csv2rdf 0.0.0

data/lib/csv2rdf.rb

@@ -0,0 +1,9 @@
+#!/usr/bin/env ruby
+
+require 'csv'
+require 'rdf'
+require 'rdf/ntriples'
+require 'active_support/inflector'
+
+require_relative 'csv2rdf/vocabs'
+require_relative 'csv2rdf/converter'

data/lib/csv2rdf/converter.rb

@@ -0,0 +1,123 @@
+require 'rdf'
+require 'rdf/ntriples'
+require 'logger'
+
+
+# Abstract super-class for all converters.
+#
+# == Usage
+#
+# Usually, a converter will be used as follows:
+#  converter = RatingConverter.new("path/to/file.csv", "path/to/output_file.nt", context_object)
+#  converter.convert
+#  converter.serialize
+#
+class Converter
+  
+  # The Logger object, defaults to STDOUT
+  attr_accessor :log
+  
+  # ==== Attributes
+  #
+  # * +csv_in+ - Path to the CSV file that is going to be converted. Is tested for existence and type, might raise an 
+  #   +IOError+.
+  # * +out_file+ - Path to the desired output file. The containing folder is tested for existence, might raise an 
+  #   +IOError+.
+  # * +context+ - Optionally a context Hash, can be used to pass information to the converter.
+  # * +log+ - the Logger object, defaults to STDOUT
+  #
+  def initialize(csv_in, out_file, context=nil)
+    unless (File.exists?(csv_in))
+      raise IOError, "file '#{csv_in}' does not exist"
+    end
+    unless (File.file?(csv_in))
+      raise IOError, "'#{csv_in}' is not a regular file"
+    end
+    unless (File.exists?(File.dirname(out_file)))
+      raise IOError, "directory '#{File.dirname(out_file)}' does not exist"
+    end
+    @csv_in = csv_in
+    @rdf_out = out_file
+    @context = context
+    @graph = RDF::Graph.new
+    @log = Logger.new(STDOUT)
+    @log.info("#{self.class}: converting #{@csv_in} to #{@rdf_out}")
+  end
+  
+  # Turn any string into a URI component, useful for creating URIs from names and titles. 
+  # The conversion is based on ActiveSupport's +parameterize+ method.
+  #
+  # ==== Attributes
+  #
+  # * +name+ - the string to be converted
+  # * +capitalize+ - if TRUE, the individual components of the name will be capitalized
+  #
+  # ==== Examples
+  #
+  #  Converter.name2uri("Knud Möller")
+  #   => "knud-moller"
+  #
+  #  Converter.name2uri("Knud Möller", TRUE)
+  #   => "Knud-Moller"
+  #
+  def Converter.name2uri(name, capitalize=FALSE)
+    name = name.parameterize.downcase
+    if capitalize
+      name = name.split("-").map { |x| x.capitalize }.join("-")
+    end
+    return name
+  end
+
+  # Convert the string of a German-style float ("1,59") into an actual float.
+  # If passed an actual float, it will just return it.
+  #
+  # ==== Attributes
+  #
+  # * +float+ - the string to be converted
+  #
+  # ==== Examples
+  #
+  #  Converter.german_to_english_float("1,59")
+  #   => 1.59
+  #
+  def Converter.german_to_english_float(float)
+    return float.to_s.gsub(",", ".").to_f
+  end
+
+  # Convert a string into a boolean
+  # For string.downcase == "ja" the method will return TRUE, for all other 
+  # values it will return FALSE
+  # 
+  # === Attributes
+  # 
+  # * +string+ - the string to be converted
+  # 
+  # === Examples
+  # 
+  #  Converter.ja_nein("ja")
+  #   => true
+  #  Converter.ja_nein("Ja")
+  #   => true
+  #  Converter.ja_nein("nein")
+  #   => false
+  #  Converter.ja_nein("hurtz")
+  #   => false
+  # 
+  def Converter.ja_nein(string)
+    return string.downcase.eql?("ja")
+  end
+
+  # The actual conversion takes place in this method. Implementations in sub-classes 
+  # need to build the output graph on the +graph+ instance variable.
+  #
+  def convert
+    raise "method #{self.class.name}#convert() is not implemented!"
+  end
+  
+  # After calling +convert+, +serialize+ will write the +graph+ to the desired output path
+  # (+rdf_out+) as an N-Triples file (http://www.w3.org/TR/n-triples/).
+  #
+  def serialize
+    RDF::Writer.open(@rdf_out) { |writer| writer << @graph }
+  end
+end

data/lib/csv2rdf/vocabs.rb

@@ -0,0 +1,10 @@
+require 'rdf'
+
+#
+# Some widely used RDF vocabulary namespaces.
+#
+
+SCHEMA = RDF::Vocabulary.new("http://schema.org/")
+GEO = RDF::Vocabulary.new("http://www.w3.org/2003/01/geo/wgs84_pos#>")
+GR = RDF::Vocabulary.new("http://purl.org/goodrelations/v1#")
+DCT = RDF::Vocabulary.new("http://purl.org/dc/terms/")

metadata

@@ -0,0 +1,80 @@
+--- !ruby/object:Gem::Specification
+name: csv2rdf
+version: !ruby/object:Gem::Version
+  version: 0.0.0
+  prerelease: 
+platform: ruby
+authors:
+- Knud Möller
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2013-12-01 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: rdf
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 1.0.9
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 1.0.9
+- !ruby/object:Gem::Dependency
+  name: activesupport
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 4.0.1
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 4.0.1
+description: A super light-weight framework for converting arbitrary CSV files to
+  RDF.
+email: knud@datalysator.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/csv2rdf.rb
+- lib/csv2rdf/vocabs.rb
+- lib/csv2rdf/converter.rb
+homepage: https://github.com/knudmoeller/csv2rdf
+licenses: []
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 1.8.25
+signing_key: 
+specification_version: 3
+summary: Convert CSV files to RDF
+test_files: []