rof 1.0.4 → 1.0.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 33761b71b6e080540bbab28fa0516ae9f327b496
-  data.tar.gz: d8dc2c60c38574a5a872622f5a0e7bdfdf7ededd
+  metadata.gz: f7f604446f5b61795e262bc554d46203d6ab3e00
+  data.tar.gz: 433b2a744b3600fe9dbd15536d95daeb2796f9f1
 SHA512:
-  metadata.gz: baa4c3529a273e6de117e1070fb7fde0eb70a50c398238b95243eae326ea34a9ba4fd2200c59b11b332b93e061cddb1f2947d9d56de0c5d3794141f875665f51
-  data.tar.gz: 9e78181e99a5516a1389b8895a74c51b20a790f1e3fa7fa38563d1e5ac2cf9e98370457d8d2255be78cb792cbcc2db60e4f502a47cf70efee9bfbc8e8e6ad43b
+  metadata.gz: 4deb205a5a32c0a3c9765bd6e949a8e29063a23e6eb6f366012f71c75636cf0c26a3c298371651d1b646dadc4ef1002d64053c0b5345fd8574e917772d13d3bd
+  data.tar.gz: 173692dabcdee2fbcb720cbdc7a991ddb3d16cd39aa3e681e9370450731e9a51c6f4a093e75df9f59f30ae130dbcd27ad66a65bde412751a3f1fa6e7d9acb253

data/lib/rof/get_from_fedora.rb

@@ -47,26 +47,30 @@ module ROF
           meta = create_meta(ds, config)
           @fedora_info["#{dsname}-meta"] = meta unless meta.empty?
 
-          # TODO(dbrower): change dump algorithm:
-          # if content is short < X bytes, save as string
-          # if content is > X bytes, save as file only if config option is given
+          # if content is short < X bytes and valid utf-8, save as string
+          # if content is > X bytes or is not utf-8, save as file only if config option is given
           content = ds.datastream_content
-          # NOTE- Entire datastream being downloaded every time.
-          content_string = content.to_s.force_encoding('UTF-8')
-          if (content.length <= 1024 || config['inline']) && content_string.valid_encoding?
-            @fedora_info[dsname] = content_string
-          elsif config['download']
-            fname = "#{@fedora_info['pid']}-#{dsname}"
-            abspath = File.join(config['download_path'], fname)
-            @fedora_info["#{dsname}-file"] = fname
-            if File.file?(config['download_path'])
-              puts "Error: --download directory #{config['download_path']} specified is an existing file."
-              exit 1
-            end
-            FileUtils.mkdir_p(config['download_path'])
-            File.open(abspath, 'w') do |f|
-              f.write(content)
+          if content.length <= 1024 || config['inline']
+            # this downloads the contents of the datastream into memory
+            content_string = content.to_s.force_encoding('UTF-8')
+            if content_string.valid_encoding?
+              @fedora_info[dsname] = content_string
+              next # we're done! move on to next datastream
             end
+            # not utf-8, so keep going and see if download option was given
+          end
+          next unless config['download']
+          # download option was given, so save this datastream as a file
+          fname = "#{@fedora_info['pid']}-#{dsname}"
+          abspath = File.join(config['download_path'], fname)
+          @fedora_info["#{dsname}-file"] = fname
+          if File.file?(config['download_path'])
+            puts "Error: --download directory #{config['download_path']} specified is an existing file."
+            exit 1
+          end
+          FileUtils.mkdir_p(config['download_path'])
+          File.open(abspath, 'w') do |f|
+            f.write(content)
           end
         end
       end
@@ -127,6 +131,10 @@ module ROF
 
       root = xml_doc.root
 
+      # check for optional embargo date - set if present
+      this_embargo = root.elements['embargo']
+      rights_array['embargo-date'] = this_embargo.elements['machine'].elements['date'][0] if has_embargo_date(this_embargo)
+
       %w(read edit).each do |access|
         this_access = root.elements["//access[@type=\'#{access}\']"]
 
@@ -152,6 +160,13 @@ module ROF
       @fedora_info['rights'] = rights_array
     end
 
+    # test for embargo xml cases
+    def self.has_embargo_date(embargo_xml)
+      return false if embargo_xml == '' || embargo_xml.nil?
+      return false unless embargo_xml.elements['machine'].has_elements? && embargo_xml.elements['machine'].elements['date'].has_text?
+      true
+    end
+
     def self.RELSEXT(ds, _config)
       # RELS-EXT is RDF-XML - parse it
       ctx = ROF::RelsExtRefContext.dup

data/lib/rof/osf_to_rof.rb

@@ -34,8 +34,8 @@ module ROF
     def self.ttl_from_targz(config, this_project, ttl_filename)
       id =  this_project['project_identifier']
       ttl_path = File.join(id,
-                      'data/obj/root',
-                      ttl_filename)
+                           'data/obj/root',
+                           ttl_filename)
       ROF::Utility.file_from_targz(File.join(config['package_dir'], id + '.tar.gz'),
                                    ttl_path)
       ttl_data = fetch_from_ttl(File.join(config['package_dir'], ttl_path))
@@ -55,13 +55,13 @@ module ROF
       metadata = {}
       metadata['@context'] = ROF::RdfContext.dup
       # metdata derived from project ttl file
-      metadata['dc:created'] = Time.iso8601(ttl_data[0][@osf_map['dc:created']][0]['@value']).to_date.iso8601 + "Z"
+      metadata['dc:created'] = Time.iso8601(ttl_data[0][@osf_map['dc:created']][0]['@value']).to_date.iso8601 + 'Z'
       metadata['dc:title'] = ttl_data[0][@osf_map['dc:title']][0]['@value']
       metadata['dc:description'] =
         ttl_data[0][@osf_map['dc:description']][0]['@value']
       metadata['dc:subject'] = map_subject(ttl_data[0])
       # metadata derived from osf_projects data, passed from UI
-      metadata['dc:source'] = "https://osf.io/" +project['project_identifier']
+      metadata['dc:source'] = 'https://osf.io/' + project['project_identifier']
       metadata['dc:creator#adminstrative_unit'] = project['administrative_unit']
       metadata['dc:creator#affiliation'] = project['affiliation']
       metadata['dc:creator'] = map_creator(config, project, ttl_data)
@@ -101,13 +101,14 @@ module ROF
     # sets the creator- needs to read another ttl for the User data
     # only contrubutors with isBibliographic true are considered
     def self.map_creator(config, project, ttl_data)
-      creator = ''
-      contributor = ttl_data[0][@osf_map['hasContributor']][0]['@id']
-      ttl_data.each do |item|
-        next unless item['@id'] == contributor
-        if item[@osf_map['isBibliographic']][0]['@value'] == 'true'
-          creator = map_user_from_ttl(config, project,
-                                      item[@osf_map['hasUser']][0]['@id'])
+      creator = []
+      ttl_data[0][@osf_map['hasContributor']].each do |contributor|
+        ttl_data.each do |item|
+          next unless item['@id'] == contributor['@id']
+          if item[@osf_map['isBibliographic']][0]['@value'] == 'true'
+            creator.push map_user_from_ttl(config, project,
+                                           item[@osf_map['hasUser']][0]['@id'])
+          end
         end
       end
       creator

data/lib/rof/rdf_context.rb

@@ -4,9 +4,12 @@ module ROF
     'dc' => 'http://purl.org/dc/terms/',
     'ebucore' => 'http://www.ebu.ch/metadata/ontologies/ebucore/ebucore#',
     'foaf' => 'http://xmlns.com/foaf/0.1/',
+    'hydramata-rel' => 'http://projecthydra.org/ns/relations#',
     'mrel' => 'http://id.loc.gov/vocabulary/relators/',
+    'ms' => 'http://www.ndltd.org/standards/metadata/etdms/1.1/',
     'nd' => 'https://library.nd.edu/ns/terms/',
     'rdfs' => 'http://www.w3.org/2000/01/rdf-schema#',
+    'ths' => 'http://id.loc.gov/vocabulary/relators/',
     'vracore' => 'http://purl.org/vra/',
 
     'dc:dateSubmitted' => {

data/lib/rof/version.rb

@@ -1,3 +1,3 @@
 module ROF
-  VERSION="1.0.4"
+  VERSION="1.0.7"
 end

data/rof.gemspec

@@ -25,13 +25,14 @@ Gem::Specification.new do |spec|
   spec.add_dependency "rdf-turtle"
   spec.add_dependency "rdf-isomorphic"
   spec.add_dependency "json-ld", "~> 2.0.0"
-  spec.add_dependency "mime-types", "~> 2.4.3"
+  spec.add_dependency "mime-types", "~> 2.4"
   spec.add_dependency "rubydora", "~> 1.8.1"
   spec.add_dependency "noids_client"
   spec.add_dependency 'deprecation', '~> 0.1'
   # only needed because we use ruby < 2.2.2 in production and that doesn't play
   # nice with rails 5
   spec.add_dependency 'activesupport', '< 5.0'
+  spec.add_dependency 'ebnf', '< 1.0.2' 
 
   spec.add_development_dependency "bundler", "~> 1.3"
   spec.add_development_dependency "rake"

data/spec/lib/rof/cli_spec.rb

@@ -32,11 +32,14 @@ describe ROF::CLI do
 	 "@context"=> {
            "bibo"=>"http://purl.org/ontology/bibo/",
            "dc"=>"http://purl.org/dc/terms/",
+           "ms" => 'http://www.ndltd.org/standards/metadata/etdms/1.1/',
            "ebucore"=>"http://www.ebu.ch/metadata/ontologies/ebucore/ebucore#",
            "foaf"=>"http://xmlns.com/foaf/0.1/",
+           'hydramata-rel' => 'http://projecthydra.org/ns/relations#',
            "mrel"=>"http://id.loc.gov/vocabulary/relators/",
 	   "nd"=>"https://library.nd.edu/ns/terms/",
            "rdfs"=>"http://www.w3.org/2000/01/rdf-schema#",
+           'ths' => 'http://id.loc.gov/vocabulary/relators/',
            "vracore"=>"http://purl.org/vra/",
            "dc:dateSubmitted" => {"@type" => "http://www.w3.org/2001/XMLSchema#date"},
            "dc:created"=>{"@type"=>"http://www.w3.org/2001/XMLSchema#date"},

data/spec/lib/rof/get_from_fedora_spec.rb

@@ -0,0 +1,22 @@
+require 'spec_helper'
+
+RSpec.describe ROF::FedoraToRof do
+  it 'handles embargo presence or absence' do
+
+    rights_tests = [
+      ['<embargo> <human/> <machine> <date>2017-08-01</date> </machine> </embargo>', true],
+      ['<embargo> <human/> <machine> <date></date> </machine> </embargo>', false],
+      ['<embargo> <human/> <machine/> </embargo>', false]
+    ]
+
+    begin
+
+      rights_tests.each do |this_test|
+        xml_doc = REXML::Document.new(this_test[0])
+        root = xml_doc.root
+        rights = ROF::FedoraToRof.has_embargo_date(root)
+        expect(rights).to eq(this_test[1])
+      end
+    end
+  end
+end

data/spec/lib/rof/osf_to_rof_spec.rb

@@ -38,9 +38,12 @@ RSpec.describe ROF::OsfToRof do
                                                  "dc"=>"http://purl.org/dc/terms/",
                                                  "ebucore"=>"http://www.ebu.ch/metadata/ontologies/ebucore/ebucore#",
                                                  "foaf"=>"http://xmlns.com/foaf/0.1/",
+                                                 'hydramata-rel' => 'http://projecthydra.org/ns/relations#',
                                                  "mrel"=>"http://id.loc.gov/vocabulary/relators/",
+                                                 'ms' => 'http://www.ndltd.org/standards/metadata/etdms/1.1/',
                                                  "nd"=>"https://library.nd.edu/ns/terms/",
                                                  "rdfs"=>"http://www.w3.org/2000/01/rdf-schema#",
+                                                 'ths' => 'http://id.loc.gov/vocabulary/relators/',
                                                  "vracore"=>"http://purl.org/vra/",
                                                  "dc:dateSubmitted"=>{"@type"=>"http://www.w3.org/2001/XMLSchema#date"},
                                                  "dc:created"=>{"@type"=>"http://www.w3.org/2001/XMLSchema#date"},
@@ -52,7 +55,7 @@ RSpec.describe ROF::OsfToRof do
                                                  "dc:source"=>"https://osf.io/b6psa",
                                                  "dc:creator#adminstrative_unit"=>"Library",
                                                  "dc:creator#affiliation"=>"OddFellows Local 151",
-                                                 "dc:creator"=>"Mark Suhovecky"},
+                                                 "dc:creator"=>["Mark Suhovecky"]},
                      "files"=>["b6psa.tar.gz"]}]
 
     FileUtils.cp('spec/fixtures/osf/b6psa.tar.gz', tar_file)

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rof
 version: !ruby/object:Gem::Version
-  version: 1.0.4
+  version: 1.0.7
 platform: ruby
 authors:
 - Jeremy Friesen
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-10-25 00:00:00.000000000 Z
+date: 2016-12-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rdf
@@ -86,14 +86,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.4.3
+        version: '2.4'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.4.3
+        version: '2.4'
 - !ruby/object:Gem::Dependency
   name: rubydora
   requirement: !ruby/object:Gem::Requirement
@@ -150,6 +150,20 @@ dependencies:
     - - "<"
       - !ruby/object:Gem::Version
         version: '5.0'
+- !ruby/object:Gem::Dependency
+  name: ebnf
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "<"
+      - !ruby/object:Gem::Version
+        version: 1.0.2
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "<"
+      - !ruby/object:Gem::Version
+        version: 1.0.2
 - !ruby/object:Gem::Dependency
   name: bundler
   requirement: !ruby/object:Gem::Requirement
@@ -310,6 +324,7 @@ files:
 - spec/lib/rof/filters/file_to_url_spec.rb
 - spec/lib/rof/filters/label_spec.rb
 - spec/lib/rof/filters/work_spec.rb
+- spec/lib/rof/get_from_fedora_spec.rb
 - spec/lib/rof/ingest_spec.rb
 - spec/lib/rof/ingesters/rels_ext_ingester_spec.rb
 - spec/lib/rof/ingesters/rights_metadata_ingester_spec.rb
@@ -357,6 +372,7 @@ test_files:
 - spec/lib/rof/filters/file_to_url_spec.rb
 - spec/lib/rof/filters/label_spec.rb
 - spec/lib/rof/filters/work_spec.rb
+- spec/lib/rof/get_from_fedora_spec.rb
 - spec/lib/rof/ingest_spec.rb
 - spec/lib/rof/ingesters/rels_ext_ingester_spec.rb
 - spec/lib/rof/ingesters/rights_metadata_ingester_spec.rb