Class: TTL2HTML::App

Inherits:
Object
  • Object
show all
Includes:
Util
Defined in:
lib/ttl2html.rb

Instance Method Summary collapse

Methods included from Util

#uri_mapping_to_path

Constructor Details

#initialize(config = "config.yml") ⇒ App

Returns a new instance of App.



18
19
20
21
22
23
24
25
26
 
# File 'lib/ttl2html.rb', line 18

def initialize(config = "config.yml")
  @config = load_config(config)
  if not @config[:base_uri]
    raise "load_config: base_uri not found"
  end
  @data = {}
  @data_inverse = {}
  @prefix = {}
end

Instance Method Details

#build_breadcrumbs(uri, template, depth = 0) ⇒ Object 



277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
 
# File 'lib/ttl2html.rb', line 277

def build_breadcrumbs(uri, template, depth = 0)
  results = []
  data = @data[uri]
  if @config[:breadcrumbs]
    if depth == 0
      first_label = template.get_title(data)
      first_label = data[@config[:breadcrumbs].first["label"]].first if @config[:breadcrumbs].first["label"] and data[@config[:breadcrumbs].first["label"]]
      results << { label: first_label }
    end
    @config[:breadcrumbs].each do |e|
      data_target = data
      data_target = @data_inverse[uri] if e["inverse"]
      if data_target
        if e["property"].kind_of? Array
          parent = nil
          data_target_sub = data_target
          e["property"].each do |prop|
            if data_target_sub[prop["property"]]
              data_target_sub[prop["property"]].each do |o|
                parent = o
                data_target_sub = @data[parent]
              end
            end
          end
          if parent
            results << build_breadcrumbs_sub(parent, template)
            results += build_breadcrumbs(parent, template, depth + 1)
            return results
          end
        elsif data_target[e["property"]]
          data_target[e["property"]].each do |parent|
            results << build_breadcrumbs_sub(parent, template, e["label"])
            results += build_breadcrumbs(parent, template, depth + 1)
            return results
          end
        end
      end
    end
  end
  results
end

#build_breadcrumbs_sub(parent, template, label_prop = nil) ⇒ Object 



318
319
320
321
322
323
324
325
326
 
# File 'lib/ttl2html.rb', line 318

def build_breadcrumbs_sub(parent, template, label_prop = nil)
  data_parent = @data[parent]
  label = template.get_title(data_parent)
  label = data_parent[label_prop].first if label_prop and data_parent[label_prop]
  {
    uri: parent,
    label: label,
  }
end

#cleanup ⇒ Object 



518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
 
# File 'lib/ttl2html.rb', line 518

def cleanup
  @data.select do |uri, v|
    uri.start_with? @config[:base_uri]
  end.sort_by do |uri, v|
    -(uri.size)
  end.each do |uri, v|
    html_file = uri_mapping_to_path(uri, @config, ".html")
    html_file = File.join(@config[:output_dir], html_file) if @config[:output_dir]
    File.unlink html_file if File.exist? html_file
    ttl_file = uri_mapping_to_path(uri, @config, ".ttl")
    ttl_file = File.join(@config[:output_dir], ttl_file) if @config[:output_dir]
    File.unlink ttl_file if File.exist? ttl_file
    dir = uri.sub(@config[:base_uri], "")
    dir = File.join(@config[:output_dir], dir) if @config[:output_dir]
    Dir.rmdir dir if File.exist? dir
  end
  index_html = "index.html"
  index_html = File.join(@config[:output_dir], "index.html") if @config[:output_dir]
  if @config[:top_class] and File.exist? index_html
    File.unlink index_html
  end
end

#each_data(label = :each_data) ⇒ Object 



130
131
132
133
134
135
136
137
138
139
140
141
142
143
 
# File 'lib/ttl2html.rb', line 130

def each_data(label = :each_data)
  progressbar_options = {
    output: $stderr,
    title: label.to_s,
    format: "(%t) %a %e %P% Processed: %c from %C"
  }
  data = @data.keys.sort_by do|uri|
    [ uri.count("/"), uri.size, uri ] 
  end.reverse
  Parallel.each(data, progress: progressbar_options) do |uri|
    next if not uri.start_with? @config[:base_uri]
    yield uri, @data[uri]
  end
end

#extract_derivedfrom(data) ⇒ Object 



425
426
427
428
429
430
431
432
433
434
435
 
# File 'lib/ttl2html.rb', line 425

def extract_derivedfrom(data)
  derivedfrom = {}
  wasDerivedFrom = data["http://www.w3.org/ns/prov#wasDerivedFrom"]&.first
  if @data[wasDerivedFrom]
    derivedfrom = {
      url: @data[wasDerivedFrom]["http://www.w3.org/1999/02/22-rdf-syntax-ns#value"]&.first,
      label: @data[wasDerivedFrom]["http://www.w3.org/2000/01/rdf-schema#label"]
    }
  end
  derivedfrom
end

#extract_license(data) ⇒ Object 



436
437
438
439
440
441
442
443
444
445
446
447
448
449
 
# File 'lib/ttl2html.rb', line 436

def extract_license(data)
  license = {}
  if data["http://purl.org/dc/terms/license"]
    license_data = @data[data["http://purl.org/dc/terms/license"].first]
    if license_data
      license[:url] = license_data["http://www.w3.org/1999/02/22-rdf-syntax-ns#value"]&.first
      license[:icon] = license_data["http://xmlns.com/foaf/0.1/thumbnail"]&.first
      license[:label] = license_data["http://www.w3.org/2000/01/rdf-schema#label"]
    elsif data["http://purl.org/dc/terms/license"].first =~ URI::regexp
      license[:url] = license[:label] = data["http://purl.org/dc/terms/license"].first
    end
  end
  license
end

#extract_toplevel ⇒ Object 



450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
 
# File 'lib/ttl2html.rb', line 450

def extract_toplevel
  result = {}
  toplevel = nil
  @data.each do |s, v|
    if @data[s]["http://purl.org/pav/hasCurrentVersion"]
      toplevel = s
    end
  end
  data  = @data[toplevel.to_s]
  if toplevel
    license = extract_license(data)
    derivedfrom = extract_derivedfrom(data)
    if data["http://purl.org/dc/terms/publisher"]
      publisher_data = @data[data["http://purl.org/dc/terms/publisher"].first]
      email = publisher_data["http://xmlns.com/foaf/0.1/mbox"]&.first
      contact = { email: email }
      name = publisher_data["http://xmlns.com/foaf/0.1/name"]
      contact[:name] = name if name
      members = []
      if publisher_data["http://xmlns.com/foaf/0.1/member"]
        publisher_data["http://xmlns.com/foaf/0.1/member"].each do |member|
          member_data = @data[member]
          members << {
            name: member_data["http://xmlns.com/foaf/0.1/name"],
            org: member_data["http://www.w3.org/2006/vcard/ns#organization-name"]
          }
        end
        contact[:members] = members
      end
    end
    if data["http://rdfs.org/ns/void#sparqlEndpoint"]
      endpoint = data["http://rdfs.org/ns/void#sparqlEndpoint"].first
    end
    result = {
      uri: toplevel.to_s,
      description: data["http://purl.org/dc/terms/description"],
      license: license,
      contact: contact,
      endpoint: endpoint,
      derivedfrom: derivedfrom,
    }
  end
  result
end

#extract_version_metadata(data) ⇒ Object 



373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
 
# File 'lib/ttl2html.rb', line 373

def extract_version_metadata(data)
  description = data["http://purl.org/dc/terms/description"]
  link = nil
  if not description
    qrev = data["http://www.w3.org/ns/prov#qualifiedRevision"]&.first
    if @data[qrev]
      description = @data[qrev]["http://www.w3.org/2000/01/rdf-schema#comment"]
      link = @data[qrev]["http://www.w3.org/2000/01/rdf-schema#seeAlso"]&.first
    end
  end
  subset = []
  if data["http://rdfs.org/ns/void#subset"]
    data["http://rdfs.org/ns/void#subset"].each do |s|
      abort "#{s} not found" if not @data[s]
      subset << extract_version_metadata(@data[s])
    end
  end
  date = data["http://purl.org/pav/createdOn"]&.first
  date = data["http://purl.org/dc/terms/issued"]&.first if date.nil?
  return {
    version: data["http://purl.org/pav/version"]&.first,
    triples: data["http://rdfs.org/ns/void#triples"]&.first,
    datadump: data["http://rdfs.org/ns/void#dataDump"]&.first,
    bytesize: data["http://www.w3.org/ns/dcat#byteSize"]&.first,
    date: date,
    description: description,
    subset: subset,
    link: link,
    license: extract_license(data),
    derivedfrom: extract_derivedfrom(data),
  }
end

#extract_versions ⇒ Object 



405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
 
# File 'lib/ttl2html.rb', line 405

def extract_versions
  versions = []
  ["http://purl.org/pav/hasVersion", "http://purl.org/pav/hasCurrentVersion", "http://purl.org/dc/terms/hasVersion"].each do |prop|
    objects = []
    @data.each do |s, v|
      if @data[s][prop]
        objects += @data[s][prop]
      end
    end
    objects.each do |o|
      uri = o.to_s
      version = @data[uri]
      next if not version
      next if not version["http://www.w3.org/1999/02/22-rdf-syntax-ns#type"]
      next if not version["http://www.w3.org/1999/02/22-rdf-syntax-ns#type"].include? "http://rdfs.org/ns/void#Dataset"
      versions << extract_version_metadata(version)
    end
  end
  versions.sort_by{|v| [ v[:date], v[:version] ] }
end

#format_turtle(subject, depth = 1) ⇒ Object 



75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
 
# File 'lib/ttl2html.rb', line 75

def format_turtle(subject, depth = 1)
  turtle = RDF::Turtle::Writer.new
  result = ""
  if subject =~ /^_:/
    result << "[\n#{"  "*depth}"
  else
    result << "<#{subject}>\n#{"  "*depth}"
  end
  result << @data[subject.to_s].keys.sort.map do |predicate|
    str = "<#{predicate}> "
    str << @data[subject.to_s][predicate].sort_by do |object|
      #p [subject, predicate, object, depth]
      if object.to_s =~ /^_:/ and @data[object.to_s]
        qb_order = "http://purl.org/linked-data/cube#order"
        schema_position = "http://schema.org/position"
        shacl_order = "http://www.w3.org/ns/shacl#order"
        [ @data[object.to_s][schema_position] ? @data[object.to_s][schema_position].first.to_i : Float::INFINITY,
          @data[object.to_s][qb_order] ? @data[object.to_s][qb_order].first.to_i : Float::INFINITY,
          @data[object.to_s][shacl_order] ? @data[object.to_s][shacl_order].first.to_i : Float::INFINITY,
          format_turtle(object, depth + 1)
        ]
      else
        object.to_s
      end
    end.map do |object|
      if /^_:/ =~ object.to_s # blank node:
        format_turtle(object, depth + 1)
      elsif RDF::URI::IRI =~ object.to_s
        turtle.format_uri(RDF::URI.new object)
      elsif object.respond_to?(:first) and object.first.kind_of?(Symbol)
        turtle.format_literal(RDF::Literal.new(object[1], language: object[0]))
      else
        turtle.format_literal(object)
      end
    end.join(", ")
    str
  end.join(";\n#{"  "*depth}")
  result << " ." if not subject =~ /^_:/
  result << "\n"
  result << "#{"  "*(depth-1)}]" if subject =~ /^_:/
  result
end

#format_turtle_inverse(object) ⇒ Object 



117
118
119
120
121
122
123
124
125
126
127
128
 
# File 'lib/ttl2html.rb', line 117

def format_turtle_inverse(object)
  result = ""
  return result if not object.start_with? @config[:base_uri]
  return result if not @data_inverse.has_key? object
  @data_inverse[object].keys.sort.each do |predicate|
    @data_inverse[object.to_s][predicate].sort.each do |subject|
      next if subject =~ /^_:/
      result << "<#{subject}> <#{predicate}> <#{object}>.\n"
    end
  end
  result
end

#load_config(file) ⇒ Object 



28
29
30
31
32
33
34
35
36
37
38
39
40
41
 
# File 'lib/ttl2html.rb', line 28

def load_config(file)
  config = { output_turtle: true }
  open(file) do |io|
    YAML.safe_load(io, permitted_classes: [Regexp]).each do |k, v|
      config[k.intern] = v
    end
  end
  [ :css_file, :javascript_file ].each do |k|
    if config[k]
      config[k] = [ config[k] ].flatten
    end
  end
  config
end

#load_turtle(file) ⇒ Object 



43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
 
# File 'lib/ttl2html.rb', line 43

def load_turtle(file)
  $stderr.puts "loading #{file}..."
  count = 0
  subjects = Set.new
  if file.end_with?(".gz")
    io = Zlib::GzipReader.open(file)
  else
    io = File.open(file)
  end
  RDF::Format.for(:turtle).reader.new(io) do |reader|
    reader.each_statement do |statement|
      s = statement.subject
      v = statement.predicate
      o = statement.object
      count += 1
      subjects << s
      @data[s.to_s] ||= {}
      @data[s.to_s][v.to_s] ||= []
      if o.is_a? RDF::URI or o.is_a? RDF::Node
        @data[s.to_s][v.to_s] << o.to_s
        @data_inverse[o.to_s] ||= {}
        @data_inverse[o.to_s][v.to_s] ||= []
        @data_inverse[o.to_s][v.to_s] << s.to_s
      else
        @data[s.to_s][v.to_s] << o
      end
    end
    @prefix.merge! reader.prefixes
  end
  $stderr.puts "#{count} triples. #{subjects.size} subjects."
  @data
end

#output_files ⇒ Object 



513
514
515
516
 
# File 'lib/ttl2html.rb', line 513

def output_files
  output_html_files
  output_turtle_files if @config[:output_turtle]
end

#output_html_files ⇒ Object 



144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
 
# File 'lib/ttl2html.rb', line 144

def output_html_files
  template = Template.new("", @config)
  shapes = []
  @data.each do |s, v|
    if v[RDF.type.to_s] and @data[s][RDF.type.to_s].include?("http://www.w3.org/ns/shacl#NodeShape")
      shapes << s
    end
  end
  labels = shapes2labels(shapes)
  versions = extract_versions
  toplevel = extract_toplevel
  @config[:labels_with_class] ||= {}
  labels.each do |klass, props|
    props.each do |property, label|
      @config[:labels_with_class][klass] ||= {}
      if @config[:labels_with_class][klass][property]
        next
      else
        @config[:labels_with_class][klass][property] = template.get_language_literal(label)
      end
    end
  end
  @config[:orders_with_class] = shapes2orders(shapes)
  Dir.mkdir @config[:output_dir] if @config[:output_dir] and not File.exist? @config[:output_dir]
  each_data(:output_html_files) do |uri, v|
    template = Template.new("default.html.erb", @config)
    param = @config.dup
    param[:uri] = uri
    param[:turtle_uri] = uri + ".ttl"
    param[:data] = v
    param[:data_inverse] = @data_inverse[uri]
    param[:data_inverse_global] = @data_inverse
    param[:data_global] = @data
    param[:title] = template.get_title(v)
    if @config[:subtitle_property] or @config[:subtitle_property_perclass]
      param[:subtitle] = template.get_subtitle(v)
    end
    if param[:breadcrumbs]
      param[:breadcrumbs_items] = build_breadcrumbs(uri, template)
    end
    file = uri_mapping_to_path(uri, @config, ".html")
    if @config[:output_dir]
      file = File.join(@config[:output_dir], file)
    end
    if template.find_template_path("_default.html.erb")
      param[:additional_content] = template.to_html_raw("_default.html.erb", param)
    end
    template.output_to(file, param)
  end
  index_html = "index.html"
  index_html = File.join(@config[:output_dir], "index.html") if @config[:output_dir]
  if @config.has_key? :top_class
    subjects = []
    @data.each do |s, v|
      if @data[s][RDF.type.to_s] and @data[s][RDF.type.to_s].include?(@config[:top_class])
        subjects << s
      end
    end
    if subjects.empty?
      $stderr.puts "WARN: top_class parameter specified as [#{@config[:top_class]}], but there is no instance data."
    else
      template = Template.new("index.html.erb", @config)
      param = @config.dup
      param[:class_label] = template.get_title(@data[@config[:top_class]], nil)
      param[:data_global] = @data
      param[:data_inverse_global] = @data_inverse
      param[:versions] = versions
      param[:toplevel] = toplevel
      param[:description] = template.to_html_raw("description.html", {}) if template.find_template_path("description.html")
      subjects.sort.each do |subject|
        objects = []
        if @config.has_key? :top_additional_property
          @config[:top_additional_property].each do |property|
            if @data[subject][property]
              objects += @data[subject][property]
            end
          end
        end
        param[:index_data] ||= []
        param[:index_data] << {
          subject.to_s => objects
        }
      end
      param[:output_file] = index_html
      param[:index_list] = template.to_html_raw("index-list.html.erb", param)
      template.output_to(index_html, param)
    end
  end
  if template.find_template_path("about.html") or shapes.size > 0 or versions.size > 0 or toplevel.size > 0
    about_html = @config[:about_file] || "about.html"
    about_html =  File.join(@config[:output_dir], about_html) if @config[:output_dir]
    template = Template.new("about.html.erb", @config)
    param = @config.dup
    param[:content] = template.to_html_raw("about.html", {}) if template.find_template_path("about.html")
    param[:data_global] = @data
    param[:versions] = versions
    param[:toplevel] = toplevel
    param[:description] = template.to_html_raw("description.html", {}) if template.find_template_path("description.html")
    param[:shapes] = {}
    shapes.each do |subject|
      orders = []
      if param[:shape_orders]
        param[:shape_orders].index(subject)
        orders << ( param[:shape_orders].index(subject) or Float::INFINITY )
      end
      orders << subject
      label = comment = nil
      target_class = @data[subject.to_s]["http://www.w3.org/ns/shacl#targetClass"]
      if target_class
        target_class = target_class.first
        if @data[target_class]
          label = template.get_title(@data[target_class], nil)
          comment = template.get_language_literal(@data[target_class]["http://www.w3.org/2000/01/rdf-schema#comment"]) if @data[target_class]["http://www.w3.org/2000/01/rdf-schema#comment"]
        else
          label = template.format_property(target_class)
        end
      else
        label = template.get_title(@data[subject.to_s])
      end
      html = template.expand_shape(@data, subject.to_s, @prefix)
      next if html.nil?
      param[:shapes][subject] = {
        label: label,
        comment: comment,
        html: html,
        target_class: target_class,
        order: orders,
      }
    end
    template.output_to(about_html, param)
  end
end

#output_turtle_files ⇒ Object 



495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
 
# File 'lib/ttl2html.rb', line 495

def output_turtle_files
  Dir.mkdir @config[:output_dir] if @config[:output_dir] and not File.exist? @config[:output_dir]
  each_data(:output_turtle_files) do |uri, v|
    file = uri_mapping_to_path(uri, @config, ".ttl")
    if @config[:output_dir]
      Dir.mkdir @config[:output_dir] if not File.exist? @config[:output_dir]
      file = File.join(@config[:output_dir], file)
    end
    dir = File.dirname(file)
    FileUtils.mkdir_p(dir) if not File.exist?(dir)
    str = format_turtle(uri)
    str << format_turtle_inverse(uri)
    open(file, "w") do |io|
      io.puts str.strip
    end
  end
end

#shapes2labels(shapes) ⇒ Object 



352
353
354
355
356
357
358
359
360
361
 
# File 'lib/ttl2html.rb', line 352

def shapes2labels(shapes)
  labels = {}
  shapes_parse(shapes) do |target_class, property|
    path = @data[property]["http://www.w3.org/ns/shacl#path"].first
    name = @data[property]["http://www.w3.org/ns/shacl#name"]
    labels[target_class] ||= {}
    labels[target_class][path] = name
  end
  labels
end

#shapes2orders(shapes) ⇒ Object 



362
363
364
365
366
367
368
369
370
371
 
# File 'lib/ttl2html.rb', line 362

def shapes2orders(shapes)
  orders = {}
  shapes_parse(shapes) do |target_class, property|
    path = @data[property]["http://www.w3.org/ns/shacl#path"].first
    order = @data[property]["http://www.w3.org/ns/shacl#order"]
    orders[target_class] ||= {}
    orders[target_class][path] = order&.first&.to_i
  end
  orders
end

#shapes_parse(shapes) ⇒ Object 



328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
 
# File 'lib/ttl2html.rb', line 328

def shapes_parse(shapes)
  shapes.each do |shape|
    target_class = @data[shape]["http://www.w3.org/ns/shacl#targetClass"]&.first
    if target_class
      properties = @data[shape.to_s]["http://www.w3.org/ns/shacl#property"]
      if @data[shape.to_s]["http://www.w3.org/ns/shacl#or"]
        properties ||= []
        node_list = @data[shape.to_s]["http://www.w3.org/ns/shacl#or"].first
        while node_list and @data[node_list] do
          sub_shape = @data[node_list]["http://www.w3.org/1999/02/22-rdf-syntax-ns#first"].first
          if @data[sub_shape] and @data[sub_shape]["http://www.w3.org/ns/shacl#property"]
            properties += @data[sub_shape]["http://www.w3.org/ns/shacl#property"]
          end
          node_list = @data[node_list]["http://www.w3.org/1999/02/22-rdf-syntax-ns#rest"].first
        end
      end
      if not properties.empty?
        properties.each do |property|
          yield target_class, property
        end
      end
    end
  end
end