Class: RDF::Borsh::Writer

Inherits:
Writer
  • Object
show all
Defined in:
lib/rdf/borsh/writer.rb

Constant Summary collapse

MAGIC = 
RDF::Borsh::Format::MAGIC
VERSION = 
RDF::Borsh::Format::VERSION
FLAGS = 
RDF::Borsh::Format::FLAGS
LZ4HC_CLEVEL_MAX = 
12

Instance Method Summary collapse

Constructor Details

#initialize(output = $stdout, **options) {|writer| ... } ⇒ void

Initializes the RDF/Borsh writer.

Parameters:

  • output (IO, StringIO) (defaults to: $stdout)
  • options (Hash{Symbol => Object})

Yields:

  • (writer)

Yield Parameters:

Yield Returns:

  • (void) 


26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
 
# File 'lib/rdf/borsh/writer.rb', line 26

def initialize(output = $stdout, **options, &block)
  output.extend(Borsh::Writable)
  output.binmode if output.respond_to?(:binmode)

  @terms_dict, @terms_map = [], {}
  @quads_set = SortedSet.new

  super(output, **options) do
    if block_given?
      case block.arity
        when 0 then self.instance_eval(&block)
        else block.call(self)
      end
      self.finish
    end
  end
end

Instance Method Details

#compress {|Borsh::Buffer| ... } ⇒ String

Yields:

  • (Borsh::Buffer)

Returns:

  • (String) 


169
170
171
172
 
# File 'lib/rdf/borsh/writer.rb', line 169

def compress(&block)
  uncompressed = Borsh::Buffer.open(&block)
  LZ4::BlockEncoder.new(LZ4HC_CLEVEL_MAX).encode(uncompressed)
end

#finish ⇒ void

This method returns an undefined value.

Finishes writing the output.



84
85
86
87
88
 
# File 'lib/rdf/borsh/writer.rb', line 84

def finish
  self.write_header
  self.write_terms
  self.write_quads
end

#flush ⇒ void

This method returns an undefined value.

Flushes the output.



75
76
77
78
 
# File 'lib/rdf/borsh/writer.rb', line 75

def flush
  self.finish
  super
end

#intern_term(term) ⇒ Integer

Interns the given RDF term.

Parameters:

  • term (RDF::Term)

Returns:

  • (Integer) 


155
156
157
158
159
160
161
162
163
164
 
# File 'lib/rdf/borsh/writer.rb', line 155

def intern_term(term)
  return 0 if term.nil? # for the default graph
  term_id = @terms_map[term]
  if !term_id
    term_id = @terms_dict.size + 1
    @terms_dict << term
    @terms_map[term] = term_id
  end
  term_id
end

#write_header ⇒ void

This method returns an undefined value.

Writes the uncompressed header.



94
95
96
97
98
 
# File 'lib/rdf/borsh/writer.rb', line 94

def write_header
  @output.binmode
  @output.write([MAGIC, VERSION, FLAGS].pack('a4CC'))
  @output.write_u32(@quads_set.size)
end

#write_quad(subject, predicate, object, context) ⇒ void

This method returns an undefined value.

Writes an RDF quad.

Parameters:

  • subject (RDF::Resource)
  • predicate (RDF::URI)
  • object (RDF::Term)
  • context (RDF::Resource) 


63
64
65
66
67
68
69
 
# File 'lib/rdf/borsh/writer.rb', line 63

def write_quad(subject, predicate, object, context)
  s = self.intern_term(subject)
  p = self.intern_term(predicate)
  o = self.intern_term(object)
  g = self.intern_term(context)
  @quads_set << [g, s, p, o]
end

#write_quads ⇒ void

This method returns an undefined value.

Writes the compressed quads set.



139
140
141
142
143
144
145
146
147
148
 
# File 'lib/rdf/borsh/writer.rb', line 139

def write_quads
  buffer = self.compress do |output|
    output.write_u32(@quads_set.size)
    @quads_set.each do |quad|
      quad.each { |tid| output.write_u16(tid) }
    end
  end
  @output.write_u32(buffer.size)
  @output.write(buffer)
end

#write_terms ⇒ void

This method returns an undefined value.

Writes the compressed terms dictionary.



104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
 
# File 'lib/rdf/borsh/writer.rb', line 104

def write_terms
  buffer = self.compress do |output|
    output.write_u32(@terms_dict.size)
    @terms_dict.each do |term|
      output.write(case
        when term.iri?
          string = term.to_s
          [1, string.bytesize, string].pack('CVa*')
        when term.node?
          string = term.id.to_s
          [2, string.bytesize, string].pack('CVa*')
        when term.literal? && term.plain?
          string = term.value.to_s
          [3, string.bytesize, string].pack('CVa*')
        when term.literal? && term.datatype?
          string = term.value.to_s
          datatype = term.datatype.to_s
          [4, string.bytesize, string, datatype.bytesize, datatype].pack('CVa*Va*')
        when term.literal? && term.language?
          string = term.value.to_s
          language = term.language.to_s
          [5, string.bytesize, string, datatype.language, language].pack('CVa*Va*')
        else
          raise RDF::WriterError, "unsupported RDF/Borsh term type: #{term.inspect}"
      end)
    end
  end
  @output.write_u32(buffer.size)
  @output.write(buffer)
end

#write_triple(subject, predicate, object) ⇒ void

This method returns an undefined value.

Writes an RDF triple.

Parameters:

  • subject (RDF::Resource)
  • predicate (RDF::URI)
  • object (RDF::Term) 


51
52
53
 
# File 'lib/rdf/borsh/writer.rb', line 51

def write_triple(subject, predicate, object)
  self.write_quad(subject, predicate, object, nil)
end