Module: Dor::Describable

Extended by:
ActiveSupport::Concern
Included in:
AdminPolicyObject, Collection, Publishable, Set
Defined in:
lib/dor/models/describable.rb

Defined Under Namespace

Classes: CrosswalkError

Constant Summary collapse

DESC_MD_FORMATS = 
{
  "http://www.tei-c.org/ns/1.0" => 'tei',
  "http://www.loc.gov/mods/v3"  => 'mods'
}

Class Method Summary collapse

Instance Method Summary collapse

Class Method Details

.get_collection_title(obj) ⇒ Object 



254
255
256
257
258
259
260
261
262
263
264
 
# File 'lib/dor/models/describable.rb', line 254

def self.get_collection_title(obj)
  xml=obj.descMetadata.ng_xml
  title=''
  title_node = xml.at_xpath('//mods:mods/mods:titleInfo/mods:title','mods' => 'http://www.loc.gov/mods/v3')
  if title_node
    title = title_node.content
    subtitle=xml.at_xpath('//mods:mods/mods:titleInfo/mods:subTitle','mods' => 'http://www.loc.gov/mods/v3')
    title += " (#{subtitle.content})" if subtitle
  end
  title
end

Instance Method Details

#add_access_conditions(doc) ⇒ Object

Note:

this method modifies the passed in doc

Create MODS accessCondition statements from rightsMetadata

Parameters:



74
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
 
# File 'lib/dor/models/describable.rb', line 74

def add_access_conditions(doc)
  # clear out any existing accessConditions
  doc.xpath('//mods:accessCondition', 'mods' => 'http://www.loc.gov/mods/v3').each {|n| n.remove}
  rights = self.datastreams['rightsMetadata'].ng_xml

  rights.xpath('//use/human[@type="useAndReproduction"]').each do |use|
    txt = use.text.strip
    next if txt.empty?
    doc.root.element_children.last.add_next_sibling doc.create_element("accessCondition", txt, :type => 'useAndReproduction')
  end
  rights.xpath('//copyright/human[@type="copyright"]').each do |cr|
    txt = cr.text.strip
    next if txt.empty?
    doc.root.element_children.last.add_next_sibling doc.create_element("accessCondition", txt, :type => 'copyright')
  end
  rights.xpath("//use/machine[#{ci_compare('type', 'creativecommons')}]").each do |lic_type|
    next if lic_type.text =~ /none/i
    lic_text = rights.at_xpath("//use/human[#{ci_compare('type', 'creativecommons')}]").text.strip
    next if lic_text.empty?
    new_text = "CC #{lic_type.text}: #{lic_text}"
    doc.root.element_children.last.add_next_sibling doc.create_element("accessCondition", new_text, :type => 'license')
  end
  rights.xpath("//use/machine[#{ci_compare('type', 'opendatacommons')}]").each do |lic_type|
    next if lic_type.text =~ /none/i
    lic_text = rights.at_xpath("//use/human[#{ci_compare('type', 'opendatacommons')}]").text.strip
    next if lic_text.empty?
    new_text = "ODC #{lic_type.text}: #{lic_text}"
    doc.root.element_children.last.add_next_sibling doc.create_element("accessCondition", new_text, :type => 'license')
  end
end

#add_collection_reference(doc) ⇒ Object

Note:

this method modifies the passed in doc

returns the desc metadata a relatedItem with information about the collection this object belongs to for use in published mods and mods to DC conversion

Parameters:



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
134
135
136
137
138
139
140
141
142
143
 
# File 'lib/dor/models/describable.rb', line 108

def add_collection_reference(doc)
  return unless self.methods.include? :public_relationships
  collections=self.public_relationships.search('//rdf:RDF/rdf:Description/fedora:isMemberOfCollection',
                                   'fedora' => 'info:fedora/fedora-system:def/relations-external#',
                                   'rdf' => 'http://www.w3.org/1999/02/22-rdf-syntax-ns#' )
  return if collections.empty?

  # Remove any existing collections in the descMetadata
  doc.search('/mods:mods/mods:relatedItem[@type="host"]/mods:typeOfResource[@collection=\'yes\']', 'mods' => 'http://www.loc.gov/mods/v3').each do |node|
    node.parent.remove
  end

  collections.each do |collection_node|
    druid=collection_node['rdf:resource']
    druid=druid.gsub('info:fedora/','')
    collection_obj=Dor::Item.find(druid)
    collection_title = Dor::Describable.get_collection_title(collection_obj)
    related_item_node=Nokogiri::XML::Node.new('relatedItem',doc)
    related_item_node['type']='host'
    title_info_node = Nokogiri::XML::Node.new('titleInfo',doc)
    title_node      = Nokogiri::XML::Node.new('title',doc)
    title_node.content=collection_title

    id_node=Nokogiri::XML::Node.new('identifier',doc)
    id_node['type'] = 'uri'
    id_node.content = "http://#{Dor::Config.stacks.document_cache_host}/#{druid.split(':').last}"

    type_node=Nokogiri::XML::Node.new('typeOfResource',doc)
    type_node['collection'] = 'yes'
    doc.root.add_child(related_item_node)
    related_item_node.add_child(title_info_node)
    title_info_node.add_child(title_node)
    related_item_node.add_child(id_node)
    related_item_node.add_child(type_node)
  end
end

#add_identifier(type, value) ⇒ Object 



218
219
220
221
222
223
224
225
226
 
# File 'lib/dor/models/describable.rb', line 218

def add_identifier(type, value)
  ds_xml=self.descMetadata.ng_xml
  ds_xml.search('//mods:mods','mods' => 'http://www.loc.gov/mods/v3').each do |node|
    new_node=Nokogiri::XML::Node.new('identifier',ds_xml) #this ends up being mods:identifier without having to specify the namespace
    new_node['type']=type
    new_node.content=value
    node.add_child(new_node)
  end
end

#build_descMetadata_datastream(ds) ⇒ Object 



33
34
35
36
37
38
39
40
 
# File 'lib/dor/models/describable.rb', line 33

def build_descMetadata_datastream(ds)
  content = fetch_descMetadata_datastream
  return nil if content.nil?
  ds.dsLabel = 'Descriptive Metadata'
  ds.ng_xml = Nokogiri::XML(content)
  ds.ng_xml.normalize_text!
  ds.content = ds.ng_xml.to_xml
end

#delete_identifier(type, value = nil) ⇒ Object 



227
228
229
230
231
232
233
234
235
236
 
# File 'lib/dor/models/describable.rb', line 227

def delete_identifier(type,value=nil)
  ds_xml=self.descMetadata.ng_xml
  ds_xml.search('//mods:identifier','mods' => 'http://www.loc.gov/mods/v3').each do |node|
    if node.content == value || value.nil?
      node.remove
      return true
    end
  end
  return false
end

#fetch_descMetadata_datastream ⇒ Object 



27
28
29
30
31
 
# File 'lib/dor/models/describable.rb', line 27

def fetch_descMetadata_datastream
  candidates = self.datastreams['identityMetadata'].otherId.collect { |oid| oid.to_s }
  metadata_id = Dor::MetadataService.resolvable(candidates).first
  return metadata_id.nil? ? nil : Dor::MetadataService.fetch(metadata_id.to_s)
end

#generate_dublin_core ⇒ Object

Generates Dublin Core from the MODS in the descMetadata datastream using the LoC mods2dc stylesheet

Should not be used for the Fedora DC datastream

Raises:

  • (Exception) —

    Raises an Exception if the generated DC is empty or has no children



45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
 
# File 'lib/dor/models/describable.rb', line 45

def generate_dublin_core
  format = self.metadata_format
  if format.nil?
    raise CrosswalkError, "Unknown descMetadata namespace: #{metadata_namespace.inspect}"
  end
  xslt = Nokogiri::XSLT(File.new(File.expand_path(File.dirname(__FILE__) + "/#{format}2dc.xslt")) )
  desc_md = self.descMetadata.ng_xml.dup(1)
  self.add_collection_reference(desc_md)
  dc_doc = xslt.transform(desc_md)
  # Remove empty nodes
  dc_doc.xpath('/oai_dc:dc/*[count(text()) = 0]').remove
  if dc_doc.root.nil? || dc_doc.root.children.size == 0
    raise CrosswalkError, "Dor::Item#generate_dublin_core produced incorrect xml:\n#{dc_doc.to_xml}"
  end
  dc_doc
end

#generate_public_desc_md ⇒ Object 



62
63
64
65
66
67
68
69
 
# File 'lib/dor/models/describable.rb', line 62

def generate_public_desc_md
  doc = self.descMetadata.ng_xml.dup(1)
  add_collection_reference(doc)
  add_access_conditions(doc)
  new_doc = Nokogiri::XML(doc.to_xml) { |x| x.noblanks }
  new_doc.encoding = 'UTF-8'
  new_doc.to_xml
end

#metadata_format ⇒ Object 



153
154
155
 
# File 'lib/dor/models/describable.rb', line 153

def metadata_format
  DESC_MD_FORMATS[metadata_namespace]
end

#metadata_namespace ⇒ Object 



144
145
146
147
148
149
150
151
 
# File 'lib/dor/models/describable.rb', line 144

def metadata_namespace
  desc_md = self.datastreams['descMetadata'].ng_xml
  if desc_md.nil? || desc_md.root.nil? || desc_md.root.namespace.nil?
    return nil
  else
    return desc_md.root.namespace.href
  end
end

#set_desc_metadata_using_label(force = false) ⇒ Object 



238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
 
# File 'lib/dor/models/describable.rb', line 238

def set_desc_metadata_using_label(force=false)
  ds=self.descMetadata
  unless force || ds.new?
    raise 'Cannot proceed, there is already content in the descriptive metadata datastream: '+ds.content.to_s
  end
  label=self.label
  builder = Nokogiri::XML::Builder.new { |xml|
    xml.mods( 'xmlns' => 'http://www.loc.gov/mods/v3', 'xmlns:xsi' => 'http://www.w3.org/2001/XMLSchema-instance',:version => '3.3', "xsi:schemaLocation" => 'http://www.loc.gov/mods/v3 http://www.loc.gov/standards/mods/v3/mods-3-3.xsd'){
      xml.titleInfo{
        xml.title label
      }
    }
  }
  self.descMetadata.content=builder.to_xml
end

#stanford_mods(content = nil, ns_aware = true) ⇒ Object

intended for read-access, “as SearchWorks would see it”, mostly for to_solr()

Parameters:

  • content (Nokogiri::XML::Document) (defaults to: nil) —

    Nokogiri descMetadata document (overriding internal data)

  • ns_aware (boolean) (defaults to: true) —

    namespace awareness toggle for from_nk_node()



20
21
22
23
24
25
 
# File 'lib/dor/models/describable.rb', line 20

def stanford_mods(content=nil, ns_aware=true)
  m = Stanford::Mods::Record.new
  desc = content.nil? ? self.descMetadata.ng_xml : content
  m.from_nk_node(desc.root, ns_aware)
  m
end

#to_solr(solr_doc = Hash.new, *args) ⇒ Object 



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
 
# File 'lib/dor/models/describable.rb', line 157

def to_solr(solr_doc=Hash.new, *args)
  super solr_doc, *args
  mods_sources = {
    'sw_language_ssim'           => :sw_language_facet,
    'sw_language_tesim'           => :sw_language_facet,
    'sw_genre_ssim'               => :sw_genre,
    'sw_genre_tesim'              => :sw_genre,
    'sw_format_ssim'              => :format_main,   # basically sw_typeOfResource_ssim
    'sw_format_tesim'             => :format_main,   # basically sw_typeOfResource_tesim
    'sw_subject_temporal_ssim'    => :era_facet,
    'sw_subject_temporal_tesim'   => :era_facet,
    'sw_subject_geographic_ssim'  => :geographic_facet,
    'sw_subject_geographic_tesim' => :geographic_facet,
    'mods_typeOfResource_ssim'    => [:term_values, :typeOfResource],
    'mods_typeOfResource_tesim'   => [:term_values, :typeOfResource]
  }
  keys = mods_sources.keys.concat(%w[ metadata_format_ssim ])
  keys.each { |key|
    solr_doc[key] ||= []     # initialize multivalue targts if necessary
  }

  solr_doc["metadata_format_ssim"] << self.metadata_format
  begin
    dc_doc = self.generate_dublin_core
    dc_doc.xpath('/oai_dc:dc/*').each do |node|
      add_solr_value(solr_doc, "public_dc_#{node.name}", node.text, :string, [:stored_searchable])
    end
    creator=''
    dc_doc.xpath('//dc:creator').each do |node|
      creator=node.text
    end
    title=''
    dc_doc.xpath('//dc:title').each do |node|
      title=node.text
    end
    creator_title=creator+title
    add_solr_value(solr_doc, 'creator_title', creator_title , :string, [:stored_sortable])
  rescue CrosswalkError => e
    ActiveFedora.logger.warn "Cannot index #{self.pid}.descMetadata: #{e.message}"
  end

  begin
    mods = self.stanford_mods
    mods_sources.each_pair do |solr_key, meth|
      vals = meth.is_a?(Array) ? mods.send(meth.shift, *meth) : mods.send(meth)
      solr_doc[solr_key].push *vals unless vals.nil? || vals.empty?
      # asterisk to avoid multi-dimensional array: push values, not the array
    end
    solr_doc['sw_pub_date_sort_ssi' ] = mods.pub_date_sort  # e.g. '0800'
    solr_doc['sw_pub_date_facet_ssi'] = mods.pub_date_facet # e.g. '9th century'
  end
  # some fields get explicit "(none)" placeholder values, mostly for faceting
  %w[sw_language_tesim sw_genre_tesim sw_format_tesim].each { |key| solr_doc[key] = ['(none)'] if solr_doc[key].empty? }
  # otherwise remove empties
  keys.each{ |key| solr_doc.delete(key) if solr_doc[key].nil? || solr_doc[key].empty?}
  solr_doc
end

#update_title(new_title) ⇒ Object 



215
216
217
 
# File 'lib/dor/models/describable.rb', line 215

def update_title(new_title)
  raise 'Descriptive metadata has no title to update!' unless update_simple_field('mods:mods/mods:titleInfo/mods:title', new_title)
end