dwc-archive 0.9.6 → 0.9.10

data/spec/lib/metadata_spec.rb

@@ -0,0 +1,80 @@
+require_relative '../spec_helper'
+
+describe DarwinCore::Metadata do
+  subject(:dwca) { DarwinCore.new(file_path) }
+  subject(:eml) { DarwinCore::Metadata.new(dwca.archive) }
+  let(:file_path) { File.join(File.expand_path('../../files', __FILE__), 
+                              file_name) }
+  let(:file_name) { 'data.tar.gz' }
+  
+  describe '.new' do
+    it 'initializes' do
+      expect(eml).to be_kind_of DarwinCore::Metadata
+    end
+  end
+
+  describe '#data' do
+    it 'returns hash of metadata' do
+      expect(eml.data).to be_kind_of Hash
+    end
+  end
+
+  describe '#id' do
+    it 'returns id of the archive if it exists' do
+      expect(eml.id).to eq 'leptogastrinae:version:2.5'
+    end
+  end
+
+  describe '#package_id' do
+    context 'no package id is given' do
+      it 'returns id of the DwCA file if exists' do
+        expect(eml.package_id).to be_nil
+      end
+    end
+  end
+
+  describe '#title' do
+    it 'returns name of the archive' do
+      expect(eml.title).
+        to eq 'Leptogastrinae (Diptera: Asilidae) Classification'
+    end
+  end
+
+  describe '#authors' do
+    it 'returns authors of the archive' do
+      expect(eml.authors).
+        to eq [
+          {:first_name=>"Keith", 
+           :last_name=>"Bayless", 
+           :email=>"keith.bayless@gmail.com"}, 
+          {:first_name=>"Torsten", 
+            :last_name=>"Dikow", 
+            :email=>"dshorthouse@eol.org"}]
+    end
+
+    describe '#abstract' do
+      it 'returns abstract of an article' do
+        expect(eml.abstract).
+          to eq 'These are all the names in the Leptogastrinae classification.'
+      end
+    end
+
+    describe '#citation' do
+      it 'returns citation of the archive' do
+        expect(eml.citation).
+          to eq 'Dikow, Torsten. 2010. The Leptogastrinae classification.'
+      end
+    end
+
+    describe '#url' do
+      it 'returns url to the archive' do
+        expect(eml.url).to eq 'http://leptogastrinae.lifedesks.org'\
+                              '/files/leptogastrinae/'\
+                              'classification_export/shared/'\
+                              'leptogastrinae.tar.gz'
+      end
+    end
+
+  end
+end
+

data/spec/lib/taxon_normalized_spec.rb

@@ -0,0 +1,145 @@
+require_relative '../spec_helper'
+# encoding: utf-8
+
+describe DarwinCore::TaxonNormalized do
+  subject(:dwca) { DarwinCore.new(file_path) }
+  subject(:normalizer) { DarwinCore::ClassificationNormalizer.new(dwca) }
+  let(:file_dir) { File.expand_path('../../files', __FILE__) }
+
+  context 'typical case' do
+    let(:file_path) { File.join(file_dir, 'data.tar.gz') }
+    let(:normalized) { normalizer.normalize }
+    let(:tn) { normalized['leptogastrinae:tid:2681'] }
+
+    it 'has TaxonNormalized type' do
+      expect(tn).to be_kind_of DarwinCore::TaxonNormalized
+    end
+
+    describe '#id' do
+      it 'returns taxon_id' do
+        expect(tn.id).to eq 'leptogastrinae:tid:2681'
+      end
+    end
+
+    describe '#local_id' do
+      it 'returns local id' do
+        expect(tn.id).to eq 'leptogastrinae:tid:2681'
+      end
+    end
+
+    describe '#global_id' do
+      it 'returns global id' do
+        expect(tn.global_id).to eq '843ff0df-4bcd-4aad-ba94-643f8695a292'
+      end
+    end
+
+    describe '#source' do
+      it 'returns url to the data' do
+        expect(tn.source).
+          to eq 'http://leptogastrinae.lifedesks.org/pages/2681'
+      end
+    end 
+
+    describe '#parent_id' do
+      it 'returns ancestor\'s id' do
+        expect(tn.parent_id).to eq 'leptogastrinae:tid:2584'
+      end
+    end
+      
+    describe '#classification_path_id' do
+      it 'returns an array' do
+        expect(tn.classification_path_id).to be_kind_of Array
+      end
+
+      it 'returns ids of classification path elements' do
+        expect(tn.classification_path_id).
+          to eq ['leptogastrinae:tid:42', 
+                 'leptogastrinae:tid:2045', 
+                 'leptogastrinae:tid:2584', 
+                 'leptogastrinae:tid:2681']
+      end
+    end
+
+    describe '#classification_path' do
+      let(:cp) { tn.classification_path }
+      it 'returns array of name strings of the classification path' do
+        expect(cp).to be_kind_of Array
+        expect(cp).to eq ["Leptogastrinae", 
+                          "Leptogastrini", 
+                          "Leptogaster", 
+                          "Leptogaster flavipes"] 
+      end
+    end
+
+    describe '#linnean_classification_path' do
+      it 'returns empty array for parent/child based classification' do
+        expect(tn.linnean_classification_path).to be_kind_of Array
+        expect(tn.linnean_classification_path).to be_empty
+      end
+    end
+
+    describe '#current_name' do
+      it 'returns name marked as currently valid for the taxon' do
+        expect(tn.current_name).to eq 'Leptogaster flavipes Loew, 1862'
+      end
+    end
+    
+    describe '#current_name_canonical' do
+      it 'returns canonical form of valid name' do
+        expect(tn.current_name_canonical).to eq 'Leptogaster flavipes'
+      end
+    end
+
+    describe '#synonyms' do
+      context 'synonyms not found' do 
+
+        it 'returns empty array' do
+          expect(tn.synonyms).to be_kind_of Array
+          synonym = tn.synonyms[0]
+          expect(synonym).to be_kind_of DarwinCore::SynonymNormalized
+          expect(synonym).to be_kind_of Struct
+        end
+      end
+
+      context 'synonyms found' do
+        let(:tn) { normalized['leptogastrinae:tid:2858'] }
+        it 'returns array of synonyms' do
+          expect(tn.synonyms).to be_kind_of Array
+          expect(tn.synonyms).to be_empty
+        end
+      end
+    end
+
+    describe '#vernacular_names' do
+      context 'vernacular names not found' do
+        it 'returns empty array' do
+          expect(tn.vernacular_names).to be_kind_of Array
+          expect(tn.vernacular_names).to be_empty
+        end
+      end
+
+      context 'vernacular names found' do
+        let(:tn) { normalized['leptogastrinae:tid:42'] }
+        it 'returns array with vernacular name structure' do
+          expect(tn.vernacular_names).to be_kind_of Array
+          vn = tn.vernacular_names[0]
+          expect(vn).to be_kind_of DarwinCore::VernacularNormalized
+          expect(vn).to be_kind_of Struct
+        end
+      end
+    end
+
+    describe '#rank' do
+      it 'returns rank of the taxon' do
+        expect(tn.rank).to eq 'species'
+      end
+    end
+
+    describe '#status' do
+      it 'returns status of taxon' do
+        expect(tn.status).to be_nil
+      end
+    end
+  end
+
+end

data/spec/lib/xml_reader_spec.rb

@@ -1,15 +1,18 @@
-require File.expand_path(File.dirname(__FILE__) + "/../spec_helper")
+require_relative '../spec_helper'
 
-describe "DarwinCore::XmlReader" do 
-  it "should parse xml to hash" do
-    DarwinCore::XmlReader.public_methods.map { |i| i.to_s }.include?("from_xml").should be_true
-  end
+describe DarwinCore::XmlReader do 
+  
+  describe '.from_xml' do
+
+    it "reads xml to hash" do
+      file = File.expand_path('../../files/meta.xml', __FILE__)
+      meta = subject.from_xml(File.read(file))
+      expect(meta.is_a? Hash).to be_true
+      expect(meta[:archive].keys.sort).to eq [:core, :extension]
+      expect(meta[:archive][:core].keys.sort).
+        to eq [:attributes, :field, :files, :id]
+    end
 
-  it "should parse xml" do
-    xml_string = open(File.dirname(__FILE__) + "/../files/meta.xml").read
-    meta = DarwinCore::XmlReader.from_xml(xml_string)
-    meta[:archive].keys.map {|k| k.to_s}.sort.should == %w(core extension)
-    meta[:archive][:core].keys.map{|k| k.to_s}.sort.should == ["attributes", "field", "files", "id"]
   end
 
 end

data/spec/spec_helper.rb

@@ -1,9 +1,78 @@
-$LOAD_PATH.unshift(File.dirname(__FILE__))
-$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
+# encoding: utf-8
+# HACK: to suppress warnings
+$VERBOSE = nil
+
+require 'coveralls'
+Coveralls.wear!
+
 require 'dwc-archive'
 require 'rspec'
-require 'rspec/autorun'
+require 'rspec/mocks'
+require 'socket'
 
 RSpec.configure do |config|
+end
 
+unless defined?(SPEC_CONSTANTS)
+  EML_DATA = {
+      id: '1234',
+      license: 'http://creativecommons.org/licenses/by-sa/3.0/',
+      title: 'Test Classification',
+      authors: [
+        { first_name: 'John',
+          last_name: 'Doe',
+          email: 'jdoe@example.com',
+          organization: 'Example',
+          position: 'Assistant Professor',
+          url: 'http://example.org' },
+          { first_name: 'Jane',
+            last_name: 'Doe',
+            email: 'jane@example.com' }
+    ],
+      metadata_providers: [
+        { first_name: 'Jim',
+          last_name: 'Doe',
+          email: 'jimdoe@example.com',
+          url: 'http://aggregator.example.org' }],
+      abstract: 'test classification',
+      citation:
+        'Test classification: Doe John, Doe Jane, Taxnonmy, 10, 1, 2010',
+      url: 'http://example.com'
+    }
+  META_DATA = {
+    extensions: [
+      {fields: [
+        'http://rs.tdwg.org/dwc/terms/TaxonID', 
+        'http://rs.tdwg.org/dwc/terms/vernacularName'], 
+        ignoreHeaderLines: 1, 
+        location: 'vern.csv', 
+        rowType: 'http://rs.gbif.org/terms/1.0/VernacularName'}], 
+        core: {
+          fields: [
+            'http://rs.tdwg.org/dwc/terms/taxonID', 
+            'http://rs.tdwg.org/dwc/terms/parentNameUsageID', 
+            'http://rs.tdwg.org/dwc/terms/scientificName', 
+            'http://rs.tdwg.org/dwc/terms/taxonRank'], 
+            ignoreHeaderLines: 1, 
+            location: 'core.csv'}}
+  CORE_DATA = [
+    ["http://rs.tdwg.org/dwc/terms/taxonID",
+     "http://rs.tdwg.org/dwc/terms/parentNameUsageID",
+     "http://rs.tdwg.org/dwc/terms/scientificName",
+     "http://rs.tdwg.org/dwc/terms/taxonRank"],
+    [1, 0, "Plantae", "kingdom"],
+    [2, 1, "Betula", "genus"],
+    [3, 2, "Betula verucosa", "species"]
+  ] 
+  EXTENSION_DATA = [
+    ["http://rs.tdwg.org/dwc/terms/TaxonID",
+    "http://rs.tdwg.org/dwc/terms/vernacularName"],
+    [1, "Plants"],
+    [1, "Растения"],
+    [2, "Birch"],
+    [2, "Береза"],
+    [3, "Wheeping Birch"],
+    [3, "Береза плакучая"]
+  ]
+  SPEC_CONSTANTS = true
 end

metadata

@@ -1,52 +1,46 @@
 --- !ruby/object:Gem::Specification
 name: dwc-archive
 version: !ruby/object:Gem::Version
-  version: 0.9.6
-  prerelease: 
+  version: 0.9.10
 platform: ruby
 authors:
 - Dmitry Mozzherin
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-06-21 00:00:00.000000000 Z
+date: 2013-12-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: '1.5'
+        version: '1.6'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: '1.5'
+        version: '1.6'
 - !ruby/object:Gem::Dependency
   name: parsley-store
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: 0.3.2
+        version: '0.3'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: 0.3.2
+        version: '0.3'
 - !ruby/object:Gem::Dependency
   name: archive-tar-minitar
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
@@ -54,102 +48,137 @@ dependencies:
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
         version: '0.5'
 - !ruby/object:Gem::Dependency
-  name: debugger
+  name: rake
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: '1.3'
+        version: '10.1'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: '1.3'
+        version: '10.1'
 - !ruby/object:Gem::Dependency
-  name: parsley-store
+  name: bundler
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ~>
       - !ruby/object:Gem::Version
-        version: 0.3.0
-  type: :runtime
+        version: '1.3'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ~>
       - !ruby/object:Gem::Version
-        version: 0.3.0
+        version: '1.3'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ~>
       - !ruby/object:Gem::Version
-        version: 1.2.9
+        version: '2.14'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ~>
       - !ruby/object:Gem::Version
-        version: 1.2.9
+        version: '2.14'
 - !ruby/object:Gem::Dependency
   name: cucumber
   requirement: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ~>
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '1.3'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
-    none: false
     requirements:
-    - - ! '>='
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.3'
+- !ruby/object:Gem::Dependency
+  name: coveralls
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
       - !ruby/object:Gem::Version
-        version: '0'
-description: Darwin Core Archive is the current standard exchange format for GLobal
-  Names Architecture modules. This gem makes it easy to incorporate files in Darwin
-  Core Archive format into a ruby project.
-email: dmozzherin at gmail dot com
+        version: '0.7'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '0.7'
+- !ruby/object:Gem::Dependency
+  name: debugger
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.6'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.6'
+- !ruby/object:Gem::Dependency
+  name: git
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.2'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.2'
+description: "Darwin Core Archive is the current standard exchange \n                          format
+  for GLobal Names Architecture modules.  \n                          This gem makes
+  it easy to incorporate files in \n                          Darwin Core Archive
+  format into a ruby project."
+email:
+- dmozzherin at gmail dot com
 executables: []
 extensions: []
-extra_rdoc_files:
-- LICENSE
-- README.md
+extra_rdoc_files: []
 files:
 - .document
+- .gitignore
+- .rspec
+- .ruby-version
 - .rvmrc
 - .travis.yml
 - CHANGELOG
 - Gemfile
-- Gemfile.lock
 - LICENSE
 - README.md
 - Rakefile
-- VERSION
+- ']'
+- dwc-archive.gemspec
 - features/dwca-creator.feature
 - features/dwca-reader.feature
 - features/step_definitions/dwc-creator_steps.rb
 - features/step_definitions/dwc-reader_steps.rb
 - features/support/env.rb
 - lib/dwc-archive.rb
-- lib/dwc-archive/.expander.rb.swo
 - lib/dwc-archive/archive.rb
 - lib/dwc-archive/classification_normalizer.rb
 - lib/dwc-archive/core.rb
@@ -161,7 +190,7 @@ files:
 - lib/dwc-archive/generator_meta_xml.rb
 - lib/dwc-archive/ingester.rb
 - lib/dwc-archive/metadata.rb
-- lib/dwc-archive/utf_regex_ruby18.rb
+- lib/dwc-archive/version.rb
 - lib/dwc-archive/xml_reader.rb
 - spec/files/broken.tar.gz
 - spec/files/data.tar.gz
@@ -185,35 +214,77 @@ files:
 - spec/files/synonyms_in_core_accepted_name_field.tar.gz
 - spec/files/synonyms_in_extension.tar.gz
 - spec/files/uncompressed
-- spec/lib/dwc-archive_spec.rb
+- spec/lib/classification_normalizer_spec.rb
+- spec/lib/core_spec.rb
+- spec/lib/darwin_core_spec.rb
+- spec/lib/generator_eml_xml_spec.rb
+- spec/lib/generator_meta_xml_spec.rb
+- spec/lib/generator_spec.rb
+- spec/lib/gnub_taxon_spec.rb
+- spec/lib/metadata_spec.rb
+- spec/lib/taxon_normalized_spec.rb
 - spec/lib/xml_reader_spec.rb
-- spec/spec.opts
 - spec/spec_helper.rb
 homepage: http://github.com/GlobalNamesArchitecture/dwc-archive
-licenses: []
+licenses:
+- MIT
+metadata: {}
 post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - '>='
     - !ruby/object:Gem::Version
       version: '0'
-      segments:
-      - 0
-      hash: 2372622078447411988
 required_rubygems_version: !ruby/object:Gem::Requirement
-  none: false
   requirements:
-  - - ! '>='
+  - - '>='
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.25
+rubygems_version: 2.0.14
 signing_key: 
-specification_version: 3
+specification_version: 4
 summary: Handler of Darwin Core Archive files
-test_files: []
+test_files:
+- features/dwca-creator.feature
+- features/dwca-reader.feature
+- features/step_definitions/dwc-creator_steps.rb
+- features/step_definitions/dwc-reader_steps.rb
+- features/support/env.rb
+- spec/files/broken.tar.gz
+- spec/files/data.tar.gz
+- spec/files/data.zip
+- spec/files/data_with_quotes.tar.gz
+- spec/files/eml.xml
+- spec/files/empty_coreid.tar.gz
+- spec/files/file with characters(3).gz
+- spec/files/flat_list.tar.gz
+- spec/files/gnub.tar.gz
+- spec/files/invalid.tar.gz
+- spec/files/junk_dir_inside.zip
+- spec/files/language_locality.tar.gz
+- spec/files/latin1.tar.gz
+- spec/files/linnean.tar.gz
+- spec/files/meta.xml
+- spec/files/minimal.tar.gz
+- spec/files/not_synonym_in_extension.tar.gz
+- spec/files/sci_name_authorship.tar.gz
+- spec/files/sci_name_authorship_dup.tar.gz
+- spec/files/synonyms_in_core_accepted_name_field.tar.gz
+- spec/files/synonyms_in_extension.tar.gz
+- spec/files/uncompressed
+- spec/lib/classification_normalizer_spec.rb
+- spec/lib/core_spec.rb
+- spec/lib/darwin_core_spec.rb
+- spec/lib/generator_eml_xml_spec.rb
+- spec/lib/generator_meta_xml_spec.rb
+- spec/lib/generator_spec.rb
+- spec/lib/gnub_taxon_spec.rb
+- spec/lib/metadata_spec.rb
+- spec/lib/taxon_normalized_spec.rb
+- spec/lib/xml_reader_spec.rb
+- spec/spec_helper.rb