different semantic model for queries. You can also specify multiple models, and one or more rulebases for this defaultModel.

    For example, the following specifies two models (named ABOX and TBOX) and an OWLPRIME rulebase for the default model. Note that models specified using the oracle:modelName predicate must exist; they will not be created automatically.

    <#oracle> rdf:type oracle:Dataset;
        joseki:poolSize 1 ; ## Number of concurrent connections allowed to this dataset.
        oracle:connection
        [ a oracle:OracleConnection ;
        ];
        oracle:defaultModel [ oracle:firstModel "ABOX";
                                          oracle:modelName "TBOX";
                                          oracle:rulebaseName "OWLPRIME" ] .
    
  • Specify named graphs in the dataset. For example, you can create a named graph called <http://G1> based on two Oracle models and an entailment, as follows.

    <#oracle> rdf:type oracle:Dataset;
        joseki:poolSize 1 ; ## Number of concurrent connections allowed to this dataset.
          oracle:connection
          [ a oracle:OracleConnection ;
          ];
          oracle:namedModel [ oracle:firstModel "ABOX";
                              oracle:modelName "TBOX";
                              oracle:rulebaseName "OWLPRIME";
                              oracle:namedModelURI <http://G1> ]  .
    

    The object of namedModel can take the same specifications as defaultModel, so virtual models are supported here as well (see also the next item).

  • Use a virtual model for queries by adding oracle:useVM "TRUE", as shown in the following example. Note that if the specified virtual model does not exist, it will automatically be created on demand.

    <#oracle> rdf:type oracle:Dataset;
        joseki:poolSize 1 ; ## Number of concurrent connections allowed to this dataset.
        oracle:connection
        [ a oracle:OracleConnection ;
        ];
        oracle:defaultModel [ oracle:firstModel "ABOX";
                                         oracle:modelName "TBOX";
                                         oracle:rulebaseName "OWLPRIME";
                                         oracle:useVM "TRUE"
        ] .
    

    For more information, see Section 7.10.1, "Virtual Models Support".

  • Specify a virtual model as the default model to answer SPARQL queries by using the predicate oracle:virtualModelName, as shown in the following example with a virtual model named TRIPLE_DATA_VM_0:

    oracle:defaultModel [ oracle:virtualModelName "TRIPLE_DATA_VM_0" ] .
    

    If the underlying data consists of quads, you can use oracle:virtualModelName with oracle:allGraphs. The presence of oracle:allGraphs causes an instantiation of DatasetGraphOracleSem objects to answer named graph queries. An example is as follows:

    oracle:allGraphs [ oracle:virtualModelName "QUAD_DATA_VM_0" ] .
    

    Note that when a virtual model name is specified as the default graph, the endpoint can serve only query requests; SPARQL Update operations are not supported.

  • Set the queryOptions and inferenceMaintenance properties to change the query behavior and inference update mode. (See the Javadoc for information about QueryOptions and InferenceMaintenanceMode.)

    By default, QueryOptions.ALLOW_QUERY_INVALID_AND_DUP and InferenceMaintenanceMode.NO_UPDATE are set, for maximum query flexibility and efficiency.

7.2.2.1 Client Identifiers

For every database connection created or used by the Jena Adapter, a client identifier is associated with the connection. The client identifier can be helpful, especially in a Real Application Cluster (Oracle RAC) environment, for isolating Jena Adapter-related activities from other database activities when you are doing performance analysis and tuning.

By default, the client identifier assigned is JenaAdapter. However, you can specify a different value by setting the Java VM clientIdentifier property using the following format:

-Doracle.spatial.rdf.client.jena.clientIdentifier=<identificationString>

To start the tracing of only Jena Adapter-related activities on the database side, you can use the DBMS_MONITOR.CLIENT_ID_TRACE_ENABLE procedure. For example:

SQL> EXECUTE DBMS_MONITOR.CLIENT_ID_TRACE_ENABLE('JenaAdapter', true, true);

7.2.2.2 Using OLTP Compression for Application Tables and Staging Tables

By default, the Jena Adapter creates the application tables and any staging tables (the latter used for bulk loading, as explained in Section 7.11) using basic table compression with the following syntax:

CREATE TABLE .... (... column definitions ...) ... compress;

However, if you are licensed to use the Oracle Advanced Compression option no the database, you can set the following JVM property to turn on OLTP compression, which compresses data during all DML operations against the underlying application tables and staging tables:

-Doracle.spatial.rdf.client.jena.advancedCompression="compress for oltp"

7.2.3 Terminating Long-Running SPARQL Queries

Because some applications need to be able to terminate long-running SPARQL queries, an abort framework has been introduced with the Jena Adapter and the Joseki setup. Basically, for queries that may take a long time to run, you must stamp each with a unique query ID (qid) value.

For example, the following SPARQL query selects out the subject of all triples. A query ID (qid) is set so that this query can be terminated upon request.

PREFIX ORACLE_SEM_FS_NS:  <http://example.com/semtech#qid=8761>
SELECT ?subject WHERE {?subject ?property ?object }

The qid attribute value is of long integer type. You can choose a value for the qid for a particular query based on your own application needs.

To terminate a SPARQL query that has been submitted with a qid value, applications can send an abort request to a servlet in the following format and specify a matching QID value

http://<hostname>:7001/joseki/querymgt?abortqid=8761

7.2.4 N-Triples Encoding for Non-ASCII Characters

For any non-ASCII characters in the lexical representation of RDF resources, \uHHHH N-Triples encoding is used when the characters are inserted into the Oracle database. (For details about N-Triples encoding, see http://www.w3.org/TR/rdf-testcases/#ntrip_grammar.) Encoding of the constant resources in a SPARQL query is handled in a similar fashion.

Using \uHHHH N-Triples encoding enables support for international characters, such as a mix of Norwegian and Swedish characters, in the Oracle database even if a supported Unicode character set is not being used.

7.3 Setting Up the Semantic Technologies Environment

To use the Jena Adapter to perform queries, you can connect as any user (with suitable privileges) and use any models in the semantic network. If your Semantic Technologies environment already meets the requirements, you can go directly to compiling and running Java code that uses the Jena Adapter. If your Semantic Technologies environment is not yet set up to be able to use the Jena Adapter, you can perform actions similar to the following example steps:

  1. Connect as SYS with the SYSDBA role:

    sqlplus sys/<password-for-sys> as sysdba
    
  2. Create a tablespace for the system tables. For example:

    CREATE TABLESPACE rdf_users datafile 'rdf_users01.dbf' 
        size 128M reuse autoextend on next 64M 
        maxsize unlimited segment space management auto;
    
  3. Create the semantic network. For example:

    EXECUTE sem_apis.create_sem_network('RDF_USERS');
    
  4. Create a database user (for connecting to the database to use the semantic network and the Jena Adapter). For example:

    CREATE USER rdfusr IDENTIFIED BY <password-for-udfusr>
                       DEFAULT TABLESPACE rdf_users;
    
  5. Grant the necessary privileges to this database user. For example:

    GRANT connect, resource TO rdfusr;
    
  6. To use the Jena Adapter with your own semantic data, perform the appropriate steps to store data, create a model, and create database indexes, as explained in Section 1.10, "Quick Start for Using Semantic Data". Then perform queries by compiling and running Java code; see Section 7.15 for information about example queries.

    To use the Jena Adapter with supplied example data, see Section 7.15.

7.4 SEM_MATCH and Jena Adapter Queries Compared

There are two ways to query semantic data stored in Oracle Database: SEM_MATCH-based SQL statements and SPARQL queries through the Jena Adapter. Queries using each approach are similar in appearance, but there are important behavioral differences. To ensure consistent application behavior, you must understand the differences and use care when dealing with query results coming from SEM_MATCH queries and SPARQL queries.

The following simple examples show the two approaches.

Query 1 (SEM_MATCH-based)

select s, p, o
    from table(sem_match('{?s ?p ?o}', sem_models('Test_Model'), ....))

Query 2 (SPARQL query through the Jena Adapter)

select ?s ?p ?o
where {?s ?p ?o}

These two queries perform the same kind of functions; however, there are some important differences. Query 1 (SEM_MATCH-based):

Query 2 (SPARQL query executed through the Jena Adapter) also reads all triples out of Test_Model (assume it executed a call to ModelOracleSem referring to the same underlying Test_Model). However, Query 2:

Blank node handling is another difference between the two approaches:

The maximum length for the name of a semantic model created using the Jena Adapter API is 22 characters.

7.5 Optimized Handling of SPARQL Queries

This section describes some performance-related features of the Jena Adapter that can enhance SPARQL query processing. These features are performed automatically by default.

This section assumes that you are familiar with SPARQL, including the CONSTRUCT feature and property paths.

7.5.1 Compilation of SPARQL queries to a single SEM_MATCH Call

SPARQL queries involving DISTINCT, OPTIONAL, FILTER, UNION, ORDER BY, and LIMIT are converted to a single Oracle SEM_MATCH table function. If a query cannot be converted directly to SEM_MATCH because it uses SPARQL features not supported by SEM_MATCH (for example, CONSTRUCT), the Jena Adapter employs a hybrid approach and tries to execute the largest portion of the query using a single SEM_MATCH function while executing the rest using the Jena ARQ query engine.

For example, the following SPARQL query is directly translated to a single SEM_MATCH table function:

PREFIX dc:  <http://purl.org/dc/elements/1.1/> 
PREFIX rdf:  <http://www.w3.org/1999/02/22-rdf-syntax-ns#> 
PREFIX foaf: <http://xmlns.com/foaf/0.1/> 
SELECT ?person ?name 
  WHERE {
                 {?alice foaf:knows ?person . }
               UNION { 
                 ?person ?p ?name. OPTIONAL { ?person ?x ?name1 } 
                     }
        }

However, the following example query is not directly translatable to a single SEM_MATCH table function because of the CONSTRUCT keyword:

PREFIX vcard: <http://www.w3.org/2001/vcard-rdf/3.0#> 
CONSTRUCT  { <http://example.org/person#Alice> vcard:FN ?obj } 
  WHERE  { { ?x <http://pred/a> ?obj.}
         UNION
         { ?x <http://pred/b> ?obj.}  }

In this case, the Jena Adapter converts the inner UNION query into a single SEM_MATCH table function, and then passes on the result set to the Jena ARQ query engine for further evaluation.

7.5.2 Optimized Handling of Property Paths

As defined in Jena, a property path is a possible route through an RDF graph between two graph nodes. Property paths are an extension of SPARQL and are more expressive than basic graph pattern queries, because regular expressions can be used over properties for pattern matching RDF graphs. For more information about property paths, see the documentation for the Jena ARQ query engine.

The Jena Adapter supports all Jena property path types through the integration with the Jena ARQ query engine, but it converts some common path types directly to native SQL hierarchical queries (not based on SEM_MATCH) to improve performance. The following types of property paths are directly converted to SQL by the Jena Adapter when dealing with triple data:

  • Predicate alternatives: (p1 | p2 | … | pn) where pi is a property URI

  • Predicate sequences: (p1 / p2 / … / pn) where pi is a property URI

  • Reverse paths : ( ^ p ) where p is a predicate URI

  • Complex paths: p+, p*, p{0, n} where p could be an alternative, sequence, reverse path, or property URI

Path expressions that cannot be captured in this grammar are not translated directly to SQL by the Jena Adapter, and they are answered using the Jena query engine.

The following example contains a code snippet using a property path expression with path sequences:

String m = "PROP_PATH";
 
ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, m);
 
GraphOracleSem graph = new GraphOracleSem(oracle, m);    
    
// populate the RDF Graph
    graph.add(Triple.create(Node.createURI("http://a"),
    Node.createURI("http://p1"),
    Node.createURI("http://b")));
 
graph.add(Triple.create(Node.createURI("http://b"),
 Node.createURI("http://p2"),
 Node.createURI("http://c")));
 
graph.add(Triple.create(Node.createURI("http://c"),
 Node.createURI("http://p5"),
 Node.createURI("http://d")));
 
String query =
" SELECT ?s  " +
" WHERE {?s (<http://p1>/<http://p2>/<http://p5>)+ <http://d>.}";
   
QueryExecution qexec = 
      QueryExecutionFactory.create(QueryFactory.create(query, 
 Syntax.syntaxARQ), model);
 
try {
  ResultSet results = qexec.execSelect();
  ResultSetFormatter.out(System.out, results);
}
finally {
  if (qexec != null)
    qexec.close();
}
     
OracleUtils.dropSemanticModel(oracle, m);
model.close();

7.6 Additions to the SPARQL Syntax to Support Other Features

The Jena Adapter allows you to pass in hints and additional query options. It implements these capabilities by overloading the SPARQL namespace prefix syntax by using Oracle-specific namespaces that contain query options. The namespaces are in the form PREFIX ORACLE_SEM_xx_NS, where xx indicates the type of feature (such as HT for hint or AP for additional predicate)

7.6.1 SQL Hints

SQL hints can be passed to a SEM_MATCH query including a line in the following form:

PREFIX ORACLE_SEM_HT_NS: <http://oracle.com/semtech#hint>

Where hint can be any hint supported by SEM_MATCH. For example:

PREFIX ORACLE_SEM_HT_NS: <http://oracle.com/semtech#leading(t0,t1)> 
SELECT ?book ?title ?isbn     
WHERE { ?book <http://title> ?title. ?book <http://ISBN> ?isbn }

In this example, t0,t1 refers to the first and second patterns in the query.

Note the slight difference in specifying hints when compared to SEM_MATCH. Due to restrictions of namespace value syntax, a comma (,) must be used to separate t0 and t1 (or other hint components) instead of a space.

For more information about using SQL hints, see Section 1.6, "Using the SEM_MATCH Table Function to Query Semantic Data", specifically the material about the HINT0 keyword in the options attribute.

7.6.2 Using Bind Variables in SPARQL Queries

In Oracle Database, using bind variables can reduce query parsing time and increase query efficiency and concurrency. Bind variable support in SPARQL queries is provided through namespace pragma specifications similar to ORACLE_SEM_FS_NS.

Consider a case where an application runs two SPARQL queries, where the second (Query 2) depends on the partial or complete results of the first (Query 1). Some approaches that do not involve bind variables include:

  • Iterating through results of Query 1 and generating a set of queries. (However, this approach requires as many queries as the number of results of Query 1.)

  • Constructing a SPARQL filter expression based on results of Query 1.

  • Treating Query 1 as a subquery.

Another approach in this case is to use bind variables, as in the following sample scenario:

Query 1:
 
  SELECT ?x
    WHERE { ... <some complex query> ... };
 
 
Query 2:
 
  SELECT ?subject ?x
    WHERE {?subject <urn:related> ?x .};

The following example shows Query 2 with the syntax for using bind variables with the Jena Adapter:

PREFIX ORACLE_SEM_FS_NS: <http://oracle.com/semtech#no_fall_back,s2s>
PREFIX ORACLE_SEM_UEAP_NS: <http://oracle.com/semtech#x$RDFVID%20in(?,?,?)>
PREFIX ORACLE_SEM_UEPJ_NS: <http://oracle.com/semtech#x$RDFVID>
PREFIX ORACLE_SEM_UEBV_NS: <http://oracle.com/semtech#1,2,3>
SELECT ?subject ?x
WHERE {
  ?subject <urn:related>  ?x
};

This syntax includes using the following namespaces:

  • ORACLE_SEM_UEAP_NS is like ORACLE_SEM_AP_NS, but the value portion of ORACLE_SEM_UEAP_NS is URL Encoded. Before the value portion is used, it must be URL decoded, and then it will be treated as an additional predicate to the SPARQL query.

    In this example, after URL decoding, the value portion (following the # character) of this ORACLE_SEM_UEAP_NS prefix becomes "x$RDFVID in(?,?,?)". The three question marks imply a binding to three values coming from Query 1.

  • ORACLE_SEM_UEPJ_NS specifies the additional projections involved. In this case, because ORACLE_SEM_UEAP_NS references the x$RDFVID column, which does not appear in the SELECT clause of the query, it must be specified. Multiple projections are separated by commas.

  • ORACLE_SEM_UEBV_NS specifies the list of bind values that are URL encoded first, and then concatenated and delimited by commas.

Conceptually, the preceding example query is equivalent to the following non-SPARQL syntax query, in which 1, 2, and 3 are treated as bind values:

SELECT ?subject ?x
  WHERE {
    ?subject <urn:related>  ?x
  }
  AND ?x$RDFVID in (1,2,3);

In the preceding SPARQL example of Query 2, the three integers 1, 2, and 3 come from Query 1. You can use the oext:build-uri-for-id function to generate such internal integer IDs for RDF resources. The following example gets the internal integer IDs from Query 1:

PREFIX oext: <http://oracle.com/semtech/jena-adaptor/ext/function#>
SELECT ?x  (oext:build-uri-for-id(?x) as ?xid)
WHERE { ... <some complex query> ... };

The values of ?xid have the form of <rdfvid:integer-value>. The application can strip out the angle brackets and the "rdfvid:" strings to get the integer values and pass them to Query 2.

Consider another case, with a single query structure but potentially many different constants. For example, the following SPARQL query finds the hobby for each user who has a hobby and who logs in to an application. Obviously, different users will provide different <uri> values to this SPARQL query, because users of the application are represented using different URIs.

SELECT ?hobby
  WHERE { <uri> <urn:hasHobby> ?hobby };

One approach, which would not use bind variables, is to generate a different SPARQL query for each different <uri> value. For example, user Jane Doe might trigger the execution of the following SPARQL query:

SELECT ?hobby WHERE {
<http://www.example.com/Jane_Doe> <urn:hasHobby> ?hobby };

However, another approach is to use bind variables, as in the following example specifying user Jane Doe:

PREFIX ORACLE_SEM_FS_NS: <http://oracle.com/semtech#no_fall_back,s2s>
PREFIX ORACLE_SEM_UEAP_NS: <http://oracle.com/semtech#subject$RDFVID%20in(ORACLE_ORARDF_RES2VID(?))>
PREFIX ORACLE_SEM_UEPJ_NS: <http://oracle.com/semtech#subject$RDFVID>
PREFIX ORACLE_SEM_UEBV_NS: <http://oracle.com/semtech#http%3a%2f%2fwww.example.com%2fJohn_Doe>
SELECT ?subject ?hobby
  WHERE {
    ?subject <urn:hasHobby>  ?hobby
  };

Conceptually, the preceding example query is equivalent to the following non-SPARQL syntax query, in which http://www.example.com/Jane_Doe is treated as a bind variable:

SELECT ?subject ?hobby
WHERE {
  ?subject <urn:hasHobby>  ?hobby
}
AND ?subject$RDFVID in (ORACLE_ORARDF_RES2VID('http://www.example.com/Jane_Doe'));

In this example, ORACLE_ORARDF_RES2VID is a function that translates URIs and literals into their internal integer ID representation. This function is created automatically when the Jena Adapter is used to connect to an Oracle database.

7.6.3 Additional WHERE Clause Predicates

The SEM_MATCH filter attribute can specify additional selection criteria as a string in the form of a WHERE clause without the WHERE keyword. Additional WHERE clause predicates can be passed to a SEM_MATCH query including a line in the following form:

PREFIX ORACLE_SEM_AP_NS: <http://oracle.com/semtech#pred>

Where pred reflects the WHERE clause content to be appended to the query. For example:

PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX ORACLE_SEM_AP_NS:<http://www.oracle.com/semtech#label$RDFLANG='fr'>  
SELECT DISTINCT ?inst ?label
  WHERE { ?inst a <http://someCLass>. ?inst rdfs:label ?label . }
  ORDER BY (?label) LIMIT 20

In this example, a restriction is added to the query that the language type of the label variable must be 'fr'.

7.6.4 Additional Query Options

Additional query options can be passed to a SEM_MATCH query including a line in the following form:

PREFIX ORACLE_SEM_FS_NS: <http://oracle.com/semtech#option>

Where option reflects a query option (or multiple query options delimited by commas) to be appended to the query. For example:

PREFIX ORACLE_SEM_FS_NS:   
<http://oracle.com/semtech#timeout=3,dop=4,INF_ONLY,ORDERED,ALLOW_DUP=T>
SELECT * WHERE {?subject ?property ?object }

The following query options are supported:

  • ALLOW_DUP=t chooses a faster way to query multiple semantic models, although duplicate results may occur.

  • BEST_EFFORT_QUERY=t, when used with the TIMEOUT=n option, returns all matches found in n seconds for the SPARQL query.

  • DEGREE=n specifies, at the statement level, the degree of parallelism (n) for the query. With multi-core or multi-CPU processors, experimenting with different DOP values (such as 4 or 8) may improve performance.

    Contrast DEGREE with DOP, which specifies parallelism at the session level. DEGREE is recommended over DOP for use with the Jena Adapter, because DEGREE involves less processing overhead.

  • DOP=n specifies, at the session level, the degree of parallelism (n) for the query. With multi-core or multi-CPU processors, experimenting with different DOP values (such as 4 or 8) may improve performance.

  • INF_ONLY causes only the inferred model to be queried.

  • JENA_EXECUTOR disables the compilation of SPARQL queries to SEM_MATCH (or native SQL); instead, the Jena native query executor will be used.

  • JOIN=n specifies how results from a SPARQL SERVICE call to a federated query can be joined with other parts of the query. For information about federated queries and the JOIN option, see Section 7.6.4.1.

  • NO_FALL_BACK causes the underlying query execution engine not to fall back on the Jena execution mechanism if a SQL exception occurs.

  • ODS=n specifies, at the statement level, the level of dynamic sampling. (For an explanation of dynamic sampling, see the section about estimating statistics with dynamic sampling in Oracle Database Performance Tuning Guide.) Valid values for n are 1 through 10. For example, you could try ODS=3 for complex queries.

  • ORDERED is translated to a LEADING SQL hint for the query triple pattern joins, while performing the necessary RDF_VALUE$ joins last.

  • PLAIN_SQL_OPT=F disables the native compilation of queries directly to SQL.

  • QID=n specifies a query ID number; this feature can be used to cancel the query if it is not responding.

  • RESULT_CACHE uses the Oracle RESULT_CACHE directive for the query.

  • REWRITE=F disables ODCI_Table_Rewrite for the SEM_MATCH table function.

  • SKIP_CLOB=T causes CLOB values not to be returned for the query.

  • S2S (SPARQL to pure SQL) causes the underlying SEM_MATCH-based query or queries generated based on the SPARQL query to be further converted into SQL queries without using the SEM_MATCH table function. The resulting SQL queries are executed by the Oracle cost-based optimizer, and the results are processed by the Jena Adapter before being passed on to the client. For more information about the S2S option, including benefits and usage information, see Section 7.6.4.2.

    S2S is enabled by default for all SPARQL queries. If you want to disable S2S, set the following JVM system property:

    -Doracle.spatial.rdf.client.jena.defaultS2S=false
    
  • TIMEOUT=n (query timeout) specifies the number of seconds (n) that the query will run until it is terminated. The underlying SQL generated from a SPARQL query can return many matches and can use features like subqueries and assignments, all of which can take considerable time. The TIMEOUT and BEST_EFFORT_QUERY=t options can be used to prevent what you consider excessive processing time for the query.

7.6.4.1 JOIN Option and Federated Queries

A SPARQL federated query, as described in W3C documents, is a query "over distributed data" that entails "querying one source and using the acquired information to constrain queries of the next source." For more information, see SPARQL 1.1 Federation Extensions (http://www.w3.org/2009/sparql/docs/fed/service).

You can use the JOIN option (described in Section 7.6.4) and the SERVICE keyword in a federated query that uses the Jena Adapter. For example, assume the following query:

SELECT ?s ?s1 ?o
 WHERE { ?s1 ?p1 ?s .
                    {
                     SERVICE <http://sparql.org/books> { ?s ?p ?o }
                    }
                 }

If the local query portion (?s1 ?p1 ?s,) is very selective, you can specify join=2, as shown in the following query:

PREFIX ORACLE_SEM_FS_NS:   <http://oracle.com/semtech#join=2>
SELECT ?s ?s1 ?o
 WHERE { ?s1 ?p1 ?s .
                    {
                     SERVICE <http://sparql.org/books> { ?s ?p ?o }
                    }
                 }

In this case, the local query portion (?s1 ?p1 ?s,) is executed locally against the Oracle database. Each binding of ?s from the results is then pushed into the SERVICE part (remote query portion), and a call is made to the service endpoint specified. Conceptually, this approach is somewhat like nested loop join.

If the remote query portion (?s ?s1 ?o) is very selective, you can specify join=3, as shown in the following query, so that the remote portion is executed first and results are used to drive the execution of local portion:

PREFIX ORACLE_SEM_FS_NS:   <http://oracle.com/semtech#join=3>
SELECT ?s ?s1 ?o
 WHERE { ?s1 ?p1 ?s .
                    {
                     SERVICE <http://sparql.org/books> { ?s ?p ?o }
                    }
                  }

In this case, a single call is made to the remote service endpoint and each binding of ?s triggers a local query. As with join=2, this approach is conceptually a nested loop based join, but the difference is that the order is switched.

If neither the local query portion nor the remote query portion is very selective, then we can choose join=1, as shown in the following query:

PREFIX ORACLE_SEM_FS_NS:   <http://oracle.com/semtech#join=1>
SELECT ?s ?s1 ?o
 WHERE { ?s1 ?p1 ?s .
                    {
                     SERVICE <http://sparql.org/books> { ?s ?p ?o }
                    }
                }

In this case, the remote query portion and the local portion are executed independently, and the results are joined together by Jena. Conceptually, this approach is somewhat like a hash join.

For debugging or tracing federated queries, you can use the HTTP Analyzer in Oracle JDeveloper to see the underlying SERVICE calls.

7.6.4.2 S2S Option Benefits and Usage Information

The S2S option, described in Section 7.6.4, provides the following potential benefits:

  • It works well with the RESULT_CACHE option to improve query performance. Using the S2S and RESULT_CACHE options is especially helpful for queries that are executed frequently.

  • It reduces the parsing time of the SEM_MATCH table function, which can be helpful for applications that involve many dynamically generated SPARQL queries.

  • It eliminates the limit of 4000 bytes for the query body (the first parameter of the SEM_MATCH table function), which means that longer, more complex queries are supported.

The S2S option causes an internal in-memory cache to be used for translated SQL query statements. The default size of this internal cache is 1024 (that is, 1024 SQL queries); however, you can adjust the size by using the following Java VM property:

-Doracle.spatial.rdf.client.jena.queryCacheSize=<size>

7.6.5 Midtier Resource Caching

When semantic data is stored, all of the resource values are hashed into IDs, which are stored in the triples table. The mappings from value IDs to full resource values are stored in the MDSYS.RDF_VALUE$ table. At query time, for each selected variable, Oracle Database must perform a join with the RDF_VALUE$ table to retrieve the resource.

However, to reduce the number of joins, you can use the midtier cache option, which causes an in-memory cache on the middle tier to be used for storing mappings between value IDs and resource values. To use this feature, include the following PREFIX pragma in the SPARQL query:

PREFIX ORACLE_SEM_FS_NS: <http://oracle.com/semtech#midtier_cache>

To control the maximum size (in bytes) of the in-memory cache, use the oracle.spatial.rdf.client.jena.cacheMaxSize system property. The default cache maximum size is 1GB.

Note that midtier resource caching is most effective for queries using ORDER BY or DISTINCT (or both) constructs, or queries with multiple projection variables. Midtier cache can be combined with the other options specified in Section 7.6.4.

If you want to pre-populate the cache with all of the resources in a model, use the GraphOracleSem.populateCache or DatasetGraphOracleSem.populateCache method. Both methods take a parameter specifying the number of threads used to build the internal midtier cache. Running either method in parallel can significantly increase the cache building performance on a machine with multiple CPUs (cores).

7.7 Functions Supported in SPARQL Queries through the Jena Adapter

SPARQL queries through the Jena Adapter can use the following kinds of functions:

7.7.1 Functions in the ARQ Function Library

SPARQL queries through the Jena Adapter can use functions in the function library of the Jena ARQ query engine. These queries are executed in the middle tier.

The following examples use the upper-case and namespace functions. In these examples, the prefix fn is <http://www.w3.org/2005/xpath-functions#> and the prefix afn is <http://jena.hpl.hp.com/ARQ/function#>.

PREFIX  fn: <http://www.w3.org/2005/xpath-functions#>
PREFIX  afn: <http://jena.hpl.hp.com/ARQ/function#>
SELECT (fn:upper-case(?object) as ?object1)
WHERE { ?subject dc:title ?object }

PREFIX  fn: <http://www.w3.org/2005/xpath-functions#>
PREFIX  afn: <http://jena.hpl.hp.com/ARQ/function#>
SELECT ?subject (afn:namespace(?object) as ?object1)
WHERE { ?subject <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> ?object } 

7.7.2 Native Oracle Database Functions for Projected Variables

SPARQL queries through the Jena Adapter can use native Oracle Database functions for projected variables. These queries and the functions are executed inside the database. Note that the functions described in this section should not be used together with ARQ functions (described in Section 7.7.1).

This section lists the supported native functions and provides some examples. In the examples, the prefix oext is <http://oracle.com/semtech/jena-adaptor/ext/function#>.

Note:

In the preceding URL, note the spelling jena-adaptor, which is retained for compatibility with existing applications and which must be used in queries. The adapter spelling is used in regular text, to follow Oracle documentation style guidelines.
  • oext:upper-literal converts literal values (except for long literals) to uppercase. For example:

    PREFIX  oext: <http://oracle.com/semtech/jena-adaptor/ext/function#>
    SELECT (oext:upper-literal(?object) as ?object1)
    WHERE { ?subject dc:title ?object }
    
  • oext:lower-literal converts literal values (except for long literals) to lowercase. For example:

    PREFIX  oext: <http://oracle.com/semtech/jena-adaptor/ext/function#>
    SELECT (oext:lower-literal(?object) as ?object1)
    WHERE { ?subject dc:title ?object }
    
  • oext:build-uri-for-id converts the value ID of a URI, bNode, or literal into a URI form. For example:

    PREFIX  oext: <http://oracle.com/semtech/jena-adaptor/ext/function#>
    SELECT (oext:build-uri-for-id(?object) as ?object1)
    WHERE { ?subject dc:title ?object }
    

    An example of the output might be: <rdfvid:1716368199350136353>

    One use of this function is to allow Java applications to maintain in memory a mapping of those value IDs to the lexical form of URIs, bNodes, or literals. The MDSYS.RDF_VALUE$ table provides such a mapping in Oracle Database.

    For a given variable ?var, if only oext:build-uri-for-id(?var) is projected, the query performance is likely to be faster because fewer internal table join operations are needed to answer the query.

  • oext:literal-strlen returns the length of literal values (except for long literals). For example:

    PREFIX  oext: <http://oracle.com/semtech/jena-adaptor/ext/function#>
    SELECT (oext:literal-strlen(?object) as ?objlen)
    WHERE { ?subject dc:title ?object }
    

7.7.3 User-Defined Functions

SPARQL queries through the Jena Adapter can use user-defined functions that are stored in the database.

In the following example, assume that you want to define a string length function (my_strlen) that handles long literals (CLOB) as well as short literals. On the SPARQL query side, this function can be referenced under the namespace of ouext, which is http://oracle.com/semtech/jena-adaptor/ext/user-def-function#.

PREFIX  ouext: <http://oracle.com/semtech/jena-adaptor/ext/user-def-function#>
SELECT ?subject ?object (ouext:my_strlen(?object) as ?obj1)
WHERE { ?subject dc:title ?object }

Inside the database, functions including my_strlen, my_strlen_cl, my_strlen_la, my_strlen_lt, and my_strlen_vt are defined to implement this capability. Conceptually, the return values of these functions are mapped as shown in Table 7-1.

Table 7-1 Functions and Return Values for my_strlen Example

Function Name Return Value

my_strlen

<VAR>

my_strlen_cl

<VAR>$RDFCLOB

my_strlen_la

<VAR>$RDFLANG

my_strlen_lt

<VAR>$RDFLTYP

my_strlen_vt

<VAR>$RDFVTYP


A set of functions (five in all) is used to implement a user-defined function that can be referenced from SPARQL, because this aligns with the internal representation of an RDF resource (in MDSYS.RDF_VALUE$). There are five major columns describing an RDF resource in terms of its value, language, literal type, long value, and value type, and these five columns can be selected out using SEM_MATCH. In this context, a user-defined function simply converts one RDF resource that is represented by five columns to another RDF resource.

These functions are defined as follows:

create or replace function my_strlen(rdfvtyp in varchar2,
                              rdfltyp in varchar2,
                              rdflang in varchar2,
                              rdfclob in clob,
                              value   in varchar2
                              ) return varchar2
 as
   ret_val  varchar2(4000);
 begin
   -- value
   if (rdfvtyp = 'LIT') then
     if (rdfclob is null) then
       return length(value);
     else
       return dbms_lob.getlength(rdfclob);
     end if;
   else
     -- Assign -1 for non-literal values so that application can
     -- easily differentiate
     return '-1';
   end if;
 end;
 /
 
 create or replace function my_strlen_cl(rdfvtyp in varchar2,
                              rdfltyp in varchar2,
                              rdflang in varchar2,
                              rdfclob in clob,
                              value   in varchar2
                              ) return clob
 as
 begin
   return null;
 end;
 /
 
 create or replace function my_strlen_la(rdfvtyp in varchar2,
                              rdfltyp in varchar2,
                              rdflang in varchar2,
                              rdfclob in clob,
                              value   in varchar2
                              ) return varchar2
 as
 begin
   return null;
 end;
 /
 
 create or replace function my_strlen_lt(rdfvtyp in varchar2,
                              rdfltyp in varchar2,
                              rdflang in varchar2,
                              rdfclob in clob,
                              value   in varchar2
                              ) return varchar2
 as
   ret_val  varchar2(4000);
 begin
   -- literal type
   return 'http://www.w3.org/2001/XMLSchema#integer';
 end;
 /
 
 create or replace function my_strlen_vt(rdfvtyp in varchar2,
                              rdfltyp in varchar2,
                              rdflang in varchar2,
                              rdfclob in clob,
                              value   in varchar2
                              ) return varchar2
 as
   ret_val  varchar2(3);
 begin
   return 'LIT';
 end;
 /

User-defined functions can also accept a parameter of VARCHAR2 type. The following five functions together define a my_shorten_str function that accepts an integer (in VARCHAR2 form) for the substring length and returns the substring. (The substring in this example is 12 characters, and it must not be greater than 4000 bytes.)

-- SPARQL query that returns the first 12 characters of literal values.
-- 
PREFIX  ouext: <http://oracle.com/semtech/jena-adaptor/ext/user-def-function#>
SELECT (ouext:my_shorten_str(?object, "12") as ?obj1) ?subject
WHERE { ?subject dc:title ?object }
 
create or replace function my_shorten_str(rdfvtyp in varchar2,
                            rdfltyp in varchar2,
                            rdflang in varchar2,
                            rdfclob in clob,
                            value   in varchar2,
                            arg     in varchar2
                            ) return varchar2
as
 ret_val  varchar2(4000);
begin
 -- value
 if (rdfvtyp = 'LIT') then
   if (rdfclob is null) then
     return substr(value, 1, to_number(arg));
   else
     return dbms_lob.substr(rdfclob, to_number(arg), 1);
   end if;
 else
   return null;
 end if;
end;
/
 
create or replace function my_shorten_str_cl(rdfvtyp in varchar2,
                            rdfltyp in varchar2,
                            rdflang in varchar2,
                            rdfclob in clob,
                            value   in varchar2,
                            arg     in varchar2
                            ) return clob
as
 ret_val  clob;
begin
 -- lob
 return null;
end;
/
 
create or replace function my_shorten_str_la(rdfvtyp in varchar2,
                            rdfltyp in varchar2,
                            rdflang in varchar2,
                            rdfclob in clob,
                            value   in varchar2,
                            arg     in varchar2
                            ) return varchar2
as
 ret_val  varchar2(4000);
begin
 -- lang
 if (rdfvtyp = 'LIT') then
   return rdflang;
 else
   return null;
 end if;
end;
/
 
create or replace function my_shorten_str_lt(rdfvtyp in varchar2,
                            rdfltyp in varchar2,
                            rdflang in varchar2,
                            rdfclob in clob,
                            value   in varchar2,
                            arg     in varchar2
                            ) return varchar2
as
 ret_val  varchar2(4000);
begin
 -- literal type
 ret_val := rdfltyp;
 return ret_val;
end;
/
 
create or replace function my_shorten_str_vt(rdfvtyp in varchar2,
                            rdfltyp in varchar2,
                            rdflang in varchar2,
                            rdfclob in clob,
                            value   in varchar2,
                            arg     in varchar2
                            ) return varchar2
as
 ret_val  varchar2(3);
begin
 return 'LIT';
end;
/

7.8 SPARQL Update Support

The Jena Adapter supports SPARQL Update (http://www.w3.org/TR/sparql11-update/), also referred to as SPARUL. The primary programming APIs involve the Jena class UpdateAction (in package com.hp.hpl.jena.update) and the Jena Adapter classes GraphOracleSem and DatasetGraphOracleSem. Example 7-1 shows a SPARQL Update operation removes all triples in named graph <http://example/graph> from the relevant model stored in the database.

Example 7-1 Simple SPARQL Update

GraphOracleSem graphOracleSem = .... ;
DatasetGraphOracleSem dsgos = DatasetGraphOracleSem.createFrom(graphOracleSem);
 
// SPARQL Update operation
String szUpdateAction = "DROP GRAPH <http://example/graph>";
 
// Execute the Update against a DatasetGraph instance (can be a Jena Model as well)
UpdateAction.parseExecute(szUpdateAction, dsgos);

Note that Oracle Database does not keep any information about an empty named graph. This implies if you invoke CREATE GRAPH <graph_name> without adding any triples into this graph, then no additional rows in the application table or the underlying RDF_LINK$ table will be created. To an Oracle database, you can safely skip the CREATE GRAPH step, as is the case in Example 7-1.

Example 7-2 shows a SPARQL Update operation (from ARQ 2.8.8) involving multiple insert and delete operations.

Example 7-2 SPARQL Update with Insert and Delete Operations

PREFIX : <http://example/>
CREATE GRAPH <http://example/graph> ;
INSERT DATA { :r :p 123 } ;
INSERT DATA { :r :p 1066 } ;
DELETE DATA { :r :p 1066 } ;
INSERT DATA {
  GRAPH <http://example/graph> { :r :p 123 . :r :p 1066 }
} ;
DELETE DATA {
  GRAPH <http://example/graph>  { :r :p 123 }
}

After running the update operation in Example 7-2 against an empty DatasetGraphOracleSem, running the SPARQL query SELECT ?s ?p ?o WHERE {?s ?p ?o} generates the following response:

-----------------------------------------------------------------------------------------------
| s                  | p                  | o                                                 |
===============================================================================================
| <http://example/r> | <http://example/p> | "123"^^<http://www.w3.org/2001/XMLSchema#decimal> |
-----------------------------------------------------------------------------------------------

Using the same data, running the SPARQL query SELECT ?g ?s ?p ?o where {GRAPH ?g {?s ?p ?o}} generates the following response:

-------------------------------------------------------------------------------------------------------------------------
| g                      | s                  | p                  | o                                                  |
=========================================================================================================================
| <http://example/graph> | <http://example/r> | <http://example/p> | "1066"^^<http://www.w3.org/2001/XMLSchema#decimal> |
-------------------------------------------------------------------------------------------------------------------------

In addition to using the Java API for SPARQL Update operations, you can configure Joseki to accept SPARQL Update operations by removing the comment (##) characters at the start of the following lines in the joseki-config.ttl file.

## <#serviceUpdate>
##     rdf:type            joseki:Service ;
##     rdfs:label          "SPARQL/Update" ;
##     joseki:serviceRef   "update/service" ;
##     # dataset part
##     joseki:dataset      <#oracle>;
##     # Service part.    
##     # This processor will not allow either the protocol,
##     # nor the query, to specify the dataset.
##     joseki:processor    joseki:ProcessorSPARQLUpdate
##     .
## 
## <#serviceRead>
##     rdf:type            joseki:Service ;
##     rdfs:label          "SPARQL" ;
##     joseki:serviceRef   "sparql/read" ;
##     # dataset part
##     joseki:dataset      <#oracle> ;     ## Same dataset
##     # Service part. 
##     # This processor will not allow either the protocol,
##     # nor the query, to specify the dataset.
##     joseki:processor    joseki:ProcessorSPARQL_FixedDS ;
##     .

After you edit the joseki-config.ttl file, you must restart the Joseki Web application. You can then try a simple update operation, as follows:

  1. In your browser, go to: http://<hostname>:7001/joseki/update.html

  2. Type or paste the following into the text box:

    PREFIX : <http://example/>
    INSERT DATA {
      GRAPH <http://example/g1> { :r :p 455 }
    }
    
  3. Click Perform SPARQL Update.

To verify that the update operation was successful, go to http://<hostname>:7001/joseki and enter the following query:

SELECT *
WHERE
  {  GRAPH <http://example/g1> {?s ?p ?o}};

The response should contain the following triple:

<http://example/r>     <http://example/p>    "455"^^<http://www.w3.org/2001/XMLSchema#decimal>

A SPARQL Update can also be sent using an HTTP POST operation to the http://<hostname>:7001/joseki/update/service. For example, you can use curl (http://en.wikipedia.org/wiki/CURL) to send an HTTP POST request to perform the update operation:

curl --data "request=PREFIX%20%3A%20%3Chttp%3A%2F%2Fexample%2F%3E%20%0AINSERT%20DATA%20%7B%0A%20%20GRAPH%20%3Chttp%3A%2F%2Fexample%2Fg1%3E%20%7B%20%3Ar%20%3Ap%20888%20%7D%0A%7D%0A"  http://hostname:7001/joseki/update/service

In the preceding example, the URL encoded string is a simple INSERT operation into a named graph. After decoding, it reads as follows:

PREFIX : <http://example/>
INSERT DATA {
  GRAPH <http://example/g1> { :r :p 888 }

7.9 Analytical Functions for RDF Data

You can perform analytical functions on RDF data by using the SemNetworkAnalyst class in the oracle.spatial.rdf.client.jena package. This support integrates the Oracle Spatial network data model (NDM) logic with the underlying RDF data structures. Therefore, to use analytical functions on RDF data, you must be familiar with the Oracle Spatial NDM, which is documented in Oracle Spatial Topology and Network Data Models Developer's Guide.

The required NDM Java libraries, including sdonm.jar and sdoutl.jar, are under the directory $ORACLE_HOME/md/jlib. Note that xmlparserv2.jar (under $ORACLE_HOME/xdk/lib) must be included in the classpath definition.

Example 7-3 uses the SemNetworkAnalyst class, which internally uses the NDM NetworkAnalyst API

Example 7-3 Performing Analytical functions on RDF Data

Oracle oracle = new Oracle(jdbcUrl, user, password);
GraphOracleSem graph = new GraphOracleSem(oracle, modelName);
 
Node nodeA = Node.createURI("http://A");
Node nodeB = Node.createURI("http://B");
Node nodeC = Node.createURI("http://C");
Node nodeD = Node.createURI("http://D");
Node nodeE = Node.createURI("http://E");
Node nodeF = Node.createURI("http://F");
Node nodeG = Node.createURI("http://G");
Node nodeX = Node.createURI("http://X");
 
// An anonymous node
Node ano = Node.createAnon(new AnonId("m1"));
 
Node relL = Node.createURI("http://likes");
Node relD = Node.createURI("http://dislikes");
Node relK = Node.createURI("http://knows");
Node relC = Node.createURI("http://differs");
 
graph.add(new Triple(nodeA, relL, nodeB));
graph.add(new Triple(nodeA, relC, nodeD));
graph.add(new Triple(nodeB, relL, nodeC));
graph.add(new Triple(nodeA, relD, nodeC));
 
graph.add(new Triple(nodeB, relD, ano));
graph.add(new Triple(nodeC, relL, nodeD));
graph.add(new Triple(nodeC, relK, nodeE));
graph.add(new Triple(ano,   relL, nodeD));
graph.add(new Triple(ano,   relL, nodeF));
graph.add(new Triple(ano,   relD, nodeB));
 
// X only likes itself
graph.add(new Triple(nodeX, relL, nodeX));
 
graph.commitTransaction();
HashMap<Node, Double> costMap = new HashMap<Node, Double>();
costMap.put(relL, Double.valueOf((double)0.5));
costMap.put(relD, Double.valueOf((double)1.5));
costMap.put(relC, Double.valueOf((double)5.5));
 
graph.setDOP(4); // this allows the underlying LINK/NODE tables
                 // and indexes to be created in parallel.
 
SemNetworkAnalyst sna = SemNetworkAnalyst.getInstance(
    graph,   // network data source
    true,    // directed graph
    true,    // cleanup existing NODE and LINK table
    costMap
    );
 
psOut.println("From nodeA to nodeC");
Node[] nodeArray = sna.shortestPathDijkstra(nodeA, nodeC);
printNodeArray(nodeArray, psOut);
 
psOut.println("From nodeA to nodeD"); 
nodeArray = sna.shortestPathDijkstra( nodeA, nodeD);
printNodeArray(nodeArray, psOut);
 
psOut.println("From nodeA to nodeF");
nodeArray = sna.shortestPathAStar(nodeA, nodeF);
printNodeArray(nodeArray, psOut);
 
psOut.println("From ano to nodeC");
nodeArray = sna.shortestPathAStar(ano, nodeC);
printNodeArray(nodeArray, psOut);
 
psOut.println("From ano to nodeX");
nodeArray = sna.shortestPathAStar(ano, nodeX);
printNodeArray(nodeArray, psOut);
 
graph.close();
oracle.dispose();
...
...
   
// A helper function to print out a path
public static void printNodeArray(Node[] nodeArray, PrintStream psOut)
{
  if (nodeArray == null) {
    psOut.println("Node Array is null");
    return;
  }
  if (nodeArray.length == 0) {psOut.println("Node Array is empty"); }
  int iFlag = 0;
  psOut.println("printNodeArray: full path starts");
  for (int iHops = 0; iHops < nodeArray.length; iHops++) {
    psOut.println("printNodeArray: full path item " + iHops + " = "
        + ((iFlag == 0) ? "[n] ":"[e] ") + nodeArray[iHops]);
    iFlag = 1 - iFlag;
  }
}

In Example 7-3:

The output of Example 7-3 is as follows. In this output, the shortest paths are listed for the given start and end nodes. Note that the return value of sna.shortestPathAStar(ano, nodeX) is null because there is no path between these two nodes.

From nodeA to nodeC
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://A           ## "n" denotes Node             
printNodeArray: full path item 1 = [e] http://likes       ## "e" denotes Edge (Link)
printNodeArray: full path item 2 = [n] http://B
printNodeArray: full path item 3 = [e] http://likes
printNodeArray: full path item 4 = [n] http://C
 
From nodeA to nodeD
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://A
printNodeArray: full path item 1 = [e] http://likes
printNodeArray: full path item 2 = [n] http://B
printNodeArray: full path item 3 = [e] http://likes
printNodeArray: full path item 4 = [n] http://C
printNodeArray: full path item 5 = [e] http://likes
printNodeArray: full path item 6 = [n] http://D
 
From nodeA to nodeF
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://A
printNodeArray: full path item 1 = [e] http://likes
printNodeArray: full path item 2 = [n] http://B
printNodeArray: full path item 3 = [e] http://dislikes
printNodeArray: full path item 4 = [n] m1
printNodeArray: full path item 5 = [e] http://likes
printNodeArray: full path item 6 = [n] http://F
 
From ano to nodeC
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] m1
printNodeArray: full path item 1 = [e] http://dislikes
printNodeArray: full path item 2 = [n] http://B
printNodeArray: full path item 3 = [e] http://likes
printNodeArray: full path item 4 = [n] http://C
 
From ano to nodeX
Node Array is null

The underlying RDF graph view (SEMM_<model_name> or RDFM_<model_name>) cannot be used directly by NDM functions, and so SemNetworkAnalyst creates necessary tables that contain the nodes and links that are derived from a given RDF graph. These tables are not updated automatically when the RDF graph changes; rather, you can set the cleanup parameter in SemNetworkAnalyst.getInstance to true, to remove old node and link tables and to rebuild updated tables.

Example 7-4 implements the NDM' nearestNeighbors function on top of semantic data. This gets a NetworkAnalyst object from the SemNetworkAnalyst instance, gets the node ID, creates PointOnNet objects, and processes LogicalSubPath objects.

Example 7-4 Implementing NDM nearestNeighbors Function on Top of Semantic Data

%cat TestNearestNeighbor.java 
 
import java.io.*;
import java.util.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import com.hp.hpl.jena.util.iterator.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.update.*;
import com.hp.hpl.jena.sparql.core.DataSourceImpl;
 
import oracle.spatial.rdf.client.jena.*;
 
import oracle.spatial.rdf.client.jena.SemNetworkAnalyst;
import oracle.spatial.network.lod.LODGoalNode;
import oracle.spatial.network.lod.LODNetworkConstraint;
import oracle.spatial.network.lod.NetworkAnalyst;
import oracle.spatial.network.lod.PointOnNet;
import oracle.spatial.network.lod.LogicalSubPath;
 
 
/**
 * This class implements a nearestNeighbors function on top of semantic data
 * using public APIs provided in SemNetworkAnalyst and Oracle Spatial NDM
 */
public class TestNearestNeighbor
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
 
    PrintStream psOut = System.out;
 
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    
    String szModelName = "test_nn";
    // First construct a TBox and load a few axioms
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, szModelName);
    String insertString =  
      " PREFIX my:  <http://my.com/> " +
      " INSERT DATA "                             +
      " { my:A   my:likes my:B .                " +
      "   my:A   my:likes my:C .                " +
      "   my:A   my:knows my:D .                " +
      "   my:A   my:dislikes my:X .             " +
      "   my:A   my:dislikes my:Y .             " +
      "   my:C   my:likes my:E .                " +
      "   my:C   my:likes my:F .                " +
      "   my:C   my:dislikes my:M .             " +
      "   my:D   my:likes my:G .                " +
      "   my:D   my:likes my:H .                " +
      "   my:F   my:likes my:M .                " +
      " }   ";
    UpdateAction.parseExecute(insertString,  model);
 
    GraphOracleSem g = model.getGraph();
    g.commitTransaction();
    g.setDOP(4);
 
    HashMap<Node, Double> costMap = new HashMap<Node, Double>();
    costMap.put(Node.createURI("http://my.com/likes"),    Double.valueOf(1.0));
    costMap.put(Node.createURI("http://my.com/dislikes"), Double.valueOf(4.0));
    costMap.put(Node.createURI("http://my.com/knows"),    Double.valueOf(2.0));
 
    SemNetworkAnalyst sna = SemNetworkAnalyst.getInstance(
        g,     // source RDF graph
        true,  // directed graph
        true,  // cleanup old Node/Link tables
        costMap
        );
 
    Node nodeStart = Node.createURI("http://my.com/A");
    long origNodeID = sna.getNodeID(nodeStart);
 
    long[] lIDs = {origNodeID};
 
    // translate from the original ID
    long nodeID = (sna.mapNodeIDs(lIDs))[0]; 
 
    NetworkAnalyst networkAnalyst = sna.getEmbeddedNetworkAnalyst();
 
    LogicalSubPath[] lsps = networkAnalyst.nearestNeighbors(
      new PointOnNet(nodeID),      // startPoint
      6,                           // numberOfNeighbors
      1,                           // searchLinkLevel
      1,                           // targetLinkLevel
      (LODNetworkConstraint) null, // constraint
      (LODGoalNode) null           // goalNodeFilter
      );
 
    if (lsps != null) {
      for (int idx = 0; idx < lsps.length; idx++) {
        LogicalSubPath lsp = lsps[idx];
        Node[] nodePath = sna.processLogicalSubPath(lsp, nodeStart);
        psOut.println("Path " + idx);
        printNodeArray(nodePath, psOut);
      }
    }
 
    g.close();
    sna.close();
    oracle.dispose();
  }
 
 
  public static void printNodeArray(Node[] nodeArray, PrintStream psOut)
  {
    if (nodeArray == null) {
      psOut.println("Node Array is null");
      return;
    }
    if (nodeArray.length == 0) {
      psOut.println("Node Array is empty");
    }
    int iFlag = 0;
    psOut.println("printNodeArray: full path starts");
    for (int iHops = 0; iHops < nodeArray.length; iHops++) {
      psOut.println("printNodeArray: full path item " + iHops + " = "
          + ((iFlag == 0) ? "[n] ":"[e] ") + nodeArray[iHops]);
      iFlag = 1 - iFlag;
    }
  }
}

The output of Example 7-4 is as follows.

Path 0
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://my.com/A
printNodeArray: full path item 1 = [e] http://my.com/likes
printNodeArray: full path item 2 = [n] http://my.com/C
 
Path 1
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://my.com/A
printNodeArray: full path item 1 = [e] http://my.com/likes
printNodeArray: full path item 2 = [n] http://my.com/B
 
Path 2
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://my.com/A
printNodeArray: full path item 1 = [e] http://my.com/knows
printNodeArray: full path item 2 = [n] http://my.com/D
 
Path 3
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://my.com/A
printNodeArray: full path item 1 = [e] http://my.com/likes
printNodeArray: full path item 2 = [n] http://my.com/C
printNodeArray: full path item 3 = [e] http://my.com/likes
printNodeArray: full path item 4 = [n] http://my.com/E
 
Path 4
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://my.com/A
printNodeArray: full path item 1 = [e] http://my.com/likes
printNodeArray: full path item 2 = [n] http://my.com/C
printNodeArray: full path item 3 = [e] http://my.com/likes
printNodeArray: full path item 4 = [n] http://my.com/F
 
Path 5
printNodeArray: full path starts
printNodeArray: full path item 0 = [n] http://my.com/A
printNodeArray: full path item 1 = [e] http://my.com/knows
printNodeArray: full path item 2 = [n] http://my.com/D
printNodeArray: full path item 3 = [e] http://my.com/likes
printNodeArray: full path item 4 = [n] http://my.com/H

7.9.1 Generating Contextual Information about a Path in a Graph

It is sometimes useful to see contextual information about a path in a graph, in addition to the path itself. The buildSurroundingSubGraph method in the SemNetworkAnalyst class can output a DOT file (graph description language file, extension .gv) into the specified Writer object. For each node in the path, up to ten direct neighbors are used to produce a surrounding subgraph for the path. Example 7-5 shows the usage, specifically the output from the analytical functions used in Example 7-3, "Performing Analytical functions on RDF Data".

Example 7-5 Generating a DOT File with Contextual Information

nodeArray = sna.shortestPathDijkstra(nodeA, nodeD);
printNodeArray(nodeArray, psOut);
 
FileWriter dotWriter = new FileWriter("Shortest_Path_A_to_D.gv");
sna.buildSurroundingSubGraph(nodeArray, dotWriter);

The generated output DOT file from Example 7-5 is straightforward, as shown in the following example:

% cat Shortest_Path_A_to_D.gv
digraph { rankdir = LR; charset="utf-8"; 
 
"Rhttp://A" [ label="http://A" shape=rectangle,color=red,style = filled, ];
"Rhttp://B" [ label="http://B" shape=rectangle,color=red,style = filled, ];
"Rhttp://A" -> "Rhttp://B" [ label="http://likes"  color=red, style=bold, ];
"Rhttp://C" [ label="http://C" shape=rectangle,color=red,style = filled, ];
"Rhttp://A" -> "Rhttp://C" [ label="http://dislikes" ];
"Rhttp://D" [ label="http://D" shape=rectangle,color=red,style = filled, ];
"Rhttp://A" -> "Rhttp://D" [ label="http://differs" ];
"Rhttp://B" -> "Rhttp://C" [ label="http://likes"  color=red, style=bold, ];
"Rm1" [ label="m1" shape=ellipse,color=blue, ];
"Rhttp://B" -> "Rm1" [ label="http://dislikes" ];
"Rm1" -> "Rhttp://B" [ label="http://dislikes" ];
"Rhttp://C" -> "Rhttp://D" [ label="http://likes"  color=red, style=bold, ];
"Rhttp://E" [ label="http://E" shape=ellipse,color=blue, ];
"Rhttp://C" -> "Rhttp://E" [ label="http://knows" ];
"Rm1" -> "Rhttp://D" [ label="http://likes" ];
}

You can also use methods in the SemNetworkAnalyst and GraphOracleSem classes to produce more sophisticated visualization of the analytical function output.

You can convert the preceding DOT file into a variety of image formats. Figure 7-1 is an image representing the information in the preceding DOT file.

Figure 7-1 Visual Representation of Analytical Function Output

Description of Figure 7-1 follows
Description of "Figure 7-1 Visual Representation of Analytical Function Output"

7.10 Support for Server-Side APIs

This section describes some of the Oracle Database Semantic Technologies features that are exposed by the Jena Adapter. For comprehensive documentation of the API calls that support the available features, see the Jena Adapter reference information (Javadoc). For additional information about the server-side features exposed by the Jena Adapter, see the relevant chapters in this manual.

7.10.1 Virtual Models Support

Virtual models (explained in Section 1.3.8) are specified in the GraphOracleSem constructor, and they are handled transparently. If a virtual model exists for the model-rulebase combination, it is used in query answering; if such a virtual model does not exist, it is created in the database.

Note:

Virtual model support through the Jena Adapter is available only with Oracle Database Release 11.2 or later.

The following example reuses an existing virtual model.

String modelName = "EX";
String m1 = "EX_1";
 
ModelOracleSem defaultModel = 
  ModelOracleSem.createOracleSemModel(oracle, modelName);
 
// create these models in case they don't exist
ModelOracleSem model1 = ModelOracleSem.createOracleSemModel(oracle, m1);
 
String vmName = "VM_" + modelName;
 
 
//create a virtual model containing EX and EX_1
oracle.executeCall(
"begin sem_apis.create_virtual_model(?,sem_models('"+ m1 + "','"+ modelName+"'),null); end;",vmName);
 
String[] modelNames = {m1};
String[] rulebaseNames = {};
 
Attachment attachment = Attachment.createInstance(modelNames, rulebaseNames, 
InferenceMaintenanceMode.NO_UPDATE, QueryOptions.ALLOW_QUERY_VALID_AND_DUP);
 
// vmName is passed to the constructor, so GraphOracleSem will use the virtual 
// model named vmname (if the current user has read privileges on it)
GraphOracleSem graph = new GraphOracleSem(oracle, modelName, attachment, vmName);
graph.add(Triple.create(Node.createURI("urn:alice"),
                        Node.createURI("http://xmlns.com/foaf/0.1/mbox"),
                        Node.createURI("mailto:alice@example")));
ModelOracleSem model = new ModelOracleSem(graph);          
 
String queryString =
 
   " SELECT ?subject ?object WHERE { ?subject ?p ?object } ";
 
Query query = QueryFactory.create(queryString) ;
QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
 
try {
   ResultSet results = qexec.execSelect() ;
   for ( ; results.hasNext() ; ) {
      QuerySolution soln = results.nextSolution() ;
      psOut.println("soln " + soln);
   }
} 
finally { 
   qexec.close() ; 
}
 
OracleUtils.dropSemanticModel(oracle, modelName);
OracleUtils.dropSemanticModel(oracle, m1);
 
oracle.dispose();

You can also use the GraphOracleSem constructor to create a virtual model, as in the following example:

GraphOracleSem graph = new GraphOracleSem(oracle, modelName, attachment, true);

In this example, the fourth parameter (true) specifies that a virtual model needs to be created for the specified modelName and attachment.

7.10.2 Connection Pooling Support

Oracle Database Connection Pooling is provided through the Jena Adapter OraclePool class. Once this class is initialized, it can return Oracle objects out of its pool of available connections. Oracle objects are essentially database connection wrappers. After dispose is called on the Oracle object, the connection is returned to the pool. More information about using OraclePool can be found in the API reference information (Javadoc).

The following example sets up an OraclePool object with five (5) initial connections.

public static void main(String[] args) throws Exception
  {      
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
 
    // test with connection properties 
    java.util.Properties prop = new java.util.Properties();
    prop.setProperty("MinLimit", "2");     // the cache size is 2 at least 
    prop.setProperty("MaxLimit", "10");
    prop.setProperty("InitialLimit", "2"); // create 2 connections at startup
    prop.setProperty("InactivityTimeout", "1800");    //  seconds
    prop.setProperty("AbandonedConnectionTimeout", "900");  //  seconds
    prop.setProperty("MaxStatementsLimit", "10");
    prop.setProperty("PropertyCheckInterval", "60"); // seconds
 
    System.out.println("Creating OraclePool");
    OraclePool op = new OraclePool(szJdbcURL, szUser, szPasswd, prop, 
               "OracleSemConnPool");
    System.out.println("Done creating OraclePool");
 
    // grab an Oracle and do something with it
    System.out.println("Getting an Oracle from OraclePool");
    Oracle oracle = op.getOracle();
    System.out.println("Done");
    System.out.println("Is logical connection:" +
        oracle.getConnection().isLogicalConnection());
    GraphOracleSem g = new GraphOracleSem(oracle, szModelName);
    g.add(Triple.create(Node.createURI("u:John"), 
                        Node.createURI("u:parentOf"), 
                        Node.createURI("u:Mary")));
    g.close();
    // return the Oracle back to the pool
    oracle.dispose();
    
    // grab another Oracle and do something else 
    System.out.println("Getting an Oracle from OraclePool");
    oracle = op.getOracle();
    System.out.println("Done");
    System.out.println("Is logical connection:" +
        oracle.getConnection().isLogicalConnection());
    g = new GraphOracleSem(oracle, szModelName);
    g.add(Triple.create(Node.createURI("u:John"), 
                        Node.createURI("u:parentOf"), 
                        Node.createURI("u:Jack")));
    g.close();
    
    OracleUtils.dropSemanticModel(oracle, szModelName); 
    
    // return the Oracle object back to the pool
    oracle.dispose();
}

7.10.3 Semantic Model PL/SQL Interfaces

Several semantic PL/SQL subprograms are available through the Jena Adapter. Table 7-2 lists the subprograms and their corresponding Java class and methods.

Table 7-2 PL/SQL Subprograms and Corresponding Jena Adapter Java Class and Methods

PL/SQL Subprogram Corresponding Java Class and Methods

SEM_APIS.DROP_SEM_MODEL

OracleUtils.dropSemanticModel

SEM_APIS.MERGE_MODELS

OracleUtils.mergeModels

SEM_APIS.SWAP_NAMES

OracleUtils.swapNames

SEM_APIS.REMOVE_DUPLICATES

OracleUtils.removeDuplicates

SEM_APIS.RENAME_MODEL

OracleUtils.renameModels


For information about these PL/SQL utility subprograms, see the reference information in Chapter 9. For information about the corresponding Java class and methods, see the Jena Adapter API Reference documentation (Javadoc).

7.10.4 Inference Options

You can add options to entailment calls by using the following methods in the Attachment class (in package oracle.spatial.rdf.client.jena):

public void setUseLocalInference(boolean useLocalInference)
public boolean getUseLocalInference()
 
public void setDefGraphForLocalInference(String defaultGraphName)
public String getDefGraphForLocalInference()
 
public String getInferenceOption()
public void setInferenceOption(String inferenceOption)

For more information about these methods, see the Javadoc.

Example 7-6 enables parallel inference (with a degree of 4) and RAW format when creating an entailment. The example also uses the performInference method to create the entailment (comparable to using the SEM_APIS.CREATE_ENTAILMENT PL/SQL procedure).

Example 7-6 Specifying Inference Options

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import com.hp.hpl.jena.util.iterator.*;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.update.*;
import com.hp.hpl.jena.sparql.core.DataSourceImpl;
 
public class TestNewInference
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
 
    PrintStream psOut = System.out;
 
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    
    String szTBoxName = "test_new_tbox";
    {
      // First construct a TBox and load a few axioms
      ModelOracleSem modelTBox = ModelOracleSem.createOracleSemModel(oracle, szTBoxName);
      String insertString =  
        " PREFIX my:  <http://my.com/> " +
        " PREFIX rdfs:  <http://www.w3.org/2000/01/rdf-schema#> " +
        " INSERT DATA "                                     +
        " { my:C1  rdfs:subClassOf my:C2 .                " +
        "   my:C2  rdfs:subClassOf my:C3 .                " +
        "   my:C3  rdfs:subClassOf my:C4 .                " +
        " }   ";
      UpdateAction.parseExecute(insertString,  modelTBox);
      modelTBox.close();
    }
 
    String szABoxName = "test_new_abox";
    {
      // Construct an ABox and load a few quads
      ModelOracleSem modelABox = ModelOracleSem.createOracleSemModel(oracle, szABoxName);
      DatasetGraphOracleSem dataset = DatasetGraphOracleSem.createFrom(modelABox.getGraph());
      modelABox.close();
 
      String insertString =  
        " PREFIX my:    <http://my.com/> " +
        " PREFIX rdf:   <http://www.w3.org/1999/02/22-rdf-syntax-ns#> " +
        " INSERT DATA                                                 " +
        " { GRAPH my:G1 { my:I1  rdf:type my:C1 .                     " +
        "                 my:I2  rdf:type my:C2 .                     " +
        "               }                                             " +
        " };                                                          " +
        " INSERT DATA                                                 " +
        " { GRAPH my:G2 { my:J1  rdf:type my:C3 .                     " +
        "               }                                             " +
        " }    ";
      UpdateAction.parseExecute(insertString,  dataset);
      dataset.close();
    }
 
 
    String[] attachedModels = new String[1];
    attachedModels[0] = szTBoxName;
 
    String[] attachedRBs = {"OWL2RL"};
 
    Attachment attachment = Attachment.createInstance(
        attachedModels, attachedRBs,
        InferenceMaintenanceMode.NO_UPDATE,
        QueryOptions.ALLOW_QUERY_INVALID);
 
    // We are going to run named graph based local inference
    attachment.setUseLocalInference(true);
 
    // Set the default graph (TBox)
    attachment.setDefGraphForLocalInference(szTBoxName);
 
    // Set the inference option to use parallel inference 
    // with a degree of 4, and RAW format.
    attachment.setInferenceOption("DOP=4,RAW8=T");
 
    GraphOracleSem graph = new GraphOracleSem(
        oracle, 
        szABoxName, 
        attachment
        );
    DatasetGraphOracleSem dsgos = DatasetGraphOracleSem.createFrom(graph);
    graph.close();
 
    // Invoke create_entailment PL/SQL API
    dsgos.performInference();
 
    psOut.println("TestNewInference: # of inferred graph " + 
        Long.toString(dsgos.getInferredGraphSize()));
 
    String queryString = 
      " SELECT ?g ?s ?p ?o WHERE {  GRAPH ?g {?s ?p ?o } } " ;
 
    Query query = QueryFactory.create(queryString, Syntax.syntaxARQ);
    QueryExecution qexec = QueryExecutionFactory.create(
        query, DataSourceImpl.wrap(dsgos));
    ResultSet results = qexec.execSelect();
 
    ResultSetFormatter.out(psOut, results);
 
    dsgos.close();
    oracle.dispose();
  }
}

The output of Example 7-6 is as follows.

TestNewInference: # of inferred graph 9
 
--------------------------------------------------------------------------------------------------------------------
| g                  | s                  | p                                                 | o                  |
====================================================================================================================
| <http://my.com/G1> | <http://my.com/I2> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C3> |
| <http://my.com/G1> | <http://my.com/I2> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C2> |
| <http://my.com/G1> | <http://my.com/I2> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C4> |
| <http://my.com/G1> | <http://my.com/I1> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C3> |
| <http://my.com/G1> | <http://my.com/I1> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C1> |
| <http://my.com/G1> | <http://my.com/I1> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C2> |
| <http://my.com/G1> | <http://my.com/I1> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C4> |
| <http://my.com/G2> | <http://my.com/J1> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C3> |
| <http://my.com/G2> | <http://my.com/J1> | <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> | <http://my.com/C4> |
--------------------------------------------------------------------------------------------------------------------

For information about using OWL inferencing, see Section 2.2.

7.10.5 PelletInfGraph Class Support Deprecated

The support for the PelletInfGraph class within the Jena Adapter is deprecated. You should instead use the more optimized Oracle/Pellet integration through the PelletDb OWL 2 reasoner for Oracle Database 11g. (For information about PelletDb, see http://clarkparsia.com/pelletdb/.)

7.11 Bulk Loading Using the Jena Adapter

To load thousands to hundreds of thousands of RDF/OWL data files into an Oracle database, you can use the prepareBulk and completeBulk methods in the OracleBulkUpdateHandler Java class to simplify the task.

The addInBulk method in the OracleBulkUpdateHandler class can load triples of a graph or model into an Oracle database in bulk loading style. If the graph or model is a Jena in-memory graph or model, the operation is limited by the size of the physical memory. The prepareBulk method bypasses the Jena in-memory graph or model and takes a direct input stream to an RDF data file, parses the data, and load the triples into an underlying staging table. If the staging table and an accompanying table for storing long literals do not already exist, they are created automatically.

The prepareBulk method can be invoked multiple times to load multiple data files into the same underlying staging table. It can also be invoked concurrently, assuming the hardware system is balanced and there are multiple CPU cores and sufficient I/O capacity.

Once all the data files are processed by the prepareBulk method, you can invoke completeBulk to load all the data into the semantic network.

Example 7-7 shows how to load all data files in directory dir_1 into the underlying staging table. Long literals are supported and will be stored in a separate table. The data files can be compressed using GZIP to save storage space, and the prepareBulk method can detect automatically if a data file is compressed using GZIP or not.

Example 7-7 Loading Data into the Staging Table (prepareBulk)

Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
GraphOracleSem graph = new GraphOracleSem(oracle, szModelName);
 
PrintStream psOut = System.out;

String dirname = "dir_1";
File fileDir = new File(dirname);
String[] szAllFiles = fileDir.list();

// loop through all the files in a directory
for (int idx = 0; idx < szAllFiles.length; idx++) {
  String szIndFileName = dirname + File.separator + szAllFiles[idx];
  psOut.println("process to [ID = " + idx + " ] file " + szIndFileName);
  psOut.flush();
 
  try {
    InputStream is = new FileInputStream(szIndFileName);
    graph.getBulkUpdateHandler().prepareBulk(
        is,                    // input stream
        "http://example.com",  // base URI
        "RDF/XML",             // data file type: can be RDF/XML, N-TRIPLE, etc.
        "SEMTS",               // tablespace
        null,                  // flags
        null,                  // listener
        null                   // staging table name.
        );
    is.close();
  }
  catch (Throwable t) {
    psOut.println("Hit exception " + t.getMessage());
  }
}
 
graph.close();
oracle.dispose();

The code in Example 7-7, starting from creating a new Oracle object and ending with disposing of the Oracle object, can be executed in parallel. Assume there is a quad-core CPU and enough I/O capacity on the database hardware system; you can divide up all the data files and save them into four separate directories: dir_1, dir_2, dir_3, and dir_4. Four Java threads of processes can be started to work on those directories separately and concurrently. (For more information, see Section 7.11.1, "Using prepareBulk in Parallel (Multithreaded) Mode".)

After all data files are processed, you can invoke, just once, the completeBulk method to load the data from staging table into the semantic network, as shown in Example 7-8. Triples with long literals will be loaded also.

Example 7-8 Loading Data from the Staging Table into the Semantic Network (completeBulk)

graph.getBulkUpdateHandler().completeBulk(
  null,  // flags for invoking SEM_APIS.bulk_load_from_staging_table
  null   // staging table name
);

The prepareBulk method can also take a Jena model as an input data source, in which case triples in that Jena model are loaded into the underlying staging table. For more information, see the Javadoc.

In addition to loading triples from Jena models and data files, the prepareBulk method supports RDFa, as shown in Example 7-9. (RDFa is explained in http://www.w3.org/TR/xhtml-rdfa-primer/.)

Example 7-9 Using prepareBulk with RDFa

graph.getBulkUpdateHandler().prepareBulk(
  rdfaUrl,   // url to a web page using RDFa
  "SEMTS",   // tablespace
  null,      // flags
  null,      // listener
  null       // staging table name
);

To parse RDFa, the relevant java-rdfa libraries must be included in the classpath. No additional setup or API calls are required. (For information about java-rdfa, see http://www.rootdev.net/maven/projects/java-rdfa/ and the other topics there under Project Information.)

Note that if the rdfaUrl is located outside a firewall, you may need to set the following HTTP Proxy-related Java VM properties:

-Dhttp.proxyPort=...
-Dhttp.proxyHost=...

The preceding examples in this section load triple data into a single graph. Loading quad data that may span across multiple named graphs (such as data in NQUADS format) requires the use of the DatasetGraphOracleSem class. The DatasetGraphOracleSem class does not use the BulkUpdateHandler API, but does provide a similar prepareBulk and completeBulk interface, as shown in Example 7-10.

Example 7-10 Loading Quads into a DatasetGraph

Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
 
// Can only create DatasetGraphOracleSem from an existing GraphOracleSem
GraphOracleSem graph = new GraphOracleSem(oracle, szModelName);
DatasetGraphOracleSem dataset = DatasetGraphOracleSem.createFrom(graph);
 
// Don't need graph anymore, close it to free resources
graph.close();
 
try {
    InputStream is = new FileInputStream(szFileName);
    // load NQUADS file into a staging table. This file can be gzipp'ed.
    dataset.prepareBulk(
        is,                // input stream
        "http://my.base/", // base URI
        "N-QUADS",         // data file type; can be "TRIG"
        "SEMTS",           // tablespace
        null,              // flags
        null,              // listener
        null,              // staging table name
        false              // truncate staging table before load
    );
    // Load quads from staging table into the dataset
    dataset.completeBulk(
        null, // flags; can be "PARSE PARALLEL_CREATE_INDEX PARALLEL=4 
              //                mbv_method=shadow" on a quad core machine
        null  // staging table name
    );
} 
catch (Throwable t) {
    System.out.println("Hit exception " + t.getMessage());
}
finally {
    dataset.close();
    oracle.dispose();
}

7.11.1 Using prepareBulk in Parallel (Multithreaded) Mode

Example 7-7, "Loading Data into the Staging Table (prepareBulk)" provided a way to load, sequentially, a set of files under a file system directory to an Oracle Database table (staging table). Example 7-11 loads, concurrently, a set of files to an Oracle table (staging table). The degree of parallelism is controlled by the input parameter iMaxThreads.

On a balanced hardware setup with 4 or more CPU cores, setting iMaxThreads to 8 (or 16) can improve significantly the speed of prepareBulk operation when there are many data files to be processed.

Example 7-11 Using prepareBulk with iMaxThreads

public void testPrepareInParallel(String jdbcUrl, String user,
                       String password, String modelName,
                       String lang,
                       String tbs,
                       String dirname,
                       int iMaxThreads,
                       PrintStream psOut)
   throws SQLException, IOException,  InterruptedException
 {
   File dir = new File(dirname);
   File[] files = dir.listFiles();

   // create a set of physical Oracle connections and graph objects
   Oracle[] oracles = new Oracle[iMaxThreads];
   GraphOracleSem[] graphs = new GraphOracleSem[iMaxThreads];
   for (int idx = 0; idx < iMaxThreads; idx++) {
     oracles[idx] = new Oracle(jdbcUrl, user, password);
     graphs[idx] = new GraphOracleSem(oracles[idx], modelName);
   }

   PrepareWorker[] workers = new PrepareWorker[iMaxThreads];
   Thread[] threads = new Thread[iMaxThreads];
   for (int idx = 0; idx < iMaxThreads; idx++) {
     workers[idx] = new PrepareWorker(
         graphs[idx],
         files,
         idx,
         iMaxThreads,
         lang,
         tbs,
         psOut
         );
     threads[idx] = new Thread(workers[idx], workers[idx].getName());
     psOut.println("testPrepareInParallel: PrepareWorker " + idx + " running");
     threads[idx].start();
   }

   psOut.println("testPrepareInParallel: all threads started");

   for (int idx = 0; idx < iMaxThreads; idx++) {
     threads[idx].join();
   }
   for (int idx = 0; idx < iMaxThreads; idx++) {
     graphs[idx].close();
     oracles[idx].dispose();
   }
 }

 static class PrepareWorker implements Runnable
 {
   GraphOracleSem graph = null;
   int idx;
   int threads;
   File[] files = null;
   String lang = null;
   String tbs  = null;
   PrintStream psOut;

   public void run()
   {
     long lStartTime = System.currentTimeMillis();
     for (int idxFile = idx; idxFile < files.length; idxFile += threads) {
       File file = files[idxFile];
       try {
         FileInputStream fis = new FileInputStream(file);
         graph.getBulkUpdateHandler().prepareBulk(
             fis,
             "http://base.com/",
             lang,
             tbs,
             null,                   // flags
             new MyListener(psOut),  // listener
             null                    // table name
             );
         fis.close();
       }
       catch (Exception e) {
         psOut.println("PrepareWorker: thread ["+getName()+"] error "+ e.getMessage());
       }
       psOut.println("PrepareWorker: thread ["+getName()+"] done to "
           + idxFile + ", file = " + file.toString()
           + " in (ms) " + (System.currentTimeMillis() - lStartTime));
     }
   }

   public PrepareWorker(GraphOracleSem graph,
                        File[] files,
                        int idx,
                        int threads,
                        String lang,
                        String tbs,
                        PrintStream psOut)
   {
     this.graph   = graph;
     this.files   = files;
     this.psOut   = psOut;
     this.idx     = idx;
     this.threads = threads;
     this.files   = files;
     this.lang    = lang;
     this.tbs     = tbs ;
   }

   public String getName()
   {
     return "PrepareWorker" + idx;
   }
 } 
 
 static class MyListener implements StatusListener 
 {
   PrintStream m_ps = null;
   public MyListener(PrintStream ps) { m_ps = ps; }
   long lLastBatch = 0;

   public void statusChanged(long count)
   {
     if (count - lLastBatch >= 10000) {
       m_ps.println("process to " + Long.toString(count));
       lLastBatch = count;
     }
   }

   public int illegalStmtEncountered(Node graphNode, Triple triple, long count)
   {
     m_ps.println("hit illegal statement with object " + triple.getObject().toString());
     return 0; // skip it
   }
 }

7.11.2 Handling Illegal Syntax During Data Loading

You can skip illegal triples and quads when using prepareBulk. This feature is useful if the source RDF data may contain syntax errors. In Example 7-12, a customized implementation of the StatusListener interface (defined in package oracle.spatial.rdf.client.jena) is passed as a parameter to prepareBulk. In this example, the illegalStmtEncountered method prints the object field of the illegal triple, and returns 0 so that prepareBulk can skip that illegal triple and move on.

Example 7-12 Skipping Triples with Illegal Syntax

....
 
Oracle oracle = new Oracle(jdbcUrl, user, password);
GraphOracleSem graph = new GraphOracleSem(oracle, modelName);
PrintStream psOut = System.err;
 
graph.getBulkUpdateHandler().prepareBulk(
  new FileInputStream(rdfDataFilename),
  "http://base.com/",    // base 
  lang,                  // data format, can be "N-TRIPLES" "RDF/XML" ...
  tbs,                   // tablespace name
  null,                  // flags
  new MyListener(psOut), // call back to show progress and also process illegal triples/quads
  null,                  // tableName, if null use default names
  false                  // truncate existing staging tables
  );
 
 graph.close();
 oracle.dispose();
 .... 
 
 // A customized StatusListener interface implementation
  public class MyListener implements StatusListener
 {
   PrintStream m_ps = null;
   public MyListener(PrintStream ps) { m_ps = ps; }
 
   public void statusChanged(long count)
   {
     // m_ps.println("process to " + Long.toString(count));
   }
 
  public int illegalStmtEncountered(Node graphNode, Triple triple, long count)
  {
    m_ps.println("hit illegal statement with object " + triple.getObject().toString());
    return 0; // skip it
  }
 }

7.12 Automatic Variable Renaming

Previously, variable names used in SPARQL queries were passed directly on to Oracle Database as a part of a SQL statement. If the variable names included a SQL or PL/SQL reserved keyword, the query failed to execute. For example, the following SPARQL query used to fail because the word date as a special meaning to the Oracle Database SQL processing engine.

select ?date { :event  :happenedOn ?date }

Currently, this query does not fail, because a "smart scan" is performed and automatic replacement is done on certain reserved variable names (or variable names that are very long) before the query is sent to Oracle database for execution. The replacement is based on a list of reserved keywords that are stored in the following file embedded in sdordfclient.jar:

oracle/spatial/rdf/client/jena/oracle_sem_reserved_keywords.lst

This file contains over 100 entries, and you can edit the file to add entries if necessary.

The following are examples of SPARQL queries that use SQL or PL/SQL reserved keywords as variables, and that will succeed because of automatic variable renaming:

7.13 JavaScript Object Notation (JSON) Format Support

JavaScript Object Notation (JSON) format is supported for SPARQL query responses. JSON data format is simple, compact, and well suited for JavaScript programs.

For example, assume the following Java code snippet, which calls the ResultSetFormatter.outputAsJSON method:

Oracle oracle = new Oracle(jdbcUrl, user, password);
 
GraphOracleSem graph = new GraphOracleSem(oracle, modelName);
ModelOracleSem model = new ModelOracleSem(graph);
 
graph.add(new Triple(
                   Node.createURI("http://ds1"),
                   Node.createURI("http://dp1"),
                   Node.createURI("http://do1")
                   )
         );
 
graph.add(new Triple(
                   Node.createURI("http://ds2"),
                   Node.createURI("http://dp2"),
                   Node.createURI("http://do2")
                   )
         );
graph.commitTransaction();
 
Query q = QueryFactory.create("select ?s ?p ?o where {?s ?p ?o}",
                              Syntax.syntaxARQ);
QueryExecution qexec = QueryExecutionFactory.create(q, model);
 
ResultSet results = qexec.execSelect();
ResultSetFormatter.outputAsJSON(System.out, results);

The JSON output is as follows:

{
  "head": {
    "vars": [ "s" , "p" , "o" ]
  } ,
  "results": {
    "bindings": [
      {
        "s": { "type": "uri" , "value": "http://ds1" } ,
        "p": { "type": "uri" , "value": "http://dp1" } ,
        "o": { "type": "uri" , "value": "http://do1" }
      } ,
      {
        "s": { "type": "uri" , "value": "http://ds2" } ,
        "p": { "type": "uri" , "value": "http://dp2" } ,
        "o": { "type": "uri" , "value": "http://do2" }
      }
    ]
  }
}

The preceding example can be changed as follows to query a remote SPARQL endpoint instead of directly against an Oracle database. (If the remote SPARQL endpoint is outside a firewall, then the HTTP Proxy probably needs to be set.)

Query q = QueryFactory.create("select ?s ?p ?o where {?s ?p ?o}",
                              Syntax.syntaxARQ);
QueryExecution qe = QueryExecutionFactory.sparqlService(sparqlURL, q);
 
ResultSet results = qexec.execSelect();
ResultSetFormatter.outputAsJSON(System.out, results);

To extend the first example in this section to named graphs, the following code snippet adds two quads to the same Oracle model, executes a named graph-based SPARQL query, and serializes the query output into JSON format:

DatasetGraphOracleSem dsgos = DatasetGraphOracleSem.createFrom(graph);
graph.close();
 
dsgos.add(new Quad(Node.createURI("http://g1"),
                   Node.createURI("http://s1"),
                   Node.createURI("http://p1"),
                   Node.createURI("http://o1")
                   )
         );
dsgos.add(new Quad(Node.createURI("http://g2"),
                   Node.createURI("http://s2"),
                   Node.createURI("http://p2"),
                   Node.createURI("http://o2")
                   )
         );
 
Query q1 = QueryFactory.create(
  "select ?g ?s ?p ?o where { GRAPH ?g {?s ?p ?o} }");
 
QueryExecution qexec1 = QueryExecutionFactory.create(q1,
    DataSourceImpl.wrap(dsgos));
 
ResultSet results1 = qexec1.execSelect();
ResultSetFormatter.outputAsJSON(System.out, results1);
 
dsgos.close();
oracle.dispose();

The JSON output is as follows:

{
  "head": {
    "vars": [ "g" , "s" , "p" , "o" ]
  } ,
  "results": {
    "bindings": [
      {
        "g": { "type": "uri" , "value": "http://g1" } ,
        "s": { "type": "uri" , "value": "http://s1" } ,
        "p": { "type": "uri" , "value": "http://p1" } ,
        "o": { "type": "uri" , "value": "http://o1" }
      } ,
      {
        "g": { "type": "uri" , "value": "http://g2" } ,
        "s": { "type": "uri" , "value": "http://s2" } ,
        "p": { "type": "uri" , "value": "http://p2" } ,
        "o": { "type": "uri" , "value": "http://o2" }
      }
    ]
  }
}

You can also get a JSON response through HTTP against a Joseki-based SPARQL endpoint, as in the following example. Normally, when executing a SPARQL query against a SPARQL Web service endpoint, the Accept request-head field is set to be application/sparql-results+xml. For JSON output format, replace the Accept request-head field with application/sparql-results+json.

http://hostname:7001/joseki/oracle?query=<URL_ENCODED_SPARQL_QUERY>&output=json

7.14 Other Recommendations and Guidelines

This section contains various recommendations and other information related to SPARQL queries.

7.14.1 BOUND or !BOUND Instead of EXISTS or NOT EXISTS

For better performance, use BOUND or !BOUND instead of EXISTS or NOT EXISTS.

7.14.2 SPARQL 1.1 SELECT Expressions

You can use SPARQL 1.1 SELECT expressions without any significant performance overhead, even if the function is not currently supported within Oracle Database. Examples include the following:

-- Quyery using MD5 and SHA1 functions
PREFIX foaf: <http://xmlns.com/foaf/0.1/>
PREFIX xsd:  <http://www.w3.org/2001/XMLSchema#>
PREFIX eg:   <http://biometrics.example/ns#>
SELECT ?name (md5(?name) as ?name_in_md5) (sha1(?email) as ?sha1) 
WHERE 
{ 
  ?x foaf:name  ?name ; eg:email ?email .
}

-- Quyery using CONCAT function
PREFIX foaf:   <http://xmlns.com/foaf/0.1/>
SELECT ( CONCAT(?G, " ", ?S) AS ?name )
WHERE  
{ 
  ?P foaf:givenName ?G ; foaf:surname ?S 
}

7.14.3 Syntax Involving Bnodes (Blank Nodes)

Syntax involving bnodes can be used freely in query patterns. For example, the following bnode-related syntax is supported at the parser level, so each is equivalent to its full triple-query-pattern-based version.

:x :q [ :p "v" ] .
 
(1 ?x 3 4) :p "w" .
 
(1 [:p :q] ( 2 ) ) .

7.14.4 Limit in the SERVICE Clause

When writing a SPARQL 1.1 federated query, you can set a limit on returned rows in the subquery inside the SERVICE clause. This can effectively constrain the amount of data to be transported between the local repository and the remote SPARQL endpoint.

For example, the following query specifies limit 100 in the subquery in the SERVICE clause:

PREFIX : <http://example.com/>
SELECT ?s ?o 
 WHERE 
     { 
       ?s :name "CA"  
       SERVICE <http://REMOTE_SPARQL_ENDPOINT_HERE>
          {
             select ?s  ?o 
                 {?s :info ?o} 
              limit 100 
           } 
     }

7.14.5 OracleGraphWrapperForOntModel Class for Better Performance

The Jena OntModel class lets you create, modify, and analyze an ontology stored in a Jena model. However, the OntModel implementation is not optimized for semantic data stored in a database. This results in suboptimal performance when using OntModel with an Oracle model. Therefore, the class OracleGraphWrapperForOntModel has been created to alleviate this performance issue.

The OracleGraphWrapperForOntModel class implements the Jena Graph interface and represents a graph backed by an Oracle RDF/OWL model that is meant for use with the Jena OntModel API. The OracleGraphWrapperForOntModel class uses two semantic stores in a hybrid approach for persisting changes and responding to queries. Both semantic stores contain the same data, but one resides in memory while the other resides in the Oracle database.

When queried through OntModel, the OracleGraphWrapperForOntModel graph runs the queries against the in-memory store to improve performance. However, the OracleGraphWrapperForOntModel class persists changes made through OntModel, such as adding or removing classes, by applying changes to both stores.

Due to its hybrid approach, an OracleGraphWrapperForOntModel graph requires that sufficient memory be allocated to the JVM to store a copy of the ontology in memory. In internal experiments, it was found that an ontology with approximately 3 million triples requires 6 or more GB of physical memory.

Example 7-13 shows how to use the OntModel APIs with an existing ontology stored in an Oracle model.

Example 7-13 Using OntModel with Ontology Stored in Oracle Database

// Set up connection to Oracle semantic store and the Oracle model
// containing the ontology
Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
GraphOracleSem oracleGraph = new GraphOracleSem(oracle, szModelName);
 
// Create a new hybrid graph using the oracle graph to persist
// changes.  This method will copy all the data from the oracle graph
// into an in-memory graph, which may significantly increase JVM memory
// usage.
Graph hybridGraph = OracleGraphWrapperForOntModel.getInstance(oracleGraph);
 
// Build a model around the hybrid graph and wrap the model with Jena's
// OntModel
Model model = ModelFactory.createModelForGraph(hybridGraph);
OntModel ontModel = ModelFactory.createOntologyModel(ontModelSpec, model);
 
// Perform operations on the ontology
OntClass personClass = ontModel.createClass("<http://someuri/person>");
ontModel.createIndividual(personClass);
 
// Close resources (will also close oracleGraph)!
hybridGraph.close();
ontModel.close();

Note that any OntModel object created using OracleGraphWrapperForOntModel will not reflect changes made to the underlying Oracle model by another process, through a separate OntModel, or through a separate Oracle graph referencing the same underlying model. All changes to an ontology should go through a single OntModel object and its underlying OracleGraphWrapperForOntModel graph until the model or graph have been closed.

If the default in-memory semantic store used by OracleGraphWrapperForOntModel is not sufficient for an ontology and system, the class provides an interface for specifying a custom graph to use as the in-memory store. Example 7-14 shows how to create an OracleGraphWrapperForOntModel that uses a custom in-memory graph to answer queries from OntModel.

Example 7-14 Using a Custom In-Memory Graph

// Set up connection to Oracle semantic store and the Oracle model
// containing the ontology
Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
GraphOracleSem oracleGraph = new GraphOracleSem(oracle, szModelName);
 
// Create a custom in-memory graph to use instead of the default
// Jena in-memory graph for quickly answering OntModel queries.
// Note that this graph does not *need* to be in-memory, but in-memory
// is preferred.
GraphBase queryGraph = new CustomInMemoryGraphImpl();
 
// Create a new hybrid graph using the oracle graph to persist
// changes and the custom in-memory graph to answer queries. 
// Also set the degree of parallelism to use when copying data from
// the oracle graph to the querying graph.
int degreeOfParallelism = 4;
Graph hybridGraph = OracleGraphWrapperForOntModel.getInstance(oracleGraph, queryGraph, degreeOfParallelism);
 
// Build a model and wrap the model with Jena's OntModel
Model model = ModelFactory.createModelForGraph(hybridGraph);
OntModel ontModel = ModelFactory.createOntologyModel(ontModelSpec, model);
 
// Perform operations on the ontology
// ...
 
// Close resources (will close oracleGraph and queryGraph)!
hybridGraph.close();
ontModel.close();

7.15 Example Queries Using the Jena Adapter

This section includes example queries using the Jena Adapter. Each example is self-contained: it typically creates a model, creates triples, performs a query that may involve inference, displays the result, and drops the model.

This section includes queries that do the following:

To run a query, you must do the following:

  1. Include the code in a Java source file. The examples used in this section are supplied in files in the examples directory of the Jena Adapter download.

  2. Compile the Java source file. For example:

    > javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test.java
    

    Note:

    The javac and java commands must each be on a single command line.
  3. Run the compiled file. For example:

    > java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
    

7.15.1 Test.java: Query Family Relationships

Example 7-15 specifies that John is the father of Mary, and it selects and displays the subject and object in each fatherOf relationship

Example 7-15 Query Family Relationships

import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.query.*;
public class Test {
 
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
          
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    Model model = ModelOracleSem.createOracleSemModel(
      oracle, szModelName);
 
    model.getGraph().add(Triple.create(
          Node.createURI("http://example.com/John"),
          Node.createURI("http://example.com/fatherOf"),
          Node.createURI("http://example.com/Mary")));
    Query query = QueryFactory.create(
        "select ?f ?k WHERE {?f <http://example.com/fatherOf> ?k .}");
    QueryExecution qexec = QueryExecutionFactory.create(query, model);
    ResultSet results = qexec.execSelect();
    ResultSetFormatter.out(System.out, results, query);
    model.close();
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-15, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
 
---------------------------------------------------------
| f                         | k                         |
=========================================================
| <http://example.com/John> | <http://example.com/Mary> |
---------------------------------------------------------

7.15.2 Test6.java: Load OWL Ontology and Perform OWLPrime inference

Example 7-16 loads an OWL ontology and performs OWLPrime inference. Note that the OWL ontology is in RDF/XML format, and after it is loaded into Oracle it will be serialized out in N-TRIPLE form. The example also queries for the number of asserted and inferred triples.

The ontology in this example can be retrieved from several locations, including http://owl.cs.manchester.ac.uk/2008/iswc-tones/ontologies/univ-bench.owl, and it describes roles, resources, and relationships in a university environment.

Example 7-16 Load OWL Ontology and Perform OWLPrime inference

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import oracle.spatial.rdf.client.jena.*;
public class Test6 {
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    Model model = ModelOracleSem.createOracleSemModel(oracle, szModelName);
        
    // load UNIV ontology
    InputStream in = FileManager.get().open("./univ-bench.owl" );
    model.read(in, null);
    OutputStream os = new FileOutputStream("./univ-bench.nt");
    model.write(os, "N-TRIPLE");
    os.close();
 
    String queryString =
      " SELECT ?subject ?prop ?object WHERE { ?subject ?prop ?object } ";
 
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
 
    try {
      int iTriplesCount = 0;
      ResultSet results = qexec.execSelect() ;
      for ( ; results.hasNext() ; ) {
        QuerySolution soln = results.nextSolution() ;
        iTriplesCount++;
      }
      System.out.println("Asserted  triples count: " + iTriplesCount);
    } 
    finally { 
      qexec.close() ; 
    }
    
    Attachment attachment = Attachment.createInstance(
        new String[] {}, "OWLPRIME",
        InferenceMaintenanceMode.NO_UPDATE, QueryOptions.DEFAULT);
 
    GraphOracleSem graph = new GraphOracleSem(oracle, szModelName, attachment);
    graph.analyze();
    graph.performInference();
 
    query = QueryFactory.create(queryString) ;
    qexec = QueryExecutionFactory.create(query,new ModelOracleSem(graph)) ;
 
    try {
      int iTriplesCount = 0;
      ResultSet results = qexec.execSelect() ;
      for ( ; results.hasNext() ; ) {
        QuerySolution soln = results.nextSolution() ;
        iTriplesCount++;
      }
      System.out.println("Asserted + Infered triples count: " + iTriplesCount);
    } 
    finally { 
      qexec.close() ; 
    }
    model.close();    
 
    OracleUtils.dropSemanticModel(oracle, szModelName);    
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-16, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test6.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test6 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
Asserted  triples count: 294
Asserted + Infered triples count: 341

Note that this output reflects an older version of the LUBM ontology. The latest version of the ontology has more triples.

7.15.3 Test7.java: Bulk Load OWL Ontology and Perform OWLPrime inference

Example 7-17 loads the same OWL ontology as in Section 7.15.2, but stored in a local file using Bulk Loader. Ontologies can also be loaded using an incremental and batch loader; these two methods are also listed in the example for completeness.

Example 7-17 Bulk Load OWL Ontology and Perform OWLPrime inference

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.rdf.model.*;
import com.hp.hpl.jena.util.*;
import oracle.spatial.rdf.client.jena.*;
 
public class Test7 
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    // in memory Jena Model
    Model model = ModelFactory.createDefaultModel();
    InputStream is = FileManager.get().open("./univ-bench.owl");
    model.read(is, "", "RDF/XML");
    is.close();
 
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem modelDest = ModelOracleSem.createOracleSemModel(oracle, 
szModelName);
 
    GraphOracleSem g = modelDest.getGraph();
    g.dropApplicationTableIndex();
 
    int method = 2; // try bulk loader
    String tbs = "SYSAUX"; // can be customized
    if (method == 0) {
      System.out.println("start incremental");
      modelDest.add(model);
      System.out.println("end size " + modelDest.size());
    }
    else if (method == 1) {
      System.out.println("start batch load");
      g.getBulkUpdateHandler().addInBatch(
          GraphUtil.findAll(model.getGraph()), tbs);
      System.out.println("end size " + modelDest.size());
    }
    else {
      System.out.println("start bulk load");
      g.getBulkUpdateHandler().addInBulk(
          GraphUtil.findAll(model.getGraph()), tbs);
      System.out.println("end size " + modelDest.size());
    }
    g.rebuildApplicationTableIndex();
 
    long lCount = g.getCount(Triple.ANY);
    System.out.println("Asserted  triples count: " + lCount);
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-17, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test7.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test7 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
start bulk load
end size 293
Asserted  triples count: 293

Note that this output reflects an older version of the LUBM ontology. The latest version of the ontology has more triples.

7.15.4 Test8.java: SPARQL OPTIONAL Query

Example 7-18 shows a SPARQL OPTIONAL query. It inserts triples that postulate the following:

  • John is a parent of Mary.

  • John is a parent of Jack.

  • Mary is a parent of Jill.

It then finds parent-child relationships, optionally including any grandchild (gkid) relationships.

Example 7-18 SPARQL OPTIONAL Query

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
 
public class Test8 
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, 
szModelName);
    GraphOracleSem g = model.getGraph();
 
    g.add(Triple.create(
          Node.createURI("u:John"), Node.createURI("u:parentOf"), 
Node.createURI("u:Mary")));
    g.add(Triple.create(
          Node.createURI("u:John"), Node.createURI("u:parentOf"), 
Node.createURI("u:Jack")));
    g.add(Triple.create(
          Node.createURI("u:Mary"), Node.createURI("u:parentOf"), 
Node.createURI("u:Jill")));
        
    String queryString =
  " SELECT ?s ?o ?gkid " +
  " WHERE { ?s <u:parentOf> ?o . OPTIONAL {?o <u:parentOf> ?gkid }} ";
 
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
 
    try {
      int iMatchCount = 0;
      ResultSet results = qexec.execSelect() ;
      ResultSetFormatter.out(System.out, results, query);
    } 
    finally { 
      qexec.close() ; 
    }
    model.close();    
 
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-18, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test8.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test8 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
----------------------------------
| s        | o        | gkid     |
==================================
| <u:John> | <u:Mary> | <u:Jill> |
| <u:Mary> | <u:Jill> |          |
| <u:John> | <u:Jack> |          |
---------------------------------- 

7.15.5 Test9.java: SPARQL Query with LIMIT and OFFSET

Example 7-19 shows a SPARQL query with LIMIT and OFFSET. It inserts triples that postulate the following:

  • John is a parent of Mary.

  • John is a parent of Jack.

  • Mary is a parent of Jill.

It then finds one parent-child relationship (LIMIT 1), skipping the first two parent-child relationships encountered (OFFSET 2), and optionally includes any grandchild (gkid) relationships for the one found.

Example 7-19 SPARQL Query with LIMIT and OFFSET

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
public class Test9 
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, 
szModelName);
    GraphOracleSem g = model.getGraph();
 
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"),
                    Node.createURI("u:Mary")));
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"),
                    Node.createURI("u:Jack")));
    g.add(Triple.create(Node.createURI("u:Mary"), Node.createURI("u:parentOf"),              
                    Node.createURI("u:Jill")));
        
    String queryString =
      " SELECT ?s ?o ?gkid " +
      " WHERE { ?s <u:parentOf> ?o . OPTIONAL {?o <u:parentOf> ?gkid }} " +
      " LIMIT 1 OFFSET 2";
 
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
 
    int iMatchCount = 0;
    ResultSet results = qexec.execSelect() ;
    ResultSetFormatter.out(System.out, results, query);
    qexec.close() ; 
    model.close();    
 
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-19, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test9.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test9 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
------------------------------
| s        | o        | gkid |
==============================
| <u:John> | <u:Jack> |      |
------------------------------ 

7.15.6 Test10.java: SPARQL Query with TIMEOUT and DOP

Example 7-20 shows the SPARQL query from Section 7.15.5 with additional features, including a timeout setting (TIMEOUT=1, in seconds) and parallel execution setting (DOP=4).

Example 7-20 SPARQL Query with TIMEOUT and DOP

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
 
public class Test10 {
  public static void main(String[] args) throws Exception  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, szModelName);
    GraphOracleSem g = model.getGraph();
 
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
                            Node.createURI("u:Mary")));
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
                        Node.createURI("u:Jack")));
    g.add(Triple.create(Node.createURI("u:Mary"), Node.createURI("u:parentOf"), 
                        Node.createURI("u:Jill")));
    String queryString =
        " PREFIX ORACLE_SEM_FS_NS: <http://oracle.com/semtech#dop=4,timeout=1> " 
      + " SELECT ?s ?o ?gkid WHERE { ?s <u:parentOf> ?o . " 
      + " OPTIONAL {?o <u:parentOf> ?gkid }} "
      + " LIMIT 1 OFFSET 2";
 
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
 
    int iMatchCount = 0;
    ResultSet results = qexec.execSelect() ;
    ResultSetFormatter.out(System.out, results, query);
    qexec.close() ; 
    model.close();    
 
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-20, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test10.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test10 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
------------------------------
| s        | o        | gkid |
==============================
| <u:Mary> | <u:Jill> |      |
------------------------------

7.15.7 Test11.java: Query Involving Named Graphs

Example 7-21 shows a query involving named graphs. It involves a default graph that has information about named graph URIs and their publishers. The query finds graph names, their publishers, and within each named graph finds the mailbox value using the foaf:mbox predicate.

Example 7-21 Named Graph Based Query

import java.io.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.sparql.core.*;
import com.hp.hpl.jena.query.*;
import oracle.spatial.rdf.client.jena.*;
 
public class Test11
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    GraphOracleSem graph = new GraphOracleSem(oracle, szModelName);
    DatasetGraphOracleSem dataset = DatasetGraphOracleSem.createFrom(graph);
    
    // don't need the GraphOracleSem anymore, release resources
    graph.close();
    
    // add data to the default graph
    dataset.add(new Quad(
          Quad.defaultGraphIRI, // specifies default graph
          Node.createURI("http://example.org/bob"),
          Node.createURI("http://purl.org/dc/elements/1.1/publisher"),
          Node.createLiteral("Bob Hacker")));
    dataset.add(new Quad(
          Quad.defaultGraphIRI, // specifies default graph
          Node.createURI("http://example.org/alice"),
          Node.createURI("http://purl.org/dc/elements/1.1/publisher"),
          Node.createLiteral("alice Hacker")));
    
    // add data to the bob named graph
    dataset.add(new Quad(
          Node.createURI("http://example.org/bob"), // graph name
          Node.createURI("urn:bob"),
          Node.createURI("http://xmlns.com/foaf/0.1/name"),
          Node.createLiteral("Bob")));
    dataset.add(new Quad(
          Node.createURI("http://example.org/bob"), // graph name
          Node.createURI("urn:bob"),
          Node.createURI("http://xmlns.com/foaf/0.1/mbox"),
          Node.createURI("mailto:bob@example")));
    
    // add data to the alice named graph
    dataset.add(new Quad(
          Node.createURI("http://example.org/alice"), // graph name
          Node.createURI("urn:alice"),
          Node.createURI("http://xmlns.com/foaf/0.1/name"),
          Node.createLiteral("Alice")));
    dataset.add(new Quad(
          Node.createURI("http://example.org/alice"), // graph name
          Node.createURI("urn:alice"),
          Node.createURI("http://xmlns.com/foaf/0.1/mbox"),
          Node.createURI("mailto:alice@example")));
    
    DataSource ds = DatasetFactory.create(dataset);
    
    String queryString =  
          " PREFIX foaf: <http://xmlns.com/foaf/0.1/> "
        + " PREFIX dc: <http://purl.org/dc/elements/1.1/> "
        + " SELECT ?who ?graph ?mbox "
        + " FROM NAMED <http://example.org/alice> "
        + " FROM NAMED <http://example.org/bob> "
        + " WHERE "
        + " { " 
        + "    ?graph dc:publisher ?who . "
        + "    GRAPH ?graph { ?x foaf:mbox ?mbox } "
        + " } ";
    
    Query query = QueryFactory.create(queryString);
    QueryExecution qexec = QueryExecutionFactory.create(query, ds);
    
    ResultSet results = qexec.execSelect();
    ResultSetFormatter.out(System.out, results, query);
    
    qexec.close();
    dataset.close();
    
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-21, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test11.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test11 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1

7.15.8 Test12.java: SPARQL ASK Query

Example 7-22 shows a SPARQL ASK query. It inserts a triple that postulates that John is a parent of Mary. It then finds whether John is a parent of Mary.

Example 7-22 SPARQL ASK Query

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
public class Test12
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, 
          szModelName);
    GraphOracleSem g = model.getGraph();
 
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
                        Node.createURI("u:Mary")));
    String queryString = " ASK { <u:John> <u:parentOf> <u:Mary> } ";
 
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
    boolean b = qexec.execAsk();
    System.out.println("ask result = " + ((b)?"TRUE":"FALSE"));
    qexec.close() ; 
    
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-22, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test12.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test12 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
ask result = TRUE

7.15.9 Test13.java: SPARQL DESCRIBE Query

Example 7-23 shows a SPARQL DESCRIBE query. It inserts triples that postulate the following:

  • John is a parent of Mary.

  • John is a parent of Jack.

  • Amy is a parent of Jack.

It then finds all relationships that involve any parents of Jack.

Example 7-23 SPARQL DESCRIBE Query

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
 
public class Test13
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, szModelName);
    GraphOracleSem g = model.getGraph();
 
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
                    Node.createURI("u:Mary")));
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
 Node.createURI("u:Jack")));
    g.add(Triple.create(Node.createURI("u:Amy"), Node.createURI("u:parentOf"), 
 Node.createURI("u:Jack")));
    String queryString = " DESCRIBE ?x WHERE {?x <u:parentOf> <u:Jack>}";
 
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
    Model m = qexec.execDescribe();
    System.out.println("describe result = " + m.toString());
 
    qexec.close() ; 
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-23, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test13.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test13 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
describe result = <ModelCom   {u:Amy @u:parentOf u:Jack; u:John @u:parentOf u:Jack; u:John @u:parentOf u:Mary} | >

7.15.10 Test14.java: SPARQL CONSTRUCT Query

Example 7-24 shows a SPARQL CONSTRUCT query. It inserts triples that postulate the following:

  • John is a parent of Mary.

  • John is a parent of Jack.

  • Amy is a parent of Jack.

  • Each parent loves all of his or her children.

It then constructs an RDF graph with information about who loves whom.

Example 7-24 SPARQL CONSTRUCT Query

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
 
public class Test14
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, szModelName);
    GraphOracleSem g = model.getGraph();
 
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
 Node.createURI("u:Mary")));
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
 Node.createURI("u:Jack")));
    g.add(Triple.create(Node.createURI("u:Amy"), Node.createURI("u:parentOf"), 
 Node.createURI("u:Jack")));
    String queryString = " CONSTRUCT { ?s <u:loves> ?o } WHERE {?s <u:parentOf> ?o}";
 
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
    Model m = qexec.execConstruct();
    System.out.println("Construct result = " + m.toString());
 
    qexec.close() ; 
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-24, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test14.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test14 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
Construct result = <ModelCom   {u:Amy @u:loves u:Jack; u:John @u:loves u:Jack; u:John @u:loves u:Mary} | >

7.15.11 Test15.java: Query Multiple Models and Specify "Allow Duplicates"

Example 7-25 queries multiple models and uses the "allow duplicates" option. It inserts triples that postulate the following:

  • John is a parent of Jack (in Model 1).

  • Mary is a parent of Jack (in Model 2).

  • Each parent loves all of his or her children.

It then finds out who loves whom. It searches both models and allows for the possibility of duplicate triples in the models (although there are no duplicates in this example).

Example 7-25 Query Multiple Models and Specify "Allow Duplicates"

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
 
public class Test15
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName1 = args[3];
    String szModelName2 = args[4];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model1 = ModelOracleSem.createOracleSemModel(oracle, szModelName1);
    model1.getGraph().add(Triple.create(Node.createURI("u:John"), 
                     Node.createURI("u:parentOf"), Node.createURI("u:Jack")));
    model1.close();
 
    ModelOracleSem model2 = ModelOracleSem.createOracleSemModel(oracle, szModelName2);
    model2.getGraph().add(Triple.create(Node.createURI("u:Mary"), 
                     Node.createURI("u:parentOf"), Node.createURI("u:Jack")));
    model2.close();
 
    String[] modelNamesList = {szModelName2};
    String[] rulebasesList  = {};
    Attachment attachment = Attachment.createInstance(modelNamesList, rulebasesList, 
              InferenceMaintenanceMode.NO_UPDATE,
              QueryOptions.ALLOW_QUERY_VALID_AND_DUP);
 
    GraphOracleSem graph = new GraphOracleSem(oracle, szModelName1, attachment);
    ModelOracleSem model = new ModelOracleSem(graph);
 
    String queryString = " CONSTRUCT { ?s <u:loves> ?o } WHERE {?s <u:parentOf> ?o}";
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
    Model m = qexec.execConstruct();
    System.out.println("Construct result = " + m.toString());
 
    qexec.close() ; 
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName1);
    OracleUtils.dropSemanticModel(oracle, szModelName2);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-25, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test15.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test15 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1 M2
Construct result = <ModelCom   {u:Mary @u:loves u:Jack; u:John @u:loves u:Jack} | >

7.15.12 Test16.java: SPARQL Update

Example 7-26 inserts two triples into a model.

Example 7-26 SPARQL Update

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import com.hp.hpl.jena.util.iterator.*;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.update.*;
 
public class Test16
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, szModelName);
    GraphOracleSem g = model.getGraph();
    String insertString =  
      " PREFIX dc: <http://purl.org/dc/elements/1.1/> "         + 
      " INSERT DATA "                                           +
      " { <http://example/book3> dc:title    \"A new book\" ; " +
      "                         dc:creator  \"A.N.Other\" . "   + 
      " }   ";
 
    UpdateAction.parseExecute(insertString,  model);
    ExtendedIterator ei = GraphUtil.findAll(g);
    while (ei.hasNext()) {
      System.out.println("Triple " + ei.next().toString());
    }
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-26, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test16.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test16 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
Triple http://example/book3 @dc:title "A new book"
Triple http://example/book3 @dc:creator "A.N.Other"

7.15.13 Test17.java: SPARQL Query with ARQ Built-In Functions

Example 7-27 inserts data about two books, and it displays the book titles in all uppercase characters and the length of each title string.

Example 7-27 SPARQL Query with ARQ Built-In Functions

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import com.hp.hpl.jena.util.iterator.*;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.update.*;
 
public class Test17 {
  public static void main(String[] args) throws Exception  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, szModelName);
    GraphOracleSem g = model.getGraph();
    String insertString =  
      " PREFIX dc: <http://purl.org/dc/elements/1.1/> "         + 
      " INSERT DATA "                                           +
      " { <http://example/book3> dc:title    \"A new book\" ; " +
      "                         dc:creator  \"A.N.Other\" . "   + 
      "   <http://example/book4> dc:title    \"Semantic Web Rocks\" ; " +
      "                         dc:creator  \"TB\" . "   + 
      " }   ";
 
    UpdateAction.parseExecute(insertString,  model);
    String queryString = "PREFIX  dc:   <http://purl.org/dc/elements/1.1/> " +
      " PREFIX  fn: <http://www.w3.org/2005/xpath-functions#> " + 
      " SELECT ?subject (fn:upper-case(?object) as ?object1)  " + 
      "                 (fn:string-length(?object) as ?strlen) " + 
      " WHERE { ?subject dc:title ?object } " 
      ;
    Query query = QueryFactory.create(queryString, Syntax.syntaxARQ);
    QueryExecution qexec = QueryExecutionFactory.create(query, model);
    ResultSet results = qexec.execSelect();
    ResultSetFormatter.out(System.out, results, query);
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-27, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test17.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test17 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
----------------------------------------------------------
| subject                | object1              | strlen |
==========================================================
| <http://example/book3> | "A NEW BOOK"         | 10     |
| <http://example/book4> | "SEMANTIC WEB ROCKS" | 18     |
----------------------------------------------------------

7.15.14 Test18.java: SELECT Cast Query

Example 7-28 "converts" two Fahrenheit temperatures (18.1 and 32.0) to Celsius temperatures.

Example 7-28 SELECT Cast Query

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import com.hp.hpl.jena.util.iterator.*;
import oracle.spatial.rdf.client.jena.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.update.*;
 
public class Test18 {
  public static void main(String[] args) throws Exception  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
    
    Oracle oracle = new Oracle(szJdbcURL, szUser, szPasswd);
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, 
szModelName);
    GraphOracleSem g = model.getGraph();
    String insertString =  
      " PREFIX xsd: <http://www.w3.org/2001/XMLSchema#> " +
      " INSERT DATA "                                     +
      " { <u:Object1> <u:temp>    \"18.1\"^^xsd:float ; " +
      "               <u:name>    \"Foo... \" . "         + 
      "   <u:Object2> <u:temp>    \"32.0\"^^xsd:float ; " +
      "               <u:name>    \"Bar... \" . "         + 
      " }   ";
 
    UpdateAction.parseExecute(insertString,  model);
    String queryString = 
      " PREFIX  fn: <http://www.w3.org/2005/xpath-functions#> " + 
      " SELECT ?subject ((?temp - 32.0)*5/9 as ?celsius_temp) " +
      "WHERE { ?subject <u:temp> ?temp } " 
      ;
    Query query = QueryFactory.create(queryString, Syntax.syntaxARQ);
    QueryExecution qexec = QueryExecutionFactory.create(query, model);
    ResultSet results = qexec.execSelect();
    ResultSetFormatter.out(System.out, results, query);
 
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-28, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test18.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test18 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
------------------------------------------------------------------------
| subject     | celsius_temp                                           |
========================================================================
| <u:Object1> | "-7.7222223"^^<http://www.w3.org/2001/XMLSchema#float> |
| <u:Object2> | "0.0"^^<http://www.w3.org/2001/XMLSchema#float>        |
------------------------------------------------------------------------

7.15.15 Test19.java: Instantiate Oracle Database Using OracleConnection

Example 7-29 shows a different way to instantiate an Oracle object using a given OracleConnection object. (In a J2EE Web application, users can normally get an OracleConnection object from a J2EE data source.)

Example 7-29 Instantiate Oracle Database Using OracleConnection

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import com.hp.hpl.jena.util.iterator.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.update.*;
import oracle.spatial.rdf.client.jena.*;
import oracle.jdbc.pool.*;
import oracle.jdbc.*;
        
public class Test19 {
  public static void main(String[] args) throws Exception {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
 
    OracleDataSource ds = new OracleDataSource();
    ds.setURL(szJdbcURL);
    ds.setUser(szUser);
    ds.setPassword(szPasswd);
    OracleConnection conn = (OracleConnection) ds.getConnection();
    Oracle oracle = new Oracle(conn);
 
    ModelOracleSem model = ModelOracleSem.createOracleSemModel(oracle, 
szModelName);
    GraphOracleSem g = model.getGraph();
 
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
                        Node.createURI("u:Mary")));
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"),   
                        Node.createURI("u:Jack")));
    g.add(Triple.create(Node.createURI("u:Mary"), Node.createURI("u:parentOf"), 
         Node.createURI("u:Jill")));
    String queryString =
       " SELECT ?s ?o  WHERE { ?s <u:parentOf> ?o .} ";
    Query query = QueryFactory.create(queryString) ;
    QueryExecution qexec = QueryExecutionFactory.create(query, model) ;
 
    ResultSet results = qexec.execSelect() ;
    ResultSetFormatter.out(System.out, results, query);
    qexec.close() ; 
    model.close();    
    OracleUtils.dropSemanticModel(oracle, szModelName);
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-29, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test19.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test19 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
-----------------------
| s        | o        |
=======================
| <u:John> | <u:Mary> |
| <u:John> | <u:Jack> |
| <u:Mary> | <u:Jill> |
-----------------------

7.15.16 Test20.java: Oracle Database Connection Pooling

Example 7-30 uses Oracle Database connection pooling.

Example 7-30 Oracle Database Connection Pooling

import java.io.*;
import com.hp.hpl.jena.query.*;
import com.hp.hpl.jena.rdf.model.Model;
import com.hp.hpl.jena.util.FileManager;
import com.hp.hpl.jena.util.iterator.*;
import com.hp.hpl.jena.graph.*;
import com.hp.hpl.jena.update.*;
import oracle.spatial.rdf.client.jena.*;
import oracle.jdbc.pool.*;
import oracle.jdbc.*;
 
public class Test20
{
  public static void main(String[] args) throws Exception
  {
    String szJdbcURL = args[0];
    String szUser    = args[1];
    String szPasswd  = args[2];
    String szModelName = args[3];
 
    // test with connection properties (taken from some example)
    java.util.Properties prop = new java.util.Properties();
    prop.setProperty("MinLimit", "2");     // the cache size is 2 at least 
    prop.setProperty("MaxLimit", "10");
    prop.setProperty("InitialLimit", "2"); // create 2 connections at startup
    prop.setProperty("InactivityTimeout", "1800");    //  seconds
    prop.setProperty("AbandonedConnectionTimeout", "900");  //  seconds
    prop.setProperty("MaxStatementsLimit", "10");
    prop.setProperty("PropertyCheckInterval", "60"); // seconds
 
    System.out.println("Creating OraclePool");
    OraclePool op = new OraclePool(szJdbcURL, szUser, szPasswd, prop, 
               "OracleSemConnPool");
    System.out.println("Done creating OraclePool");
 
    // grab an Oracle and do something with it
    System.out.println("Getting an Oracle from OraclePool");
    Oracle oracle = op.getOracle();
    System.out.println("Done");
    System.out.println("Is logical connection:" +
        oracle.getConnection().isLogicalConnection());
    GraphOracleSem g = new GraphOracleSem(oracle, szModelName);
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
                        Node.createURI("u:Mary")));
    g.close();
    // return the Oracle back to the pool
    oracle.dispose();
    
    // grab another Oracle and do something else 
    System.out.println("Getting an Oracle from OraclePool");
    oracle = op.getOracle();
    System.out.println("Done");
    System.out.println("Is logical connection:" +
        oracle.getConnection().isLogicalConnection());
    g = new GraphOracleSem(oracle, szModelName);
    g.add(Triple.create(Node.createURI("u:John"), Node.createURI("u:parentOf"), 
                        Node.createURI("u:Jack")));
    g.close();
    
    OracleUtils.dropSemanticModel(oracle, szModelName); 
    
    // return the Oracle back to the pool
    oracle.dispose();
  }
}

The following are the commands to compile and run Example 7-30, as well as the expected output of the java command.

javac -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:/slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar Test20.java
java -classpath ./:./jena-2.6.4.jar:./sdordfclient.jar:./ojdbc6.jar:./slf4j-api-1.5.8.jar:./slf4j-log4j12-1.5.8.jar:./arq-2.8.8.jar:./xercesImpl-2.7.1.jar:./iri-0.8.jar:./icu4j-3.4.4.jar:./log4j-1.2.14.jar  Test20 jdbc:oracle:thin:@localhost:1521:orcl scott <password-for-scott> M1
Creating OraclePool
Done creating OraclePool
Getting an Oracle from OraclePool
Done
Is logical connection:true
Getting an Oracle from OraclePool
Done
Is logical connection:true

7.16 SPARQL Gateway and Semantic Data

SPARQL Gateway is a J2EE web application that is included with the Jena Adapter for Oracle Database. It is designed to make semantic data (RDF/OWL/SKOS) easily available to applications that operate on relational and XML data, including Oracle Business Intelligence Enterprise Edition (OBIEE) 11g.

This section includes the following major topics:

7.16.1 SPARQL Gateway Features and Benefits Overview

SPQRQL Gateway handles several challenges in exposing semantic data to a non-semantic application:

  • RDF syntax, SPARQL query syntax and SPARQL protocol must be understood.

  • The SPARQL query response syntax must be understood.

  • A transformation must convert a SPARQL query response to something that the application can consume.

To address these challenges, SPARQL Gateway manages SPARQL queries and XSLT operations, executes SPARQL queries against any arbitrary standard-compliant SPARQL endpoints, and performs necessary XSL transformations before passing the response back to applications. Applications can then consume semantic data as if it is coming from an existing data source.

Different triple stores or quad stores often have different capabilities. For example, the SPARQL endpoint supported by Oracle Database, with the Jena Adapter and Joseki, allows parallel execution, query timeout, dynamic sampling, result cache, and other features, in addition to the core function of parsing and answering a given standard-compliant SPARQL query. However, these features may not be available from another given semantic data store.

With the Oracle Semantic Technologies SPARQL Gateway, you get certain highly desirable capabilities, such as the ability to set a timeout on a long running query and the ability to get partial results from a complex query in a given amount of time. Waiting indefinitely for a query to finish is a challenge for end users, as is an application with a response time constraint. SPARQL Gateway provides both timeout and best effort query functions on top of a SPARQL endpoint. This effectively removes some uncertainty from consuming semantic data through SPARQL query executions. (See Section 7.16.3.2, "Specifying a Timeout Value" and Section 7.16.3.3, "Specifying Best Effort Query Execution".)

7.16.2 Installing and Configuring SPARQL Gateway

To install and configure SPARQL Gateway, follow these major steps, which are explained in subsections that follow:

  1. Download the Jena Adapter .zip File (if Not Already Done)

  2. Deploy SPARQL Gateway in WebLogic Server

  3. Modify Proxy Settings, if Necessary

  4. Configure the OracleSGDS Data Source, if Necessary

  5. Add and Configure the SparqlGatewayAdminGroup Group, if Desired

7.16.2.1 Download the Jena Adapter .zip File (if Not Already Done)

If you have not already done so, download the Jena Adapter (jena_adaptor_for_release11.2.0.3.zip) from the Oracle Database Semantic Technologies page and unzip it into a temporary directory, as explained in Section 7.1.

Note that the SPARQL Gateway Java class implementations are embedded in sdordfclient.jar (see Section 7.16.5, "Using the SPARQL Gateway Java API").

7.16.2.2 Deploy SPARQL Gateway in WebLogic Server

Deploy SPARQL Gateway in Oracle WebLogic Server, as follows:

  1. Create a temporary directory, for example, /tmp/sg.

  2. From the directory into which you unzipped the SPARQL Gateway file that you downloaded, copy the sparqlgateway directory and everything under it (including subdirectories) into the temporary directory you created in step 1. For example:

    cd /tmp
    cp -rf  sparqlgateway/*  /tmp/sg
    

    (This sample directory structure is assumed in some later steps.)

  3. If you do not already have the ARQ 2.8.8 package, download it from http://sourceforge.net/projects/jena/files/ARQ/ARQ-2.8.8/arq-2.8.8.zip/download.

  4. Copy the following libraries from the ARQ 2.8.8 package to /tmp/sg/WEB-INF/lib:

    arq-2.8.8.jar
    icu4j-3.4.4.jar
    iri-0.8.jar
    jena-2.6.4.jar
    junit-4.5.jar
    log4j-1.2.14.jar
    lucene-core-2.3.1.jar
    slf4j-api-1.5.8.jar
    slf4j-log4j12-1.5.8.jar
    stax-api-1.0.1.jar
    wstx-asl-3.2.9.jar
    xercesImpl-2.7.1.jar
    
  5. Copy the following libraries to /tmp/sg/WEB-INF/lib:

    ojdbc6.jar  (from $ORACLE_HOME/jdbc/lib/ojdbc6.jar )
    sdordfclient.jar  (from jena_adaptor_for_release11.2.0.3.zip under the jar/ directory)
    sdordf.jar  from $ORACLE_HOME/md/jlib/sdordf.jar )
    
  6. Customize the /tmp/sg/WEB-INF/web.xml file as needed. Be sure to specify appropriate values for the sparql_gateway_repository_filedir and sparql_gateway_repository_url parameters.

  7. Add XSLT files or SPARQL query files to the /tmp/sg directory, if necessary.

    The following files are provided by Oracle in that directory: default.xslt, noop.xslt, and qb1.sparql. The default.xslt file is intended mainly for transforming SPARQL query responses (XML) to a format acceptable to Oracle Business Intelligence Enterprise Edition (OBIEE).

    (These files are described in Section 7.16.3.1, "Storing SPARQL Queries and XSL Transformations"; using SPARQL Gateway with OBIEE is explained in Section 7.16.7, "Using SPARQL Gateway as an XML Data Source to OBIEE".)

  8. Go to the autodeploy directory of WebLogic Server, create directory sparqlgateway.war, and copy files, as follows. (For information about auto-deploying applications in development domains, see: http://download.oracle.com/docs/cd/E11035_01/wls100/deployment/autodeploy.html)

    cd <domain_name>/autodeploy
    mkdir sparqlgateway.war  
    cp -rf  /tmp/sg/*  <domain_name>/autodeploy/sparqgateway.war
      
    

    In this example, <domain_name> is the name of a WebLogic Server domain.

  9. If you want to build a .war file from the /tmp/sg directory, enter the following commands:

    cd /tmp/sg
    jar cvf /tmp/sparqlgateway.war *
    
  10. Start or restart WebLogic Server.

  11. Verify your deployment by using your Web browser to connect to a URL in the following format (assume that the Web application is deployed at port 7001):

    http://<hostname>:7001/sparqlgateway
    

7.16.2.3 Modify Proxy Settings, if Necessary

If your SPARQL Gateway is behind a firewall and you want SPARQL Gateway to communicate with SPARQL endpoints on the Internet as well as those inside the firewall, you probably need to use the following JVM settings:

-Dhttp.proxyHost=<your_proxy_host>
-Dhttp.proxyPort=<your_proxy_port>
-Dhttp.nonProxyHosts=127.0.0.1|<hostname_1_for_sparql_endpoint_inside_firewall>|<hostname_2_for_sparql_endpoint_inside_firewall>|...|<hostname_n_for_sparql_endpoint_inside_firewall>

You can specify these settings in the startWebLogic.sh script.

7.16.2.4 Configure the OracleSGDS Data Source, if Necessary

If an Oracle database is used for storage of and access to SPARQL queries and XSL transformations for SPARQL Gateway, then you must configure a data source named OracleSGDS.

To create this data source, follow the instructions in Section 7.2.1, "Creating the Required Data Source Using WebLogic Server"; however, specify OracleSGDS as the data source name instead of OracleSemDS.

If the OracleSGDS data source is configured and available, SPARQL Gateway servlet will automatically create all the necessary tables and indexes upon initialization.

7.16.2.5 Add and Configure the SparqlGatewayAdminGroup Group, if Desired

The following JSP files in SPARQL Gateway can help you to view, edit, and update SPARQL queries and XSL transformations that are stored in an Oracle database:

http://<host>:7001/sparqlgateway/admin/sparql.jsp
http://<host>:7001/sparqlgateway/admin/xslt.jsp

These files are protected by HTTP Basic Authentication. In WEB-INF/weblogic.xml, a principal named SparqlGatewayAdminGroup is defined.

To be able to log in to either of these JSP pages, you must use the WebLogic Server to add a group named SparqlGatewayAdminGroup, and create a new user or assign an existing user to this group.

7.16.3 Using SPARQL Gateway with Semantic Data

The primary interface for an application to interact with SPARQL Gateway is through a URL with the following format:

http://host:port/sparqlgateway/sg?<SPARQL_ENDPOINT>&<SPARQL_QUERY>&<XSLT>

In the preceding format:

  • <SPARQL_ENDPOINT> specifies the ee parameter, which contains a URL encoded form of a SPARQL endpoint.

    For example, ee=http%3A%2F%2Fsparql.org%2Fbooks is the URL encoded string for SPARQL endpoint http://sparql.org/books. It means that SPARQL queries are to be executed against endpoint http://sparql.org/books.

  • <SPARQL_QUERY> specifies either the SPARQL query, or the location of the SPARQL query.

    If it is feasible for an application to accept a very long URL, you can encode the whole SPARQL query and set eq=<encoded_SPARQL_query> in the URL If it is not feasible for an application to accept a very long URL, you can store the SPARQL queries and make them available to SPARQL Gateway using one of the approaches described in Section 7.16.3.1.

  • <XSLT> specifies either the XSL transformation, or the location of the XSL transformation.

    If it is feasible for an application to accept a very long URL, you can encode the whole XSL transformation and set ex=<encoded_XSLT> in the URL If it is not feasible for an application to accept a very long URL, you can store the XSL transformations and make them available to SPARQL Gateway using one of the approaches described in Section 7.16.3.1.

Related topics:

7.16.3.1 Storing SPARQL Queries and XSL Transformations

If it is not feasible for an application to accept a very long URL, you can specify the location of the SPARQL query and the XSL transformation in the <SPARQL_QUERY> and <XSLT> portions of the URL format described in Section 7.16.3, using any of the following approaches:

  • Store the SPARQL queries and XSL transformations in the SPARQL Gateway Web application itself.

    To do this, unpack the sparqlgateway.war file, and store the SPARQL queries and XSL transformations in the top-level directory; then pack the sparqlgateway.war file and redeploy it.

    The sparqlgateway.war file includes the following example files: qb1.sparql (SPARQL query) and default.xslt (XSL transformation).

    Tip:

    Use the file extension .sparql for SPARQL query files, and the file extension .xslt for XSL transformation files.

    The syntax for specifying these files (using the provided example file names) is wq=qb1.sparql for a SPARQL query file and wx=default.xslt for an XSL transformation file.

    If you want to customize the default XSL transformations, see the examples in Section 7.16.4, "Customizing the Default XSLT File".

    If you specify wx=noop.xslt, XSL transformation is not performed and the SPARQL response is returned "as is" to the client.

  • Store the SPARQL queries and XSL transformations in a file system directory, and make sure that the directory is accessible for the deployed SPARQL Gateway Web application.

    By default, the directory is set to /tmp, as shown in the following <init-param> setting:

    <init-param>
       <param-name>sparql_gateway_repository_filedir</param-name>
       <param-value>/tmp/</param-value>
    </init-param>
    

    It is recommended that you customize this directory before deploying the SPARQL Gateway. To change the directory setting, edit the text in between the <param-value> and </param-value> tags.

    The follwoing examples specify a SPARQL query file and an XSL transflrmation file that are in the directory specified in the <init-param> element for sparql_gateway_repository_filedir:

    fq=qb1.sparql
    fx=myxslt1.xslt
    
  • Make the SPARQL queries and XSL transformations accessible from a website.

    By default, the website directory is set to http://127.0.0.1/queries/, as shown in the following <init-param> setting:

    <init-param>
       <param-name>sparql_gateway_repository_url</param-name>
       <param-value>http://127.0.0.1/queries/</param-value>
    </init-param>
    

    Customize this directory before deploying the SPARQL Gateway. To change the website setting, edit the text in between the <param-value> and </param-value> tags.

    The follwoing example specifies a SPARQL query file and an XSL transflrmation file that are in the URL specified in the <init-param> element for sparql_gateway_repository_url.

    uq=qb1.sparql
    ux=myxslt1.xslt
    

    Internally, SPARQL Gateway computes the appropriate complete URL, fetches the content, starts query execution, and applies the XSL transformation to the query response XML.

  • Store the SPARQL queries and XSL transformations in an Oracle database.

    This approach requires that the J2EE data source OracleSGDS be defined. After SPARQL Gateway retrieves a database connection from the OracleSGDS data source, a SPARQL query is read from the database table ORACLE_ORARDF_SG_QUERY using the integer ID provided.

    The syntax for fetching a SPARQL query from an Oracle database is dq=<integer-id>, and the syntax for fetching an XSL transformation from an Oracle database is dx=<integer-id>.

    Upon servlet initialization, the following tables are created automatically if they do not already exist (you do not need to create them manually):

    • ORACLE_ORARDF_SG_QUERY with a primary key of QID (integer type)

    • ORACLE_ORARDF_SG_XSLT with a primary key of XID (integer type)

7.16.3.2 Specifying a Timeout Value

When you submit a potentially long-running query using the URL format described in Section 7.16.3, you can limit the execution time by specifying a timeout value in milliseconds. For example, the following shows the URL format and a timeout specification that the SPARQL query execution started from SPARQL Gateway is to be ended after 1000 milliseconds (1 second):

http://host:port/sparqlgateway/sg?<SPARQL_ENDPOINT>&<SPARQL_QUERY>&<XSLT>&t=1000

If a query does not finish when timeout occurs, then an empty SPARQL response is constructed by SPARQL Gateway.

Note that even if SPARQL Gateway times out a query execution at the HTTP connection level, the query may still be running on the server side. The actual behavior will be vendor-dependent.

7.16.3.3 Specifying Best Effort Query Execution

Note:

You can specify best effort query execution only if you also specify a timeout value (described in Section 7.16.3.2).

When you submit a potentially long-running query using the URL format described in Section 7.16.3, if you specify a timeout value, you can also specify a "best effort" limitation on the query. For example, the following shows the URL format with a timeout specification of 1000 milliseconds (1 second) and a best effort specification (&b=t):

http://host:port/sparqlgateway/sg?<SPARQL_ENDPOINT>&<SPARQL_QUERY>&<XSLT>&t=1000&b=t

The web.xml file includes two parameter settings that affect the behavior of the best effort option: sparql_gateway_besteffort_maxrounds and sparql_gateway_besteffort_maxthreads. The following show the default definitions:

<init-param>
  <param-name>sparql_gateway_besteffort_maxrounds</param-name>
  <param-value>10</param-value>
</init-param>
 
<init-param>
  <param-name>sparql_gateway_besteffort_maxthreads</param-name>
  <param-value>3</param-value>
</init-param>

When a SPARQL SELECT query is executed in best effort style, a series of queries will be executed with an increasing LIMIT value setting in the SPARQL query body. (The core idea is based on the observation that a SPARQL query runs faster with a smaller LIMIT setting.) SPARQL Gateway starts query execution with a "LIMIT 1" setting. Ideally, this query can finish before the timeout is due. Assume that is the case, the next query will have its LIMIT setting is increased, and subsequent queries have higher limits. The maximum number of query executions is controlled by the sparql_gateway_besteffort_maxrounds parameter.

If it is possible to run the series of queries in parallel, the sparql_gateway_besteffort_maxthreads parameter controls the degree of parallelism.

7.16.3.4 Specifying a Content Type Other Than text/xml

By default, SPARQL Gateway assumes that XSL transformations generate XML, and so the default content type set for HTTP response is text/xml. However, if your application requires a response format other than XML, you can specify the format in an additional URL parameter (with syntax &rt=), using the following format:

http://host:port/sparqlgateway/sg?<SPARQL_ENDPOINT>&<SPARQL_QUERY>&<XSLT>&rt=<content_type>

Note that <content_type> must be URL encoded.

7.16.4 Customizing the Default XSLT File

You can customize the default XSL transformation file (the one referenced using wx=default.xslt). This section presents some examples of customizations.

The following example implements this namespace prefix replacement logic: if a variable binding returns a URI that starts with http://purl.org/goodrelations/v1#, that portion is replaced by gr:; and if a variable binding returns a URI that starts with http://www.w3.org/2000/01/rdf-schema#, that portion is replaced by rdfs:.

<xsl:when test="starts-with(text(),'http://purl.org/goodrelations/v1#')">
   <xsl:value-of select="concat('gr:',substring-after(text(),'http://purl.org/goodrelations/v1#'))"/>
</xsl:when>
...
<xsl:when test="starts-with(text(),'http://www.w3.org/2000/01/rdf-schema#')">
   <xsl:value-of select="concat('rdfs:',substring-after(text(),'http://www.w3.org/2000/01/rdf-schema#'))"/>
</xsl:when>

The following example implements logic to trim a leading http://localhost/ or a leading http://127.0.0.1/.

<xsl:when test="starts-with(text(),'http://localhost/')">
  <xsl:value-of select="substring-after(text(),'http://localhost/')"/>
</xsl:when>
<xsl:when test="starts-with(text(),'http://127.0.0.1/')">
  <xsl:value-of select="substring-after(text(),'http://127.0.0.1/')"/>
</xsl:when>

7.16.5 Using the SPARQL Gateway Java API

In addition to a Web interface, the SPARQL Gateway administration service provides a convenient Java application programming interface (API) for managing SPARQL queries and their associated XSL transformations. The Java API is included in the Jena Adapter library, sdordfclient.jar.

Java API reference information is available in the javadoc_sparqlgateway.zip file that is included in the SPARQL Gateway .zip file (described in Section 7.16.2.1).

The main entry point for this API is the oracle.spatial.rdf.client.jena.SGDBHandler class (SPARQL Gateway Database Handler), which provides the following static methods for managing queries and transformations:

  • deleteSparqlQuery(Connection, int)

  • deleteXslt(Connection, int)

  • insertSparqlQuery(Connection, int, String, String, boolean)

  • insertXslt(Connection, int, String, String, boolean)

  • getSparqlQuery(Connection, int, StringBuilder, StringBuilder)

  • getXslt(Connection, int, StringBuilder, StringBuilder)

These methods manipulate and retrieve entries in the SPARQL Gateway associated tables that are stored in an Oracle Database instance. To use these methods, the necessary associated tables must already exist. If the tables do not exist, deploy the SPARQL Gateway on a Web server and access a URL in the following format:

http://<host>:<port>/sparqlgateway/sg?

where <host> is the host name of the Web server and <port> is the listening port of the Web server. Accessing this URL will automatically create the necessary tables if they do not already exist.

Any changes made through the Java API affect the SPARQL Gateway Web service in the same way as changes made through the administration Web interface. This provides the flexibility to manage queries and transformations using the interface you find most convenient.

Note that the insert methods provided by the Java API will not replace existing queries or transformations stored in the tables. Attempting to replace an existing query or transformation will fail. To replace a query or transformation, you must remove the existing entry in the table using one of the delete methods, and then insert the new query or transformation using one of the insert methods.

The following examples demonstrate how to perform common management tasks using the Java API. The examples assume a connection has already been established to the underlying Oracle Database instance backing the SPARQL Gateway.

Example 7-31 adds a query and an XSL transformation to the database backing the SPARQL Gateway. After the query and transformation are added, other programs can use the query and transformation through the gateway by specifying the appropriate query ID (qid) and XSL transformation ID (xid) in the request URL.

Note:

Although Example 7-31 inserts both a query and transformation, the query and transformation are not necessarily related and do not need to be used together when accessing SPARQL Gateway. Any query in the database can be used with any transformation in the database when submitting a request to SPARQL Gateway.

Example 7-31 Storing a SPARQL Query and an XSL Transformation

String query = "PREFIX ... SELECT ..."; // full SPARQL query text
String xslt  = "<?xml ...> ...";        // full XSLT transformation text
 
String queryDesc = "Conference attendee information"; // description of SPARQL query
String xsltDesc = "BIEE table widget transformation"; // description of XSLT transformation
 
int queryId = queryIdCounter++; // assign a unique ID to this query
int xsltId  = xsltIdCounter++;  // assign a unique ID to this transformation
 
// Inserting a query or transformation will fail if the table already contains
// an entry with the same ID.  Setting this boolean to true will ignore these
// exceptions (but the table will remain unchanged). Here we specify that we
// want an exception thrown if we encounter a duplicate ID.
boolean ignoreDupException = false;
 
// add the query
try {
  // Delete query if one already exists with this ID (this will not throw an
  // error if no such entry exists)
  SGDBHandler.deleteSparqlQuery( connection, queryId );
  SGDBHandler.insertSparqlQuery( connection, queryId, query, queryDesc, ignoreDupException );
} catch( SQLException sqle ) {
  // Handle exception
} catch( QueryException qe ) {
  // Handle query syntax exception
}
 
// add the XSLT
try {
  // Delete xslt if one already exists with this ID (this will not throw an
  // error if no such entry exists)
  SGDBHandler.deleteXslt( connection, xsltId );
  SGDBHandler.insertXslt( connection, xsltId, xslt, xsltDesc, ignoreDupException );
}  catch( SQLException sqle ) {
  // Handle database exception
} catch( TransformerConfigurationException tce ) {
  // Handle XSLT syntax exception
}

Example 7-32 retrieves an existing query from the database, modifies it, then stores the updated version of the query back in the database. These steps simulate editing a query and saving the changes. (Note that if the query does not exist, an exception is thrown.)

Example 7-32 Modifying a Query

StringBuilder query;
StringBuilder description;
 
// Populate these with the query text and description from the database
query = new StringBuilder( );
description = new StringBuilder( );
 
// Get the query from the database
try {
  SGDBHandler.getSparqlQuery( connection, queryId, query, description );
} catch( SQLException sqle ) {
  // Handle exception
  // NOTE: exception is thrown if query with specified ID does not exist
}
 
// The query and description should be populated now
 
// Modify the query 
String updatedQuery = query.toString( ).replaceAll("invite", "attendee");
 
// Insert the query back into the database
boolean ignoreDup = false;
try {
  // First must delete the old query
  SGDBHandler.deleteSparqlQuery( connection, queryId );
  // Now we can add
  SGDBHandler.insertSparqlQuery( connection, queryId, updatedQuery, description.toString( ), ignoreDup );
} catch( SQLException sqle ) {
  // Handle exception
} catch( QueryException qe ) {
  // Handle query syntax exception
}

Example 7-33 retrieves an existing XSL transformation and prints it to standard output. (Note that if the transformation does not exist, an exception is thrown.)

Example 7-33 Retrieving and Printing an XSL Transformation

StringBuilder xslt;
StringBuilder description;
 
// Populate these with the XSLT text and description from the database
xslt = new StringBuilder( );
description = new StringBuilder( );
 
try {
  SGDBHandler.getXslt( connection, xsltId, xslt, description );
} catch( SQLException sqle ) {
  // Handle exception
  // NOTE: exception is thrown if transformation with specified ID does not exist
}
 
// Print it to standard output
System.out.printf( "XSLT description: %s\n", description.toString( ) );
System.out.printf( "XSLT body:\n%s\n", xslt.toString( ) );

7.16.6 Using the SPARQL Gateway Graphical Web Interface

SPARQL Gateway provides several browser-based interfaces to help you test queries, navigate semantic data, and manage SPQARQL query and XSLT files:

7.16.6.1 Main Page (index.html)

http://<host>:<port>/sparqlgateway/index.html provides a simple interface for executing SPARQL queries and then applying the transformations in the default.xslt file to the response. Figure 7-2 shows this interface for executing a query.

Figure 7-2 Graphical Interface Main Page (index.html)

Description of Figure 7-2 follows
Description of "Figure 7-2 Graphical Interface Main Page (index.html)"

Enter or select a SPARQL Endpoint, specify the SPARQL SELECT Query Body, and press Submit Query.

For example, if you specify http://dbpedia.org/sparql as the SPARQL endpoint and use the SPARQL query body from Figure 7-2, the response will be similar to Figure 7-3. Note that the default transformations (in default.xslt) have been applied to the XML output in this figure.

Figure 7-3 SPARQL Query Main Page Response

Description of Figure 7-3 follows
Description of "Figure 7-3 SPARQL Query Main Page Response"

7.16.6.2 Navigation and Browsing Page (browse.jsp)

http://<host>:<port>/sparqlgateway/browse.jsp provides navigation and browsing capabilities for semantic data. It works against any standard compliant SPARQL endpoint. Figure 7-4 shows this interface for executing a query.

Figure 7-4 Graphical Interface Navigation and Browsing Page (browse.jsp)

Description of Figure 7-4 follows
Description of "Figure 7-4 Graphical Interface Navigation and Browsing Page (browse.jsp)"

Enter or select a SPARQL Endpoint, specify the SPARQL SELECT Query Body, optionally specify a Timeout (ms) value in milliseconds and the Best Effort option, and press Submit Query.

The SPARQL response is parsed and then presented in table form, as shown in Figure 7-5.

Figure 7-5 Browsing and Navigation Page: Response

Description of Figure 7-5 follows
Description of "Figure 7-5 Browsing and Navigation Page: Response"

In Figure 7-5, note that URIs are clickable to allow navigation, and that when users move the cursor over a URI, tool tips are shown for the URIs which have been shortened for readability (as in http://purl.org.dc/elements/1.1/title being displayed as the tool tip for dc:title in the figure).

If you click the URI http://example.org/book/book5 in the output shown in Figure 7-5, a new SPARQL query is automatically generated and executed. This generated SPARQL query has three query patterns that use this particular URI as subject, predicate, and object, as shown in Figure 7-6. Such a query can give you a good idea about how this URI is used and how it is related to other resources in the data set.

Figure 7-6 Query and Response from Clicking URI Link

Description of Figure 7-6 follows
Description of "Figure 7-6 Query and Response from Clicking URI Link"

When there are many matches of a query, the results are organized in pages and you can click on any page. The page size by default is 50 results. To display more (or fewer) than 50 rows per page in a response with the Browsing and Navigation Page (browse.jsp), you can specify the &resultsPerPage parameter in the URL. For example, to allow 100 rows per page, include the following in the URL:

&resultsPerPage=100

7.16.6.3 XSLT Management Page (xslt.jsp)

http://<host>:<port>/sparqlgateway/admin/xslt.jsp provides a simple XSLT management interface. You can enter an XSLT ID (integer) and click Get XSLT to retrieve both the Description and XSLT Body. You can modify the XSLT Body text and then save the changes by clicking Save XSLT. Note that there is a previewer to help you navigate among available XSLT definitions.

Figure 7-7 shows the XSLT Management Page.

Figure 7-7 XSLT Management Page

Description of Figure 7-7 follows
Description of "Figure 7-7 XSLT Management Page"

7.16.6.4 SPARQL Management Page (sparql.jsp)

http://<host>:<port>/sparqlgateway/admin/xslt.jsp provides a simple SPARQL management interface. You can enter a SPARQL ID (integer) and click Get SPARQL to retrieve both the Description and SPARQL Body. You can modify the SPARQL Body text and then save the changes by clicking Save SPARQL. Note that there is a previewer to help you navigate among available SPARQL queries.

Figure 7-8 shows the SPARQL Management Page.

Figure 7-8 SPARQL Management Page

Description of Figure 7-8 follows
Description of "Figure 7-8 SPARQL Management Page"

7.16.7 Using SPARQL Gateway as an XML Data Source to OBIEE

This section explains how to create an XML Data source for Oracle Business Intelligence Enterprise Edition (OBIEE), by integrating OBIEE with RDF using SPARQL Gateway as a bridge. (The specific steps and illustrations reflect the Oracle BI Administration Tool Version 11.1.1.3.0.100806.0408.000.)

  1. Start the Oracle BI Administration Tool.

  2. Click File, then Import Metadata. The first page of the Import Metadata wizard is displayed, as shown in Figure 7-9.

    Figure 7-9 Import Metadata - Select Data Source

    Description of Figure 7-9 follows
    Description of "Figure 7-9 Import Metadata - Select Data Source"

    Connection Type: Select XML.

    URL: URL for an application to interact with SPARQL Gateway, as explained in Section 7.16.3. You can also include the timeout and best effort options.

    Ignore the User Name and Password fields.

  3. Click Next. The second page of the Import Metadata wizard is displayed, as shown in Figure 7-10.

    Figure 7-10 Import Metadata - Select Metadata Types

    Description of Figure 7-10 follows
    Description of "Figure 7-10 Import Metadata - Select Metadata Types"

    Select the desired metadata types to be imported. Be sure that Tables is included in the selected types.

  4. Click Next. The third page of the Import Metadata wizard is displayed, as shown in Figure 7-11.

    Figure 7-11 Import Metadata - Select Metadata Objects

    Description of Figure 7-11 follows
    Description of "Figure 7-11 Import Metadata - Select Metadata Objects"

    In the Data Source View, expand the node that has the table icon, select the column names (mapped from projected variables defined in the SPARQL SELECT statement), and click the right-arrow (>) button to move the selected columns to the Repository View.

  5. Click Finish.

  6. Complete the remaining steps for the usual BI Business Model work and Mapping and Presentation definition work, which are not specific to SPARQL Gateway or RDF data.