antoniogarrote / clj-r2rml

# clj-r2rml

Implementation of 'R2RML: RDB to RDF Mapping Language' W3C proposal for mapping relational databases into RDF in the Clojure programming language.
It covers:

 - support for custom mappings
 - support for the default direct mapping as described in 'A Direct Mapping of Relational Data to RDF'

Relevant W3C document URIs:

http://www.w3.org/TR/2010/WD-r2rml-20101028/
http://www.w3.org/TR/2010/WD-rdb-direct-mapping-20101118/

The current implementation is MySQL centered.

## Usage

To create a mapping use Clojure maps describing the RDF triples in the mapping.
For example the mapping:

  {:logical-table   "select concat('_:Department',deptno) AS deptid, deptno, dname, loc from Dept"
   :class           "xyz:dept"
   :table-graph-iri "xyz:DeptGraph"
   :subject-map     {:column "deptid"}
   :property-object-map [{:property "dept:deptno"
                          :column   "deptno"
                          :datatype "xsd:positiveInteger"}
                         {:property "dept:name"
                          :column   "dname"}
                         {:property "dept:location"
                          :column   "loc"}
                         {:property       "dept:COMPANY"
                          :constant-value "XYZ Corporation"}]}

is equivalent to the following mapping expressed using the Turtle RDF syntax:

  <#TriplesMap1>
      a rr:TriplesMap;
      rr:logicalTable "
         Select ('_:Department' || deptno) AS deptid
              , deptno
              , dname
              , loc
           from dept
         ";
      rr:class xyz:dept;
      rr:tableGraphIRI xyz:DeptGraph;
      rr:subjectMap [ a rr:BlankNodeMap; rr:column "deptid";
                      rr:InverseExpression "{alias.}deptno = substr({alias.}deptid,length('_:Department')+1)"];
      rr:propertyObjectMap [ rr:property dept:deptno; rr:column "deptno"; rr:datatype xsd:positiveInteger ];
      rr:propertyObjectMap [ rr:property dept:name; rr:column "dname" ];
      rr:propertyObjectMap [ rr:property dept:location; rr:column "loc" ];
      rr:propertyObjectMap [ rr:property dept:COMPANY; rr:constantValue "XYZ Corporation" ];
  .

To run the mapping over some relational data, you need to create a mapping context containing a DB connection and call the run-mapping function,
the resulting triples will be stored in the :results key of the context:

  (def *context* (make-context *db-spec* {}))
  (def *res* (:results (run-mapping clj-r2rml.core/test-spec *context* {})))

The result will be stored in Quad records, containing a subject, a predicate, an object and a graph value. The object can containg a Literal record with value, datatype, and language.
For the previous table with the following tuples:


  Column Name	Column Datatype	   Column Key Constraints
  deptno	INTEGER	           UNIQUE
  dname	        VARCHAR(30)
  loc	        VARCHAR(100)

  deptno	  dname	       loc
  10	        APPSERVER    NEW YORK

The following triples are the generated in this case:

  (#:clj-r2rml.core.Quad{:subject "_:Department10",
                         :predicate "dept:deptno",
                         :object #:clj-r2rml.core.Literal{:value "10",
                                                          :datatype "xsd:positiveInteger",
                                                          :language nil},
                         :graph "xyz:DeptGraph"}
   #:clj-r2rml.core.Quad{:subject "_:Department10",
                         :predicate "dept:name",
                         :object #:clj-r2rml.core.Literal{:value "APPSERVER",
                                                          :datatype nil,
                                                          :language nil},
                         :graph "xyz:DeptGraph"}
   #:clj-r2rml.core.Quad{:subject "_:Department10",
                         :predicate "dept:location",
                         :object #:clj-r2rml.core.Literal{:value "NEW YORK",
                                                          :datatype nil,
                                                          :language nil},
                         :graph "xyz:DeptGraph"}
   #:clj-r2rml.core.Quad{:subject "_:Department10",
                         :predicate "dept:COMPANY",
                         :object #:clj-r2rml.core.Literal{:value "XYZ Corporation",
                                                          :datatype nil,
                                                          :language nil},
                         :graph "xyz:DeptGraph"}
   #:clj-r2rml.core.Quad{:subject "_:Department10",
                         :predicate "rdf:type",
                         :object "xyz:dept",
                         :graph "xyz:DeptGraph"})

The default mapping can be applied using the default-mapping-triples function, providing a default
namespace for the triples, a map of virtual tables and a context containing a database connection:

  (default-mapping-triples "http://clj-r2rml.test" {"Dept" "select * from Dept"} *context*)

The generated triples in this case are the following:

  (#:clj-r2rml.core.Quad{:subject "_:0",
                         :predicate "http://clj-r2rml.test/Dept#deptno",
                         :object #:clj-r2rml.core.Literal{:value "10",
                                                          :datatype "http://www.w3.org/TR/xmlschema-2/#integer",
                                                          :language nil}, :graph nil}
   #:clj-r2rml.core.Quad{:subject "_:0",
                         :predicate "http://clj-r2rml.test/Dept#dname",
                         :object #:clj-r2rml.core.Literal{:value "APPSERVER",
                                                          :datatype nil,
                                                          :language nil},
                         :graph nil}
   #:clj-r2rml.core.Quad{:subject "_:0",
                         :predicate "http://clj-r2rml.test/Dept#loc",
                         :object #:clj-r2rml.core.Literal{:value "NEW YORK",
                                                          :datatype nil,
                                                          :language nil},
                         :graph nil}
   #:clj-r2rml.core.Quad{:subject "_:0",
                         :predicate "rdf:type",
                         :object "http://clj-r2rml.test/Dept",
                         :graph nil})

A collection of Quad records can be transformed into RDF in the turtle format using the to-rdf function.

## Installation

Generate a jar using leiningen and add it to the dependencies of your project. Not in clojars yet.

## License

Copyright (C) 2010 Antonio Garrote

Deistributed under the LGPL license