MIRIAM XML to RDF

miriam_xml2rdf.rb

#!/usr/bin/env ruby
#
# Convert MIRIAM Registry XML file (http://www.ebi.ac.uk/miriam/main/export/) to RDF
#
# Copyright (C) 2012 Toshiaki Katayama <[email protected]>
#
# Pre requirements:
#  % curl http://www.ebi.ac.uk/miriam/main/export/xml/ > miriam.xml
#  % gem install nokogiri
#  % miriam_xml2rdf.rb miriam.xml > miriam.rdf
#

require 'rubygems'
require 'nokogiri'
require 'erb'

$turtle_backward_compatible = true

xml = Nokogiri::XML(ARGF)
erb_template = DATA.read

ns = xml.namespaces
meta = xml.xpath('//xmlns:miriam', ns).first

def clean(str)
  return str.gsub(/\s+/, ' ').strip.gsub('"', '\\"')
end

puts <<HEADER
@prefix : <http://identifiers.org/dataset/> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix skos: <http://www.w3.org/2004/02/skos/core#> .
@prefix void: <http://rdfs.org/ns/void#> .
@prefix dcterms: <http://purl.org/dc/terms/> .
@prefix miriam: <#{ns['xmlns']}> .

<>
  dcterms:date "#{meta['date']}" ;
  dcterms:hasVersion "#{meta['data-version']}" .

HEADER

xml.xpath('//xmlns:datatype', ns).each do |datatype|
  next if datatype["obsolete"] == "true"

  path = './/xmlns:uris/xmlns:uri[@type="URN"]'
  path2 = './/xmlns:uris/xmlns:uri[@deprecated="true"]'
  urns = datatype.xpath(path, ns).collect(&:text) - datatype.xpath(path2, ns).collect(&:text)

  path = './/xmlns:uris/xmlns:uri[@type="URL"]'
  path2 = './/xmlns:uris/xmlns:uri[@deprecated="true"]'
  urls = datatype.xpath(path, ns).collect(&:text) - datatype.xpath(path2, ns).collect(&:text)

  namespace = datatype.at('namespace').content
  classname = namespace.split(/\.|\-/).map{|x| x.capitalize}.join
  if classname[/^\d/] and $turtle_backward_compatible
    classname = "DB" + classname
  end

  rdfs_label = clean( datatype.at('name').content )
  rdfs_comment = clean( datatype.at('definition').content )
  datatype_id = datatype["id"]
  urn = urns.first
  url = urls.first

  erb = ERB.new(erb_template)
  puts erb.result(binding)
end

__END__
:<%= classname %>
  rdf:type void:Dataset ;
  rdfs:label "<%= rdfs_label %>" ;
  rdfs:comment "<%= rdfs_comment %>" ;
  miriam:datatype "<%= datatype_id %>" ;
  miriam:urn <<%= urn %>> ;
  miriam:url <<%= url %>> ;
  miriam:namespace "<%= namespace %>" .