relaton 3.1.5 → 3.1.6

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -87,6 +87,15 @@ module Relaton
87
87
  # Cloudflare; tune via RELATON_CIE_CONCURRENCY (see .concurrency).
88
88
  DEFAULT_CONCURRENCY = 5
89
89
 
90
+ # Languages a catalogue translation can carry (relaton-cie#13). A hit for
91
+ # one of them (`cie-232-2019-de?product_id=…`, `CIE 232:2019(DE)`) has the
92
+ # primary id of its English document, so it merges into that record.
93
+ TRANSLATION_LANGUAGES = %w[de fr es it ru ja zh pt ko].freeze
94
+
95
+ # ISO 15924 script of a translation language that is not Latin.
96
+ SCRIPTS = { "ru" => "Cyrl", "ja" => "Jpan", "zh" => "Hans",
97
+ "ko" => "Kore" }.freeze
98
+
90
99
  # Per-worker pacing bounds (seconds). Each worker starts at BASE_GAP and
91
100
  # doubles its own gap up to MAX_GAP on trouble (see Pacing / #time_req).
92
101
  BASE_GAP = 1
@@ -99,7 +108,7 @@ module Relaton
99
108
  [(ENV["RELATON_CIE_CONCURRENCY"] || DEFAULT_CONCURRENCY).to_i, 1].max
100
109
  end
101
110
 
102
- def initialize(output, format)
111
+ def initialize(output, format) # rubocop:disable Metrics/MethodLength
103
112
  super
104
113
  # Guards the shared bookkeeping (@files, @errors, @seen) and the index
105
114
  # mutation while workers fetch detail pages in parallel. The slow
@@ -109,6 +118,10 @@ module Relaton
109
118
  # duplicate primary id resolves to the same last-by-position winner the
110
119
  # serial crawl would pick, regardless of worker completion order.
111
120
  @seen = {}
121
+ # output file => the item #write_file last wrote there, and
122
+ # output file => [[pos, translation item], ...] for #merge_translations.
123
+ @items = {}
124
+ @translations = Hash.new { |h, k| h[k] = [] }
112
125
  end
113
126
 
114
127
  # Factory for a fresh stealth browser agent. The pool builds one per worker
@@ -243,14 +256,15 @@ module Relaton
243
256
 
244
257
  # @param doc [Mechanize::Page]
245
258
  # @return [Array<Relaton::Bib::Title>]
246
- def fetch_title(doc)
259
+ def fetch_title(doc, lang = nil)
247
260
  t = doc.at_xpath("//hgroup/h2/text()|//hgroup/h1/text()")
248
261
  unless t && !t.text.strip.empty?
249
262
  @errors[:title] &&= true
250
263
  return []
251
264
  end
252
265
 
253
- result = Bib::Title.from_string t.text.strip
266
+ script = script_for(lang) if lang
267
+ result = Bib::Title.from_string t.text.strip, lang, script
254
268
  @errors[:title] &&= result.empty?
255
269
  result
256
270
  end
@@ -310,20 +324,45 @@ module Relaton
310
324
  [Bib::Uri.new(type: "src", content: url)]
311
325
  end
312
326
 
313
- # @param doc [Mechanize::Page]
327
+ # An XPath test that the element's class attribute has the exact token.
328
+ def self.class_token(token)
329
+ "contains(concat(' ', normalize-space(@class), ' '), ' #{token} ')"
330
+ end
331
+
332
+ # The document description: the div whose class has the exact token
333
+ # `description`. A class substring match also takes the
334
+ # `available-format-description` divs of the "Document Delivery Options"
335
+ # popup, and wrote a format blurb as the abstract (relaton-cie#7).
336
+ ABSTRACT_XPATH = "//div[#{class_token 'description'}]".freeze
337
+ # The text of the description, without its "See more" toggle link.
338
+ ABSTRACT_TEXT_XPATH =
339
+ ".//text()[not(ancestor::a[#{class_token 'full-description__more'}]) " \
340
+ "and not(ancestor::script) and not(ancestor::style)]".freeze
341
+
342
+ # @param doc [Moxml::Document, Nokogiri::HTML::Document]
314
343
  # @return [Array<Relaton::Bib::LocalizedMarkedUpString>]
315
- def fetch_abstract(doc)
316
- content = doc.at_xpath('//div[contains(@class,"description")]')&.text&.strip
344
+ def fetch_abstract(doc, lang = "en")
345
+ content = abstract_text doc
317
346
  if content.nil? || content.empty?
318
347
  @errors[:abstract] &&= true
319
348
  return []
320
349
  end
321
350
 
322
- result = [Bib::Abstract.new(content: content, language: "en", script: "Latn")]
351
+ result = [Bib::Abstract.new(content: content, language: lang,
352
+ script: script_for(lang))]
323
353
  @errors[:abstract] &&= result.empty?
324
354
  result
325
355
  end
326
356
 
357
+ # @param doc [Moxml::Document, Nokogiri::HTML::Document]
358
+ # @return [String, nil]
359
+ def abstract_text(doc)
360
+ div = doc.at_xpath ABSTRACT_XPATH
361
+ return unless div
362
+
363
+ div.xpath(ABSTRACT_TEXT_XPATH).map(&:text).join.strip
364
+ end
365
+
327
366
  # @param doc [Mechanize::Page]
328
367
  # @return [Array<Relaton::Bib::Contributor>]
329
368
  def fetch_contributor(doc) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength,Metrics/CyclomaticComplexity
@@ -391,6 +430,7 @@ module Relaton
391
430
  # Only the last-by-position hit writes the shared file's content.
392
431
  return if superseded? file, pos
393
432
 
433
+ @items[file] = bib
394
434
  if @files.include? file
395
435
  Util.warn do
396
436
  "File #{file} exists. Docid: #{bib.docidentifier[0].content}\n" \
@@ -416,6 +456,86 @@ module Relaton
416
456
  false
417
457
  end
418
458
 
459
+ # The language of a translation hit, or nil for an English one. The
460
+ # listing URL's slug names it (`…-de?product_id=`; English pages end in
461
+ # `-en` or have no slug); a `(DE)` suffix in the code is the fallback.
462
+ #
463
+ # @param hit [Moxml::Element]
464
+ # @return [String, nil]
465
+ def translation_language(hit)
466
+ link = hit.at_xpath(".//h3/a")
467
+ code = link.text.strip.delete_suffix("\u25b9").strip
468
+ lang = link[:href].to_s[/-([a-z]{2})\?product_id=/, 1] ||
469
+ code[/\(([a-z]{2})\)\z/i, 1]&.downcase
470
+ lang if TRANSLATION_LANGUAGES.include?(lang)
471
+ end
472
+
473
+ # @param lang [String, nil]
474
+ # @return [String] the ISO 15924 script of the language
475
+ def script_for(lang) = SCRIPTS.fetch(lang, "Latn")
476
+
477
+ # Merge every deferred translation into the English record of its file
478
+ # (relaton-cie#13), in catalogue order so the output does not depend on
479
+ # worker completion order. A file with no English record keeps its first
480
+ # translation as the record, so no document is lost.
481
+ def merge_translations
482
+ @translations.each { |file, list| merge_file_translations file, list }
483
+ end
484
+
485
+ # One file's merge. A failure is logged and stops only this file, so the
486
+ # crawl still saves its index (#fetch calls index.save next).
487
+ def merge_file_translations(file, list)
488
+ sorted = list.sort_by(&:first).map(&:last)
489
+ base = translation_base file, sorted
490
+ return if sorted.empty?
491
+
492
+ sorted.each { |trans| merge_translation base, trans }
493
+ File.write file, serialize(base), encoding: "UTF-8"
494
+ rescue StandardError => e
495
+ Util.error { "Translations of #{file}: #{e.message}" }
496
+ end
497
+
498
+ # The English record of `file`, or else the first translation in
499
+ # `sorted`, which is then written as the record and taken off the list.
500
+ def translation_base(file, sorted)
501
+ @items[file] || sorted.shift.tap { |item| write_file item }
502
+ end
503
+
504
+ # A translation is its own publication (its own ISBN, date and product
505
+ # page), with no CIE number of its own. So the English record keeps its
506
+ # identity, takes the translated title, and links the translation as a
507
+ # `hasTranslation` relation that keeps the translation's own metadata.
508
+ #
509
+ # @param base [Relaton::Cie::ItemData] the record that stays
510
+ # @param trans [Relaton::Cie::ItemData] its translation
511
+ def merge_translation(base, trans)
512
+ base.title = merge_list(base.title, trans.title,
513
+ :type, :content, :language)
514
+ rel = Bib::Relation.new(type: "hasTranslation",
515
+ bibitem: translation_bibitem(trans))
516
+ base.relation = merge_relation(base.relation, rel)
517
+ end
518
+
519
+ # Add `rel`, unless a relation of its type already links its source.
520
+ def merge_relation(list, rel)
521
+ (list + [rel]).uniq { |r| [r.type, r.bibitem&.source&.map(&:content)] }
522
+ end
523
+
524
+ # The translation as a relation's bibitem. Its abstract is left out: the
525
+ # catalogue shows only a placeholder text on a translation's page.
526
+ def translation_bibitem(trans)
527
+ ItemData.new(
528
+ type: trans.type, docidentifier: trans.docidentifier,
529
+ title: trans.title, date: trans.date, edition: trans.edition,
530
+ source: trans.source, language: trans.language, script: trans.script
531
+ )
532
+ end
533
+
534
+ # Concatenate two element lists, without an element that repeats `keys`.
535
+ def merge_list(list, other, *keys)
536
+ (list + other).uniq { |e| keys.map { |k| e.send(k) } }
537
+ end
538
+
419
539
  def to_xml(bib) = bib.to_xml(bibdata: true)
420
540
  def to_yaml(bib) = bib.to_yaml
421
541
  def to_bibxml(bib) = bib.to_rfcxml
@@ -431,17 +551,18 @@ module Relaton
431
551
  # rubocop:disable Metrics/AbcSize, Metrics/MethodLength
432
552
  def parse_page(hit, pos = nil, worker_agent = agent, pacing = nil)
433
553
  url = hit.at_xpath('.//h3/a')[:href]
554
+ lang = translation_language hit
434
555
  doc = time_req(pacing) { worker_agent.get url }
435
556
  @mutex.synchronize do
436
557
  item = ItemData.new(
437
558
  type: "standard", source: fetch_source(url), docnumber: fetch_docnumber(hit),
438
- docidentifier: fetch_docid(hit, doc), title: fetch_title(doc),
439
- abstract: fetch_abstract(doc), date: fetch_date(doc),
440
- edition: fetch_edition(doc), contributor: fetch_contributor(doc),
441
- relation: fetch_relation(doc), language: "en", script: "Latn",
442
- ext: fetch_ext
559
+ docidentifier: fetch_docid(hit, doc), title: fetch_title(doc, lang),
560
+ abstract: fetch_abstract(doc, lang || "en"),
561
+ date: fetch_date(doc), edition: fetch_edition(doc),
562
+ contributor: fetch_contributor(doc), relation: fetch_relation(doc),
563
+ language: lang || "en", script: script_for(lang), ext: fetch_ext
443
564
  )
444
- write_file item, pos
565
+ lang ? defer_translation(item, pos) : write_file(item, pos)
445
566
  end
446
567
  rescue StandardError => e
447
568
  Util.error do
@@ -450,8 +571,16 @@ module Relaton
450
571
  end
451
572
  # rubocop:enable Metrics/AbcSize, Metrics/MethodLength
452
573
 
574
+ # A translation has its English document's primary id: keep it for
575
+ # #merge_translations instead of letting it overwrite that record.
576
+ def defer_translation(item, pos)
577
+ file = output_file item.docidentifier[0].content
578
+ @translations[file] << [pos || 0, item]
579
+ end
580
+
453
581
  def fetch(_source = nil)
454
582
  process_hits collect_hits
583
+ merge_translations
455
584
  index.save
456
585
  report_errors
457
586
  ensure
@@ -243,9 +243,13 @@ module Relaton
243
243
  "also works). Default: author-date"
244
244
  option :output, aliases: :o, desc: "Output to the specified file"
245
245
 
246
- def convert(file) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
247
- xml = Nokogiri::XML(File.read(file, encoding: "UTF-8"))
248
- item = Relaton::Cli.parse_xml xml
246
+ def convert(file) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength,Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity
247
+ item = Relaton::Cli.parse_xml File.read(file, encoding: "UTF-8")
248
+ unless item
249
+ Util.warn "#{file} is not a readable Relaton XML document"
250
+ return
251
+ end
252
+
249
253
  result = if /yaml|yml/.match?(options[:format])
250
254
  item.to_yaml
251
255
  elsif options[:format] == "citation"
@@ -1,4 +1,4 @@
1
- require "nokogiri"
1
+ require "moxml"
2
2
  require "pathname"
3
3
 
4
4
  module Relaton
@@ -89,16 +89,6 @@ module Relaton
89
89
  )
90
90
  end
91
91
 
92
- # Turn an XML string into a Nokogiri XML document.
93
- #
94
- # @param document [String] XML
95
- # @param file [String, nil] path to file
96
- # @return [Nokogiri::XML::Document]
97
- def nokogiri_document(document, file = nil)
98
- document ||= File.read(file, encoding: "utf-8")
99
- Nokogiri.XML(document)
100
- end
101
-
102
92
  def select_source_files
103
93
  if File.file?(source)
104
94
  [source]
@@ -108,33 +98,35 @@ module Relaton
108
98
  end
109
99
 
110
100
  def relaton_collection
111
- @relaton_collection ||= Bibcollection.from_xml(nokogiri_document(nil, source))
101
+ @relaton_collection ||=
102
+ Bibcollection.from_xml(File.read(source, encoding: "utf-8"))
112
103
  end
113
104
 
114
- def extract_and_write_to_files # rubocop:disable Metrics/AbcSize, Metrics/MethodLength
105
+ def extract_and_write_to_files
115
106
  select_source_files.each do |file|
116
- xml = nokogiri_document(nil, file)
117
- xml.remove_namespaces!
118
-
119
- if (bib = xml.at("//bibdata"))
120
- bib = nokogiri_document(bib.to_xml)
121
- elsif (rfc = xml.at("//rfc"))
122
- require_relative "../ietf"
123
- require_relative "../ietf/bibxml_parser"
124
- ietf = Relaton::Ietf::BibXMLParser.parse_rfc rfc.to_xml
125
- bib = nokogiri_document(ietf.to_xml(bibdata: true))
126
- else
127
- next
128
- end
129
-
130
- bib.remove_namespaces!
131
-
132
- bibdata = Relaton::Bibdata.from_xml(bib.root)
133
- if bibdata
134
- build_bibdata_relaton(bibdata, file)
135
-
136
- write_to_file(bibdata.send(output_type), outdir, build_filename(file))
137
- end
107
+ xml = extract_bibdata_xml(File.read(file, encoding: "utf-8"), file)
108
+ bibdata = xml && Relaton::Bibdata.from_xml(xml)
109
+ next unless bibdata
110
+
111
+ build_bibdata_relaton(bibdata, file)
112
+ write_to_file(bibdata.send(output_type), outdir, build_filename(file))
113
+ end
114
+ end
115
+
116
+ # The bibdata XML of the first `bibdata` element, or of the first `rfc`
117
+ # element, in any namespace.
118
+ #
119
+ # @param content [String] XML
120
+ # @param file [String] path to file
121
+ # @return [String, nil]
122
+ def extract_bibdata_xml(content, file)
123
+ doc = parse_xml_document(content, file)
124
+ return unless doc
125
+
126
+ if (bib = find_element(doc, "bibdata"))
127
+ element_xml(bib)
128
+ elsif (rfc = find_element(doc, "rfc"))
129
+ rfc_bibdata_xml(element_xml(rfc))
138
130
  end
139
131
  end
140
132
 
@@ -142,21 +134,79 @@ require_relative "../ietf/bibxml_parser"
142
134
  #
143
135
  # @return [Array<Relaton::Bibdata>]
144
136
  def concatenate_files
145
- xml_files = [convert_rxl_to_xml, convert_yamls_to_xml, convert_xml_to_xml]
137
+ xml_files = [convert_rxl_to_xml, convert_yamls_to_xml,
138
+ convert_xml_to_xml]
146
139
 
147
140
  xml_files.flatten.reduce([]) do |mem, xml|
148
- doc = nokogiri_document(xml[:content])
149
- if (rfc = doc.at("/rfc"))
150
- require_relative "../ietf"
151
- require_relative "../ietf/bibxml_parser"
152
- ietf = Relaton::Ietf::BibXMLParser.parse_rfc rfc.to_xml
153
- d = nokogiri_document ietf.to_xml(bibdata: true)
154
- mem << bibdata_instance(d, xml[:file])
155
- elsif %w[bibitem bibdata].include? doc&.root&.name
156
- mem << bibdata_instance(doc, xml[:file])
157
- else mem
158
- end
141
+ bib = concatenate_bibdata_xml(xml[:content], xml[:file])
142
+ bib ? mem << bibdata_instance(bib, xml[:file]) : mem
143
+ end
144
+ end
145
+
146
+ # The bibdata XML of a document whose root is `rfc`, `bibitem` or
147
+ # `bibdata`, in any namespace.
148
+ #
149
+ # @param content [String] XML
150
+ # @param file [String] path to file
151
+ # @return [String, nil]
152
+ def concatenate_bibdata_xml(content, file)
153
+ root = parse_xml_document(content, file)&.root
154
+ case root&.name
155
+ when "rfc" then rfc_bibdata_xml(element_xml(root))
156
+ when "bibitem", "bibdata" then element_xml(root)
157
+ end
158
+ end
159
+
160
+ # Moxml does not recover a malformed document: warn
161
+ # and skip the file. An empty file is skipped without a warning.
162
+ #
163
+ # @param content [String] XML
164
+ # @param file [String] path to file
165
+ # @return [Moxml::Document, nil] nil when the content is not XML
166
+ def parse_xml_document(content, file)
167
+ Moxml.new.parse(content)
168
+ rescue Moxml::ParseError => e
169
+ unless content.to_s.strip.empty?
170
+ Util.warn "Skipping #{file}: not well-formed XML " \
171
+ "(#{e.message.lines.first&.strip})"
172
+ end
173
+ nil
174
+ end
175
+
176
+ # The root element when it has the name, else the first descendant
177
+ # element with that local name, in any namespace.
178
+ #
179
+ # @param doc [Moxml::Document]
180
+ # @param name [String]
181
+ # @return [Moxml::Element, nil]
182
+ def find_element(doc, name)
183
+ return doc.root if doc.root&.name == name
184
+
185
+ doc.xpath("//*[local-name()='#{name}']").first
186
+ end
187
+
188
+ # Serialize an element as a self-contained document: a sub-element
189
+ # inherits its namespace declarations from an ancestor, and `to_xml`
190
+ # does not write them.
191
+ #
192
+ # @param element [Moxml::Element]
193
+ # @return [String]
194
+ def element_xml(element)
195
+ declared = element.namespace_definitions.map(&:prefix)
196
+ element.in_scope_namespaces.each do |ns|
197
+ next if declared.include?(ns.prefix)
198
+
199
+ element.add_namespace(ns.prefix, ns.uri)
159
200
  end
201
+ element.to_xml
202
+ end
203
+
204
+ # @param rfc [String] RFC XML
205
+ # @return [String] bibdata XML
206
+ def rfc_bibdata_xml(rfc)
207
+ require_relative "../ietf"
208
+ require_relative "../ietf/bibxml_parser"
209
+ Relaton::Ietf::BibXMLParser.parse_rfc(rfc).to_xml(bibdata: true)
160
210
  end
161
211
 
162
212
  def concatenate_and_write_to_files
@@ -194,16 +244,15 @@ require_relative "../ietf/bibxml_parser"
194
244
  end
195
245
  end
196
246
 
197
- # Create a bibdata instance from the XML document,
247
+ # Create a bibdata instance from the XML,
198
248
  # while also adding file paths of related artifacts
199
249
  # as URI elements to the bibdata instance.
200
250
  #
201
- # @param document [Nokogiri::XML::Document]
251
+ # @param xml [String] bibdata XML
202
252
  # @param file [String] path to file
203
253
  # @return [Relaton::Bibdata]
204
- def bibdata_instance(document, file)
205
- document = clean_nokogiri_document(document)
206
- bibdata = Relaton::Bibdata.from_xml document.root
254
+ def bibdata_instance(xml, file)
255
+ bibdata = Relaton::Bibdata.from_xml xml
207
256
  build_bibdata_relaton(bibdata, file) if bibdata
208
257
 
209
258
  bibdata
@@ -223,18 +272,6 @@ require_relative "../ietf/bibxml_parser"
223
272
  end
224
273
  end
225
274
 
226
- # Force a namespace otherwise Nokogiri won't parse.
227
- # The reason is we use Bibcollection's from_xml, but that one
228
- # has an xmlns. We don't want to change the code for bibdata
229
- # hence this hack #bibdata_doc.root['xmlns'] = "xmlns"
230
- #
231
- # @param document [Nokogiri::XML::Document]
232
- # @return [Nokogiri::XML::Document]
233
- def clean_nokogiri_document(document)
234
- document.remove_namespaces!
235
- nokogiri_document(document.to_xml)
236
- end
237
-
238
275
  def convert_rxl_to_xml
239
276
  select_files_with("{rxl}").map do |file|
240
277
  { file: file, content: File.read(file, encoding: "utf-8") }
@@ -146,8 +146,8 @@ module Relaton
146
146
 
147
147
  def import(file)
148
148
  xml = read_xml file
149
- if xml.at "relaton-collection"
150
- imported = import_collection(xml)
149
+ if Moxml.new.parse(xml).root&.name == "relaton-collection"
150
+ imported = Relaton::Bibcollection.from_xml(xml)
151
151
  store.update(options[:collection]) do |coll|
152
152
  next imported unless coll # a new collection is the imported one
153
153
 
@@ -173,20 +173,10 @@ module Relaton
173
173
 
174
174
  private
175
175
 
176
- # Parse a namespace-free collection XML document.
177
- # Bibcollection.from_xml expects namespaced docs (via apply_namespace),
178
- # so we use plain XPath on the namespace-stripped document instead.
179
- def import_collection(xml)
180
- title = xml.at("relaton-collection/title")&.text
181
- author = xml.at_xpath(
182
- "./relaton-collection/contributor[role/@type='author']"\
183
- "/organization/name",
184
- )&.text
185
- items = xml.xpath("relaton-collection/relation").map do |rel|
186
- el = rel.at("bibdata") || rel.at("bibitem")
187
- Relaton::Bibdata.from_xml(el || rel)
188
- end.compact
189
- Relaton::Bibcollection.new(title: title, author: author, items: items)
176
+ # The XML file `import` reads: user input, not a collection.
177
+ # @return [String]
178
+ def read_xml(file)
179
+ File.read file, encoding: "UTF-8"
190
180
  end
191
181
 
192
182
  # @return [String]
@@ -210,14 +200,6 @@ module Relaton
210
200
  end
211
201
  end
212
202
 
213
- # The XML file `import` reads: user input, not a collection.
214
- # @return [Nokogiri::XML::Document]
215
- def read_xml(file)
216
- xml = Nokogiri::XML File.read(file, encoding: "UTF-8")
217
- xml.remove_namespaces!
218
- xml
219
- end
220
-
221
203
  # @return [Array<Hash>]
222
204
  def collections
223
205
  store.names(options.fetch(:collection, "*")).filter_map do |name|
@@ -1,4 +1,4 @@
1
- require "nokogiri"
1
+ require "moxml"
2
2
  require_relative "base_convertor"
3
3
 
4
4
  module Relaton
@@ -29,10 +29,11 @@ module Relaton
29
29
  "yaml"
30
30
  end
31
31
 
32
- # @param content [Nokogiri::XML::Document]
33
- # @return [Hash]
32
+ # @param content [String] Relaton XML: a bibliographic item or a
33
+ # collection
34
+ # @return [Relaton::Bibdata, Relaton::Bibcollection, nil]
34
35
  def convert_content(content)
35
- if %w[bibitem bibdata].include? content.root.name
36
+ if %w[bibitem bibdata].include? Moxml.new.parse(content).root&.name
36
37
  Bibdata.from_xml(content)
37
38
  else
38
39
  Bibcollection.from_xml(content)
@@ -40,7 +41,7 @@ module Relaton
40
41
  end
41
42
 
42
43
  def file_content
43
- Nokogiri::XML(File.read(file, encoding: "utf-8"))
44
+ File.read(file, encoding: "utf-8")
44
45
  end
45
46
  end
46
47
  end
@@ -1,4 +1,3 @@
1
- require "nokogiri"
2
1
  require "liquid"
3
2
 
4
3
  module Relaton::Cli
@@ -112,7 +111,7 @@ module Relaton::Cli
112
111
  # @param source [String] Relaton XML
113
112
  # @return [Relaton::Bibcollection]
114
113
  def build_bibcollection(source)
115
- Relaton::Bibcollection.from_xml(Nokogiri::XML(source))
114
+ Relaton::Bibcollection.from_xml(source)
116
115
  end
117
116
 
118
117
  # @param bibcollection [Relaton::Bibcollection]