rdf_context 0.5.9.1 → 0.5.10

data/lib/rdf_context/rdfxmlparser.rb

@@ -1,4 +1,4 @@
-require File.join(File.dirname(__FILE__), 'parser')
+require File.join(File.dirname(__FILE__), "nokogiri_hacks")
 
 module RdfContext
   class RdfXmlParser < Parser

data/lib/rdf_context/resource.rb

@@ -0,0 +1,56 @@
+module RdfContext
+  # Common super-class for things that are RDF Resources
+  class Resource
+
+    ##
+    # Returns `true` to indicate that this value is a resource.
+    #
+    # @return [Boolean]
+    def resource?
+      true
+    end
+
+    ##
+    # Returns `false`, overridden in Literal
+    #
+    # @return [Boolean]
+    def literal?
+      false
+    end
+
+    ##
+    # Returns `false`, overridden in URIref
+    #
+    # @return [Boolean]
+    def uri?
+      false
+    end
+
+    ##
+    # Returns `false`, overridden in BNode
+    #
+    # @return [Boolean]
+    def bnode?
+      false
+    end
+    
+    ##
+    # Returns `false`, overridden in BNode
+    #
+    # @return [Boolean]
+    def graph?
+      false
+    end
+    
+    # Parse a string to a resource, in NTriples format
+    def self.parse(str)
+      case str
+      when /^_:/    then BNode.parse(str)
+      when /^</     then URIRef.parse(str)
+      when /^http:/ then URIRef.parse(str)
+      when /^\"/    then Literal.parse(str)
+      else               Literal.parse(str)
+      end
+    end
+  end
+end

data/lib/rdf_context/serializer/abstract_serializer.rb

@@ -1,5 +1,3 @@
-require File.join(File.dirname(__FILE__), '..', 'uriref')
-
 module RdfContext
   # Abstract serializer
   class AbstractSerializer

data/lib/rdf_context/serializer/nt_serializer.rb

@@ -1,5 +1,3 @@
-require File.join(File.dirname(__FILE__), 'abstract_serializer')
-
 module RdfContext
   # Serialize RDF graphs in NTriples format
   class NTSerializer < AbstractSerializer

data/lib/rdf_context/serializer/recursive_serializer.rb

@@ -1,7 +1,3 @@
-require File.join(File.dirname(__FILE__), 'abstract_serializer')
-require File.join(File.dirname(__FILE__), '..', 'bnode')
-require File.join(File.dirname(__FILE__), '..', 'literal')
-
 module RdfContext
   # Recursive serializer
   class RecursiveSerializer < AbstractSerializer

data/lib/rdf_context/serializer/turtle_serializer.rb

@@ -1,5 +1,3 @@
-require File.join(File.dirname(__FILE__), 'recursive_serializer')
-
 module RdfContext
   # Abstract serializer
   class TurtleSerializer < RecursiveSerializer

data/lib/rdf_context/serializer/xml_serializer.rb

@@ -1,5 +1,3 @@
-require File.join(File.dirname(__FILE__), 'abstract_serializer')
-
 module RdfContext
   # Serialize RDF graphs in NTriples format
   class XmlSerializer < RecursiveSerializer

data/lib/rdf_context/store/abstract_sql_store.rb

@@ -1,7 +1,8 @@
-require File.join(File.dirname(__FILE__), 'abstract_store')
 require 'digest/sha1'
 
 module RdfContext
+  autoload :AbstractStore, File.join(File.dirname(__FILE__), 'abstract_store')
+
   # SQL-92 formula-aware implementation of an RDF Store.
   # It stores it's triples in the following partitions:
   # - Asserted non rdf:type statements
@@ -904,6 +905,7 @@ module RdfContext
     # Takes a term value, and term type
     # and Creates a term object.  QuotedGraphs are instantiated differently
     def createTerm(termString,termType,objLanguage=nil,objDatatype=nil)
+      #puts "createTerm(#{termString}, #{termType}, ...)" if ::RdfContext::debug?
       case termType
       when "L"
         @literalCache[[termString, objLanguage, objDatatype]] ||= Literal.n3_encoded(termString, objLanguage, objDatatype)

data/lib/rdf_context/store/active_record_store.rb

@@ -0,0 +1,272 @@
+gem 'activerecord'
+require 'active_record'
+
+module RdfContext
+  autoload :AbstractSQLStore, File.join(File.dirname(__FILE__), 'abstract_sql_store')
+
+  # ActiveRecord store context-ware and formula-aware implementation.
+  #
+  # Uses an ActiveRecord::Base.connection to run statements
+  #
+  # It stores it's triples in the following partitions:
+  # - Asserted non rdf:type statements
+  # - Asserted rdf:type statements (in a table which models Class membership). The motivation for this partition is primarily query speed and scalability as most graphs will always have more rdf:type statements than others
+  # - All Quoted statements
+  #
+  #  In addition it persists namespace mappings in a seperate table
+  class ActiveRecordStore < AbstractSQLStore
+    # Create a new ActiveRecordStore Store
+    # @param [URIRef] identifier
+    # @param[Hash] configuration Specific to type of storage
+    # @return [ActiveRecordStore]
+    def initialize(identifier = nil, configuration = {})
+      super(identifier, configuration)
+      @autocommit_default = false
+    end
+
+    def open(options)
+    end
+
+    def close(commit_pending_transactions = false)
+      ActiveRecord::Base.connection.commit_db_transaction if commit_pending_transactions
+    end
+
+    # Create necessary tables and indecies for this database
+    def setup
+      executeSQL(CREATE_ASSERTED_STATEMENTS_TABLE % @internedId)
+      executeSQL(CREATE_ASSERTED_TYPE_STATEMENTS_TABLE % @internedId)
+      executeSQL(CREATE_QUOTED_STATEMENTS_TABLE % @internedId)
+      executeSQL(CREATE_NS_BINDS_TABLE % @internedId)
+      executeSQL(CREATE_LITERAL_STATEMENTS_TABLE % @internedId)
+      
+      # Create indicies
+      {
+        asserted_table => {
+          "#{@internedId}_A_termComb_index" => %w(termComb),
+          "#{@internedId}_A_s_index" => %w(subject),
+          "#{@internedId}_A_p_index" => %w(predicate),
+          "#{@internedId}_A_o_index" => %w(object),
+          "#{@internedId}_A_c_index" => %w(context),
+        },
+        asserted_type_table => {
+          "#{@internedId}_T_termComb_index" => %w(termComb),
+          "#{@internedId}_T_member_index" => %w(member),
+          "#{@internedId}_T_klass_index" => %w(klass),
+          "#{@internedId}_T_c_index" => %w(context),
+        },
+        literal_table => {
+          "#{@internedId}_L_termComb_index" => %w(termComb),
+          "#{@internedId}_L_s_index" => %w(subject),
+          "#{@internedId}_L_p_index" => %w(predicate),
+          "#{@internedId}_L_c_index" => %w(context),
+        },
+        quoted_table => {
+          "#{@internedId}_Q_termComb_index" => %w(termComb),
+          "#{@internedId}_Q_s_index" => %w(subject),
+          "#{@internedId}_Q_p_index" => %w(predicate),
+          "#{@internedId}_Q_o_index" => %w(object),
+          "#{@internedId}_Q_c_index" => %w(context),
+        },
+        namespace_binds => {
+          "#{@internedId}_uri_index" => %w(uri),
+        }
+      }.each_pair do |tablename, indicies|
+        indicies.each_pair do |index, columns|
+          executeSQL("CREATE INDEX #{index} on #{tablename} ('#{columns.join(', ')}')")
+        end
+      end
+    end
+    
+    # Teardown DB files
+    def teardown
+      # Drop indicies
+      {
+        asserted_table => {
+          "#{@internedId}_A_termComb_index" => %w(termComb),
+          "#{@internedId}_A_s_index" => %w(subject),
+          "#{@internedId}_A_p_index" => %w(predicate),
+          "#{@internedId}_A_o_index" => %w(object),
+          "#{@internedId}_A_c_index" => %w(context),
+        },
+        asserted_type_table => {
+          "#{@internedId}_T_termComb_index" => %w(termComb),
+          "#{@internedId}_T_member_index" => %w(member),
+          "#{@internedId}_T_klass_index" => %w(klass),
+          "#{@internedId}_T_c_index" => %w(context),
+        },
+        literal_table => {
+          "#{@internedId}_L_termComb_index" => %w(termComb),
+          "#{@internedId}_L_s_index" => %w(subject),
+          "#{@internedId}_L_p_index" => %w(predicate),
+          "#{@internedId}_L_c_index" => %w(context),
+        },
+        quoted_table => {
+          "#{@internedId}_Q_termComb_index" => %w(termComb),
+          "#{@internedId}_Q_s_index" => %w(subject),
+          "#{@internedId}_Q_p_index" => %w(predicate),
+          "#{@internedId}_Q_o_index" => %w(object),
+          "#{@internedId}_Q_c_index" => %w(context),
+        },
+        namespace_binds => {
+          "#{@internedId}_uri_index" => %w(uri),
+        }
+      }.each_pair do |tablename, indicies|
+        tn = "#{@internedId}_#{tablename}"
+        indicies.each_pair do |index, columns|
+          executeSQL("DROP INDEX #{index} ON #{tn}")
+        end
+      end
+      
+      # Drop tables
+      executeSQL("DROP TABLE #{namespace_binds}")
+      executeSQL("DROP TABLE #{quoted_table}")
+      executeSQL("DROP TABLE #{literal_table}")
+      executeSQL("DROP TABLE #{asserted_type_table}")
+      executeSQL("DROP TABLE #{asserted_table}")
+    end
+    
+    # Destroy database
+    #
+    # @option options[String] :path Path to database file defaults to a file in the current directory based on a hash of the store identifier
+    def destroy(options = {})
+    end
+
+    protected
+
+    # Where clase utility functions
+    def buildSubjClause(subject, tableName)
+      case subject
+  #    when REGEXTerm
+  #    when Array
+      when Graph
+         ["#{tableName}.subject=?", self.normalizeTerm(subject.identifier)]
+      else
+        ["#{tableName}.subject=?", subject] if subject
+      end
+    end
+
+    def buildPredClause(predicate, tableName)
+  #    case predicate
+  #    when REGEXTerm
+  #    when Array
+  #    else
+        ["#{tableName}.predicate=?", predicate] if predicate
+  #    end
+    end
+
+    # Where clase utility functions
+    def buildObjClause(object, tableName)
+      case object
+  #    when REGEXTerm
+  #    when Array
+    when Graph
+      ["#{tableName}.object=?", self.normalizeTerm(object.identifier)]
+      else
+        ["#{tableName}.object=?", object] if object
+      end
+    end
+
+    # Where clase utility functions
+    def buildContextClause(context, tableName)
+      context = normalizeTerm(context) if context
+
+  #    case context
+  #    when REGEXTerm
+  #    when Array
+  #    else
+        ["#{tableName}.context=?", context] if context
+  #    end
+    end
+
+    # Where clase utility functions
+    def buildTypeMemberClause(subject, tableName)
+  #    case context
+  #    when REGEXTerm
+  #    when Array
+  #    else
+        ["#{tableName}.member=?", subject] if subject
+  #    end
+    end
+
+    # Where clase utility functions
+    def buildTypeClassClause(object, tableName)
+  #    case context
+  #    when REGEXTerm
+  #    else
+        ["#{tableName}.klass=?", object] if object
+  #    end
+    end
+
+    # This takes the query string and parameters and (depending on the SQL implementation) either fill in
+    # the parameter in-place or pass it on to the DB impl (if it supports this).
+    # The default (here) is to fill the parameters in-place surrounding each param with quote characters
+    #
+    # Yields each row
+    def executeSQL(qStr, *params, &block)
+      conn = ActiveRecord::Base.connection
+      
+      puts "executeSQL: params=#{params.flatten.inspect}" if ::RdfContext::debug?
+      puts "executeSQL: '#{params.dup.unshift(qStr).join("', '")}'" if ::RdfContext::debug?
+      qStr = ActiveRecord::Base.send(:replace_bind_variables, qStr, params.flatten)
+      puts "  => '#{qStr}'" if ::RdfContext::debug?
+
+      if block_given?
+        conn.select_rows(qStr).each do |row|
+          yield(row)
+        end
+      else
+        puts "executeSQL no block given" if ::RdfContext::debug?
+        conn.select_rows(qStr)
+      end
+    rescue StandardError => e
+      puts "SQL Exception (ignored): #{e.message}" if ::RdfContext::debug?
+    end
+
+    CREATE_ASSERTED_STATEMENTS_TABLE = %(
+    CREATE TABLE %s_asserted_statements (
+        subject       text not NULL,
+        predicate     text not NULL,
+        object        text not NULL,
+        context       text not NULL,
+        termComb      tinyint unsigned not NULL))
+
+    CREATE_ASSERTED_TYPE_STATEMENTS_TABLE = %(
+    CREATE TABLE %s_type_statements (
+        member        text not NULL,
+        klass         text not NULL,
+        context       text not NULL,
+        termComb      tinyint unsigned not NULL))
+
+    CREATE_LITERAL_STATEMENTS_TABLE = %(
+    CREATE TABLE %s_literal_statements (
+        subject       text not NULL,
+        predicate     text not NULL,
+        object        text,
+        context       text not NULL,
+        termComb      tinyint unsigned not NULL,
+        objLanguage   varchar(3),
+        objDatatype   text))
+
+    CREATE_QUOTED_STATEMENTS_TABLE = %(
+    CREATE TABLE %s_quoted_statements (
+        subject       text not NULL,
+        predicate     text not NULL,
+        object        text,
+        context       text not NULL,
+        termComb      tinyint unsigned not NULL,
+        objLanguage   varchar(3),
+        objDatatype   text))
+
+    CREATE_NS_BINDS_TABLE = %(
+    CREATE TABLE %s_namespace_binds (
+        prefix        varchar(20) UNIQUE not NULL,
+        uri           text,
+        PRIMARY KEY (prefix)))
+
+    DROP_ASSERTED_STATEMENTS_TABLE = %(DROP TABLE %s_asserted_statements)
+    DROP_ASSERTED_TYPE_STATEMENTS_TABLE = %(DROP TABLE %s_type_statements)
+    DROP_LITERAL_STATEMENTS_TABLE = %(DROP TABLE %s_literal_statements)
+    DROP_QUOTED_STATEMENTS_TABLE = %(DROP TABLE %s_quoted_statements)
+    DROP_NS_BINDS_TABLE = %(DROP TABLE %s_namespace_binds)
+  end
+end

data/lib/rdf_context/store/list_store.rb

@@ -1,6 +1,6 @@
-require File.join(File.dirname(__FILE__), 'abstract_store')
-
 module RdfContext
+  autoload :AbstractStore, File.join(File.dirname(__FILE__), 'abstract_store')
+
   # List storage, most efficient, but slow storage model. Works well for basic parse and serialize.
   class ListStore < AbstractStore
     def initialize(identifier = nil, configuration = {})

data/lib/rdf_context/store/memory_store.rb

@@ -1,6 +1,6 @@
-require File.join(File.dirname(__FILE__), 'abstract_store')
-
 module RdfContext
+  autoload :AbstractStore, File.join(File.dirname(__FILE__), 'abstract_store')
+
   # An integer-key-optimized-context-aware-in-memory store.
   #
   # Uses nested dictionaries to store triples and context. Each triple

data/lib/rdf_context/store/sqlite3_store.rb

@@ -1,8 +1,8 @@
 gem 'sqlite3-ruby', "=1.2.5"  # XXX version 1.3.0 & 1.3.1 break executes with multiple arguments: http://groups.google.com/group/sqlite3-ruby/browse_frm/thread/93494ce04bc07fd5
 require 'sqlite3'
-require File.join(File.dirname(__FILE__), 'abstract_sql_store')
 
 module RdfContext
+  autoload :AbstractSQLStore, File.join(File.dirname(__FILE__), 'abstract_sql_store')
   # SQLite3 store context-ware and formula-aware implementation.
   # It stores it's triples in the following partitions:
   # - Asserted non rdf:type statements
@@ -13,16 +13,18 @@ module RdfContext
   #
   # Based on Python RdfLib SQLite
   class SQLite3Store < AbstractSQLStore
-    # Create a new SQLite3Store Store, should be subclassed
+    # Create a new SQLite3Store Store
     # @param [URIRef] identifier
     # @param[Hash] configuration Specific to type of storage
+    # @option configuration [String] :db Path to database file
+    # @option configuration [String] :connection ActiveRecord::Base.connection.raw_connection
     # @return [SQLite3Store]
     def initialize(identifier = nil, configuration = {})
+      @path = configuration[:path] ||= File.join(Dir.getwd, "#{@internedId}.db")
+
       super(identifier, configuration)
 
       @autocommit_default = false
-      @path = configuration[:path] || File.join(Dir.getwd, "#{@internedId}.db")
-      @db = open(:path => @path) unless @db
     end
     
     # Opens the store specified by the configuration hash. If
@@ -34,57 +36,113 @@ module RdfContext
     #
     # @option options[String] :path Path to database file defaults to a file in the current directory based on a hash of the store identifier
     def open(options)
-      unless File.exist?(options[:path])
+      @db ||= options[:connection]
+      return @db if @db
+
+      if options[:path] && !File.exist?(options[:path])
         @db = SQLite3::Database.new(options[:path])
-        executeSQL(CREATE_ASSERTED_STATEMENTS_TABLE % @internedId)
-        executeSQL(CREATE_ASSERTED_TYPE_STATEMENTS_TABLE % @internedId)
-        executeSQL(CREATE_QUOTED_STATEMENTS_TABLE % @internedId)
-        executeSQL(CREATE_NS_BINDS_TABLE % @internedId)
-        executeSQL(CREATE_LITERAL_STATEMENTS_TABLE % @internedId)
-        
-        # Create indicies
-        {
-          asserted_table => {
-            "#{@internedId}_A_termComb_index" => %w(termComb),
-            "#{@internedId}_A_s_index" => %w(subject),
-            "#{@internedId}_A_p_index" => %w(predicate),
-            "#{@internedId}_A_o_index" => %w(object),
-            "#{@internedId}_A_c_index" => %w(context),
-          },
-          asserted_type_table => {
-            "#{@internedId}_T_termComb_index" => %w(termComb),
-            "#{@internedId}_T_member_index" => %w(member),
-            "#{@internedId}_T_klass_index" => %w(klass),
-            "#{@internedId}_T_c_index" => %w(context),
-          },
-          literal_table => {
-            "#{@internedId}_L_termComb_index" => %w(termComb),
-            "#{@internedId}_L_s_index" => %w(subject),
-            "#{@internedId}_L_p_index" => %w(predicate),
-            "#{@internedId}_L_c_index" => %w(context),
-          },
-          quoted_table => {
-            "#{@internedId}_Q_termComb_index" => %w(termComb),
-            "#{@internedId}_Q_s_index" => %w(subject),
-            "#{@internedId}_Q_p_index" => %w(predicate),
-            "#{@internedId}_Q_o_index" => %w(object),
-            "#{@internedId}_Q_c_index" => %w(context),
-          },
-          namespace_binds => {
-            "#{@internedId}_uri_index" => %w(uri),
-          }
-        }.each_pair do |tablename, indicies|
-          indicies.each_pair do |index, columns|
-            executeSQL("CREATE INDEX #{index} on #{tablename} ('#{columns.join(', ')}')")
-          end
-        end
+        setup
       end
 
       raise StoreException.new("Attempt to open missing database file #{options[:path]}") unless File.exist?(options[:path])
       @db = SQLite3::Database.new(options[:path])
     end
 
-    # Destroy databse
+    # Create necessary tables and indecies for this database
+    def setup
+      executeSQL(CREATE_ASSERTED_STATEMENTS_TABLE % @internedId)
+      executeSQL(CREATE_ASSERTED_TYPE_STATEMENTS_TABLE % @internedId)
+      executeSQL(CREATE_QUOTED_STATEMENTS_TABLE % @internedId)
+      executeSQL(CREATE_NS_BINDS_TABLE % @internedId)
+      executeSQL(CREATE_LITERAL_STATEMENTS_TABLE % @internedId)
+      
+      # Create indicies
+      {
+        asserted_table => {
+          "#{@internedId}_A_termComb_index" => %w(termComb),
+          "#{@internedId}_A_s_index" => %w(subject),
+          "#{@internedId}_A_p_index" => %w(predicate),
+          "#{@internedId}_A_o_index" => %w(object),
+          "#{@internedId}_A_c_index" => %w(context),
+        },
+        asserted_type_table => {
+          "#{@internedId}_T_termComb_index" => %w(termComb),
+          "#{@internedId}_T_member_index" => %w(member),
+          "#{@internedId}_T_klass_index" => %w(klass),
+          "#{@internedId}_T_c_index" => %w(context),
+        },
+        literal_table => {
+          "#{@internedId}_L_termComb_index" => %w(termComb),
+          "#{@internedId}_L_s_index" => %w(subject),
+          "#{@internedId}_L_p_index" => %w(predicate),
+          "#{@internedId}_L_c_index" => %w(context),
+        },
+        quoted_table => {
+          "#{@internedId}_Q_termComb_index" => %w(termComb),
+          "#{@internedId}_Q_s_index" => %w(subject),
+          "#{@internedId}_Q_p_index" => %w(predicate),
+          "#{@internedId}_Q_o_index" => %w(object),
+          "#{@internedId}_Q_c_index" => %w(context),
+        },
+        namespace_binds => {
+          "#{@internedId}_uri_index" => %w(uri),
+        }
+      }.each_pair do |tablename, indicies|
+        indicies.each_pair do |index, columns|
+          executeSQL("CREATE INDEX #{index} on #{tablename} ('#{columns.join(', ')}')")
+        end
+      end
+    end
+    
+    # Teardown DB files
+    def teardown
+      # Drop indicies
+      {
+        asserted_table => {
+          "#{@internedId}_A_termComb_index" => %w(termComb),
+          "#{@internedId}_A_s_index" => %w(subject),
+          "#{@internedId}_A_p_index" => %w(predicate),
+          "#{@internedId}_A_o_index" => %w(object),
+          "#{@internedId}_A_c_index" => %w(context),
+        },
+        asserted_type_table => {
+          "#{@internedId}_T_termComb_index" => %w(termComb),
+          "#{@internedId}_T_member_index" => %w(member),
+          "#{@internedId}_T_klass_index" => %w(klass),
+          "#{@internedId}_T_c_index" => %w(context),
+        },
+        literal_table => {
+          "#{@internedId}_L_termComb_index" => %w(termComb),
+          "#{@internedId}_L_s_index" => %w(subject),
+          "#{@internedId}_L_p_index" => %w(predicate),
+          "#{@internedId}_L_c_index" => %w(context),
+        },
+        quoted_table => {
+          "#{@internedId}_Q_termComb_index" => %w(termComb),
+          "#{@internedId}_Q_s_index" => %w(subject),
+          "#{@internedId}_Q_p_index" => %w(predicate),
+          "#{@internedId}_Q_o_index" => %w(object),
+          "#{@internedId}_Q_c_index" => %w(context),
+        },
+        namespace_binds => {
+          "#{@internedId}_uri_index" => %w(uri),
+        }
+      }.each_pair do |tablename, indicies|
+        tn = "#{@internedId}_#{tablename}"
+        indicies.each_pair do |index, columns|
+          executeSQL("DROP INDEX #{index} ON #{tn}")
+        end
+      end
+      
+      # Drop tables
+      executeSQL("DROP TABLE #{namespace_binds}")
+      executeSQL("DROP TABLE #{quoted_table}")
+      executeSQL("DROP TABLE #{literal_table}")
+      executeSQL("DROP TABLE #{asserted_type_table}")
+      executeSQL("DROP TABLE #{asserted_table}")
+    end
+    
+    # Destroy database
     #
     # @option options[String] :path Path to database file defaults to a file in the current directory based on a hash of the store identifier
     def destroy(options = {})
@@ -171,6 +229,7 @@ module RdfContext
       if block_given?
         @db.execute(@statement_cache[qStr], *params) do |row|
           puts "executeSQL res: #{row.inspect}" if ::RdfContext::debug?
+          row = row.keys.select{|k| k.is_a?(Integer)}.sort.map{|k| row[k]} if row.is_a?(Hash)
           yield(row)
         end
       else
@@ -222,5 +281,10 @@ module RdfContext
         uri           text,
         PRIMARY KEY (prefix)))
 
+    DROP_ASSERTED_STATEMENTS_TABLE = %(DROP TABLE %s_asserted_statements)
+    DROP_ASSERTED_TYPE_STATEMENTS_TABLE = %(DROP TABLE %s_type_statements)
+    DROP_LITERAL_STATEMENTS_TABLE = %(DROP TABLE %s_literal_statements)
+    DROP_QUOTED_STATEMENTS_TABLE = %(DROP TABLE %s_quoted_statements)
+    DROP_NS_BINDS_TABLE = %(DROP TABLE %s_namespace_binds)
   end
 end