bolognese 1.9.18 → 1.10.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 53a45d740ac30bb235316e61ad8650ae88bd789ca1bafcc09a004b1ac9bf76cc
-  data.tar.gz: 771f8867608f7b3984861f10ceecc0b48e8ac17763f1b936e1d9f79002e16f13
+  metadata.gz: 532586fba70709be6c7581b595a2c0d72ad1000c9c54b951ad5e4758715e81c1
+  data.tar.gz: 3029c21ac460e41479459df642e3d501b1195e6cc8a0e26d6a4ba5bba91ce13f
 SHA512:
-  metadata.gz: 1268a2e57636ec811b9004143c7092325e249050ffa4669119491e290604b5cc14c3e13983f2014196919b5bb97bb197e76c7620ced11dc34103e601b923ecb5
-  data.tar.gz: fc1a9df23987eb69d5797102fb537bcbe65a5f565d19413e7b018af8442effa295cbaf5efeba04450ac8ae017d8b017141f009694aabde102114bfccafcba5d1
+  metadata.gz: 3893a351e49ad8e8acb0ef5265fb3acdc149e67a99174de0d0855d594e7d9c16f09baeda4d9b1aba7235aef661135e7d1bd1ac401df81601c99217e155153395
+  data.tar.gz: f6225b13dae84dcc897fa1dda7aadb266e667f8bf6a3a2da2e3f6ad134f314f17c3c6613ef8c2dec54091d9f4b4f055023bdb00ace39baf018b1654820cf7eff

data/.github/workflows/build.yml

@@ -5,33 +5,19 @@ on:
     branches:
       - "master"
 jobs:
+  ci:
+    uses: ./.github/workflows/ci.yml
   build:
+    needs: ci
     runs-on: ubuntu-latest
     env:
       SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL }}
     steps:
-    - uses: actions/checkout@v2
-    - name: Set up Ruby 2.6
-      uses: actions/setup-ruby@v1
-      with:
-        ruby-version: 2.6.x
-
-    - name: Build and test
-      run: |
-        gem install bundler
-        bundle install
-        bundle exec rspec
-
-    - name: Publish code coverage
-      uses: paambaati/codeclimate-action@v2.7.4
-      env:
-        CC_TEST_REPORTER_ID: ${{ secrets.CC_TEST_REPORTER_ID }}
-
     - name: Notify Slack
       uses: adamkdean/simple-slack-notify@1.0.4
       with:
         channel: '#ops'
         username: 'GitHub Actions'
         color: 'good'
-        text: 'A new version of the bolognese gem has been built.'
+        text: 'A new version of the bolognese gem is ready for release.'
 

data/.github/workflows/ci.yml

@@ -0,0 +1,22 @@
+name: Run Tests
+on:
+  workflow_call:
+
+jobs:
+  test:
+    strategy:
+      fail-fast: false
+      matrix:
+        ruby: ["2.6", "2.7", "3.0", "3.1"]
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v3
+    - name: Set up Ruby ${{matrix.ruby}}
+      uses: ruby/setup-ruby@v1
+      with:
+        ruby-version: ${{matrix.ruby}}
+        bundler-cache: true
+
+    - name: Run tests
+      run: |
+        bundle exec rspec

data/.github/workflows/pull-request.yml

@@ -5,17 +5,5 @@ on:
     branches:
       - master
 jobs:
-  build:
-    runs-on: ubuntu-latest
-    steps:
-    - uses: actions/checkout@v2
-    - name: Set up Ruby 2.6
-      uses: actions/setup-ruby@v1
-      with:
-        ruby-version: 2.6.x
-
-    - name: Install and test
-      run: |
-        gem install bundler
-        bundle install
-        bundle exec rspec
+  ci:
+    uses: ./.github/workflows/ci.yml

data/.github/workflows/release.yml

@@ -4,28 +4,23 @@ on:
   release:
     types: [published]
 jobs:
+  ci:
+    uses: ./.github/workflows/ci.yml
   build:
+    needs: ci
     runs-on: ubuntu-latest
     env:
       SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK_URL }}
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v3
     - name: Set up Ruby 2.6
-      uses: actions/setup-ruby@v1
+      uses: ruby/setup-ruby@v1
       with:
-        ruby-version: 2.6.x
+        ruby-version: "2.6"
 
-    - name: Build and test
+    - name: Build
       run: |
-        gem install bundler
         bundle install
-        bundle exec rspec spec
-
-    - name: Code Climate Test Reporter
-      uses: aktions/codeclimate-test-reporter@v1
-      with:
-        codeclimate-test-reporter-id: ${{ secrets.CC_TEST_REPORTER_ID }}
-        command: after-build
 
     - name: Publish to RubyGems
       run: |
@@ -44,4 +39,4 @@ jobs:
         channel: '#ops'
         username: 'GitHub Actions'
         color: 'good'
-        text: 'A new version of the bolognese gem has been released.'
+        text: 'A new version of the bolognese gem has been released.'

data/CHANGELOG.md

@@ -1,20 +1,25 @@
 # Changelog
 
-## [Unreleased](https://github.com/datacite/bolognese/tree/HEAD)
+## [1.10.0](https://github.com/datacite/bolognese/tree/1.10.0) (2022-10-14)
 
-[Full Changelog](https://github.com/datacite/bolognese/compare/1.9.18...HEAD)
+[Full Changelog](https://github.com/datacite/bolognese/compare/1.9.18...1.10.0)
 
-**Fixed bugs:**
+**Merged pull requests:**
 
-- Fix related items handling to not include elements when they are blank [\#140](https://github.com/datacite/bolognese/issues/140)
-- Related Items error out if RelatedItemIdentifier is nil [\#139](https://github.com/datacite/bolognese/issues/139)
+- Add DFG-to-FOS mappings utility function [\#147](https://github.com/datacite/bolognese/pull/147) ([jrhoads](https://github.com/jrhoads))
 
 ## [1.9.18](https://github.com/datacite/bolognese/tree/1.9.18) (2022-10-11)
 
 [Full Changelog](https://github.com/datacite/bolognese/compare/1.9.17...1.9.18)
 
+**Fixed bugs:**
+
+- Fix related items handling to not include elements when they are blank [\#140](https://github.com/datacite/bolognese/issues/140)
+- Related Items error out if RelatedItemIdentifier is nil [\#139](https://github.com/datacite/bolognese/issues/139)
+
 **Merged pull requests:**
 
+- bump version number [\#145](https://github.com/datacite/bolognese/pull/145) ([digitaldogsbody](https://github.com/digitaldogsbody))
 - Fix relatedItems handling [\#144](https://github.com/datacite/bolognese/pull/144) ([digitaldogsbody](https://github.com/digitaldogsbody))
 - Fixes normalization behavior that omitted non-URL funding identifiers when reading from DataCite XML. Adds normalization for Crossref Funder ID and ROR funderIdentifierTypes. [\#143](https://github.com/datacite/bolognese/pull/143) ([codycooperross](https://github.com/codycooperross))
 - Including a schemeURI with a nameIdentifier that is a URL will concat… [\#137](https://github.com/datacite/bolognese/pull/137) ([codycooperross](https://github.com/codycooperross))
@@ -59,7 +64,6 @@
 **Merged pull requests:**
 
 - update nokogiri to latest, including updating maremma [\#130](https://github.com/datacite/bolognese/pull/130) ([orangewolf](https://github.com/orangewolf))
-- mapping for DFG-to-OECD [\#125](https://github.com/datacite/bolognese/pull/125) ([kjgarza](https://github.com/kjgarza))
 
 ## [1.9.12](https://github.com/datacite/bolognese/tree/1.9.12) (2021-08-19)
 
@@ -75,6 +79,7 @@
 
 **Merged pull requests:**
 
+- Test against multiple versions of ruby \(2.6, 2.7, 3.0, and 3.1\) [\#141](https://github.com/datacite/bolognese/pull/141) ([jrhoads](https://github.com/jrhoads))
 - Allow Orcid URLs to have been issued from the sandbox [\#121](https://github.com/datacite/bolognese/pull/121) ([prdanelli](https://github.com/prdanelli))
 
 ## [1.9.10](https://github.com/datacite/bolognese/tree/1.9.10) (2021-07-20)
@@ -110,6 +115,7 @@
 
 **Merged pull requests:**
 
+- mapping for DFG-to-OECD [\#125](https://github.com/datacite/bolognese/pull/125) ([kjgarza](https://github.com/kjgarza))
 - Fix missing affiliation identifier [\#117](https://github.com/datacite/bolognese/pull/117) ([richardhallett](https://github.com/richardhallett))
 - Add support for book title and additional metadata when reading Crossref Book Chapter DOI [\#115](https://github.com/datacite/bolognese/pull/115) ([prdanelli](https://github.com/prdanelli))
 - Valid Funder DOIs are being rejected for not starting with a 5 [\#114](https://github.com/datacite/bolognese/pull/114) ([prdanelli](https://github.com/prdanelli))
@@ -130,6 +136,10 @@
 
 - Singular geoLocationPolygons are wrapped in additional array [\#110](https://github.com/datacite/bolognese/issues/110)
 
+**Merged pull requests:**
+
+- Support returning single geoLocationPolygon [\#111](https://github.com/datacite/bolognese/pull/111) ([richardhallett](https://github.com/richardhallett))
+
 ## [1.9.3](https://github.com/datacite/bolognese/tree/1.9.3) (2021-03-26)
 
 [Full Changelog](https://github.com/datacite/bolognese/compare/1.9.2...1.9.3)
@@ -159,7 +169,6 @@
 
 **Merged pull requests:**
 
-- Support returning single geoLocationPolygon [\#111](https://github.com/datacite/bolognese/pull/111) ([richardhallett](https://github.com/richardhallett))
 - Change detection for crosscite format [\#107](https://github.com/datacite/bolognese/pull/107) ([richardhallett](https://github.com/richardhallett))
 - Add support for multiple geolocation polygons [\#106](https://github.com/datacite/bolognese/pull/106) ([richardhallett](https://github.com/richardhallett))
 - Support schema 4 4 [\#105](https://github.com/datacite/bolognese/pull/105) ([richardhallett](https://github.com/richardhallett))
@@ -1380,15 +1389,15 @@
 
 ## [v.0.9.70](https://github.com/datacite/bolognese/tree/v.0.9.70) (2018-01-12)
 
-[Full Changelog](https://github.com/datacite/bolognese/compare/v.0.6.69...v.0.9.70)
+[Full Changelog](https://github.com/datacite/bolognese/compare/v.0.9.69...v.0.9.70)
 
-## [v.0.6.69](https://github.com/datacite/bolognese/tree/v.0.6.69) (2018-01-11)
+## [v.0.9.69](https://github.com/datacite/bolognese/tree/v.0.9.69) (2018-01-11)
 
-[Full Changelog](https://github.com/datacite/bolognese/compare/v.0.9.69...v.0.6.69)
+[Full Changelog](https://github.com/datacite/bolognese/compare/v.0.6.69...v.0.9.69)
 
-## [v.0.9.69](https://github.com/datacite/bolognese/tree/v.0.9.69) (2018-01-11)
+## [v.0.6.69](https://github.com/datacite/bolognese/tree/v.0.6.69) (2018-01-11)
 
-[Full Changelog](https://github.com/datacite/bolognese/compare/v.0.9.68...v.0.9.69)
+[Full Changelog](https://github.com/datacite/bolognese/compare/v.0.9.68...v.0.6.69)
 
 ## [v.0.9.68](https://github.com/datacite/bolognese/tree/v.0.9.68) (2018-01-11)
 

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    bolognese (1.9.18)
+    bolognese (1.10.0)
       activesupport (>= 4.2.5)
       benchmark_methods (~> 0.7)
       bibtex-ruby (>= 5.1.0)
@@ -219,4 +219,4 @@ DEPENDENCIES
   webmock (~> 3.0, >= 3.0.1)
 
 BUNDLED WITH
-   2.3.23
+   2.3.12

data/bolognese.gemspec

@@ -13,7 +13,7 @@ Gem::Specification.new do |s|
   s.version       = Bolognese::VERSION
   s.extra_rdoc_files = ["README.md"]
   s.license       = 'MIT'
-  s.required_ruby_version = '~> 2.3'
+  s.required_ruby_version = ['>=2.3']
 
   # Declare dependencies here, rather than in the Gemfile
   s.add_dependency 'maremma', '>= 4.9.4', '< 5'

data/lib/bolognese/doi_utils.rb

@@ -2,6 +2,10 @@
 
 module Bolognese
   module DoiUtils
+    class << self
+      include DoiUtils
+    end
+
     def validate_doi(doi)
       doi = Array(/\A(?:(http|https):\/(\/)?(dx\.)?(doi.org|handle.stage.datacite.org|handle.test.datacite.org)\/)?(doi:)?(10\.\d{4,5}\/.+)\z/.match(doi)).last
       # remove non-printing whitespace and downcase

data/lib/bolognese/utils.rb

@@ -2,6 +2,10 @@
 
 module Bolognese
   module Utils
+    class << self
+      include Utils
+    end
+
     NORMALIZED_LICENSES = {
       "https://creativecommons.org/licenses/by/1.0" => "https://creativecommons.org/licenses/by/1.0/legalcode",
       "https://creativecommons.org/licenses/by/2.0" => "https://creativecommons.org/licenses/by/2.0/legalcode",
@@ -484,6 +488,28 @@ module Bolognese
       ":etal" => "too numerous to list (et alia)"
     }
 
+    RESOURCE_PATHS = {
+      spdx: 'spdx/licenses.json',
+      fos: 'oecd/fos-mappings.json',
+      for: 'oecd/for-mappings.json',
+      dfg: 'oecd/dfg-mappings.json'
+    }
+
+    def resources_dir_path
+      File.expand_path('../../../resources', __FILE__) + '/'
+    end
+
+    def resource_file( extra_path )
+      File.read(resources_dir_path + extra_path)
+    end
+
+    def resource_json( resource_symbol )
+      if RESOURCE_PATHS.keys().include?(resource_symbol)
+        JSON.load(resource_file(RESOURCE_PATHS[resource_symbol]))
+      end
+    end
+
+
     def find_from_format(id: nil, string: nil, ext: nil, filename: nil)
       if id.present?
         find_from_format_by_id(id)
@@ -615,7 +641,7 @@ module Bolognese
       return nil unless id.present?
 
       # check for valid DOI
-      doi = normalize_doi(id, options)
+      doi = DoiUtils::normalize_doi(id, options)
       return doi if doi.present?
 
       # check for valid HTTP uri
@@ -674,8 +700,8 @@ module Bolognese
     def normalize_ids(ids: nil, relation_type: nil)
       Array.wrap(ids).select { |idx| idx["@id"].present? }.map do |idx|
         id = normalize_id(idx["@id"])
-        related_identifier_type = doi_from_url(id).present? ? "DOI" : "URL"
-        id = doi_from_url(id) || id
+        related_identifier_type = DoiUtils::doi_from_url(id).present? ? "DOI" : "URL"
+        id = DoiUtils::doi_from_url(id) || id
 
         { "relatedIdentifier" => id,
           "relationType" => relation_type,
@@ -1233,7 +1259,7 @@ module Bolognese
     end
 
     def name_to_spdx(name)
-      spdx = JSON.load(File.read(File.expand_path('../../../resources/spdx/licenses.json', __FILE__))).fetch("licenses")
+      spdx = resource_json(:spdx).fetch("licenses")
       license = spdx.find { |l| l["name"] == name || l["licenseId"] == name || l["seeAlso"].first == normalize_cc_url(name) }
 
       if license
@@ -1249,7 +1275,7 @@ module Bolognese
     end
 
     def hsh_to_spdx(hsh)
-      spdx = JSON.load(File.read(File.expand_path('../../../resources/spdx/licenses.json', __FILE__))).fetch("licenses")
+      spdx = resource_json(:spdx).fetch("licenses")
       license = spdx.find { |l| l["licenseId"].casecmp?(hsh["rightsIdentifier"]) || l["seeAlso"].first == normalize_cc_url(hsh["rightsURI"]) || l["name"] == hsh["rights"] || l["seeAlso"].first == normalize_cc_url(hsh["rights"]) }
 
       if license
@@ -1273,7 +1299,7 @@ module Bolognese
 
     def name_to_fos(name)
       # first find subject in Fields of Science (OECD)
-      fos = JSON.load(File.read(File.expand_path('../../../resources/oecd/fos-mappings.json', __FILE__))).fetch("fosFields")
+      fos = resource_json(:fos).fetch("fosFields")
 
       subject = fos.find { |l| l["fosLabel"] == name || "FOS: " + l["fosLabel"] == name }
 
@@ -1289,7 +1315,7 @@ module Bolognese
 
       # if not found, look in Fields of Research (Australian and New Zealand Standard Research Classification)
       # and map to Fields of Science. Add an extra entry for the latter
-      fores = JSON.load(File.read(File.expand_path('../../../resources/oecd/for-mappings.json', __FILE__)))
+      fores = resource_json(:for)
       for_fields = fores.fetch("forFields")
       for_disciplines = fores.fetch("forDisciplines")
 
@@ -1311,7 +1337,7 @@ module Bolognese
 
     def hsh_to_fos(hsh)
       # first find subject in Fields of Science (OECD)
-      fos = JSON.load(File.read(File.expand_path('../../../resources/oecd/fos-mappings.json', __FILE__))).fetch("fosFields")
+      fos = resource_json(:fos).fetch("fosFields")
       subject = fos.find { |l| l["fosLabel"] == hsh["__content__"] || "FOS: " + l["fosLabel"] == hsh["__content__"] || l["fosLabel"] == hsh["subject"]}
 
       if subject
@@ -1330,7 +1356,7 @@ module Bolognese
 
       # if not found, look in Fields of Research (Australian and New Zealand Standard Research Classification)
       # and map to Fields of Science. Add an extra entry for the latter
-      fores = JSON.load(File.read(File.expand_path('../../../resources/oecd/for-mappings.json', __FILE__)))
+      fores = resource_json(:for)
       for_fields = fores.fetch("forFields")
       for_disciplines = fores.fetch("forDisciplines")
 
@@ -1369,5 +1395,20 @@ module Bolognese
           "lang" => hsh["lang"] }.compact]
       end
     end
+
+    def dfg_ids_to_fos(dfg_ids)
+      dfgs = resource_json(:dfg).fetch("dfgFields")
+      ids = Array.wrap(dfg_ids)
+
+      subjects = dfgs.select { |l| ids.include?(l["dfgId"])}
+      subjects.map do |subject|
+        {
+          "classificationCode" => subject["fosId"],
+          "subject" =>  subject["fosLabel"],
+          "subjectScheme" => "Fields of Science and Technology (FOS)",
+          "schemeUri" => "http://www.oecd.org/science/inno/38235147.pdf"
+        }
+      end
+    end
   end
 end

data/lib/bolognese/version.rb

@@ -1,3 +1,3 @@
 module Bolognese
-  VERSION = "1.9.18"
+  VERSION = "1.10.0"
 end

data/spec/utils_spec.rb

@@ -2,10 +2,8 @@
 
 require 'spec_helper'
 
-describe Bolognese::Metadata, vcr: true do
-  let(:input) { "https://doi.org/10.1101/097196" }
-
-  subject { Bolognese::Metadata.new(input: input, from: "crossref") }
+describe Bolognese::Utils do
+  subject { Bolognese::Utils }
 
   context "validate url" do
     it "DOI" do
@@ -614,4 +612,55 @@ describe Bolognese::Metadata, vcr: true do
     end
 
   end
+
+  context "DFG mappings" do
+    it "dfg_id_to_fos match" do
+      dfg_id = "20108"
+      response = subject.dfg_ids_to_fos(dfg_id)
+      expect(response).to eq([
+        {
+          "classificationCode" =>"1.6",
+          "subject" => "Biological sciences",
+          "subjectScheme" => "Fields of Science and Technology (FOS)",
+          "schemeUri" => "http://www.oecd.org/science/inno/38235147.pdf"
+        },
+        {
+          "classificationCode" =>"3.1",
+          "subject" => "Basic medicine",
+          "subjectScheme" => "Fields of Science and Technology (FOS)",
+          "schemeUri" => "http://www.oecd.org/science/inno/38235147.pdf"
+        },
+      ])
+    end
+    it "dfg_id_to_fos no match" do
+      dfg_id = "000"
+      response = subject.dfg_ids_to_fos(dfg_id)
+      expect(response).to eq([])
+    end
+
+    it "dfg_id_to_fos match list" do
+      dfg_ids = ["101", "20108"]
+      response = subject.dfg_ids_to_fos(dfg_ids)
+      expect(response).to eq([
+        {"classificationCode"=>"6.1",
+         "schemeUri"=>"http://www.oecd.org/science/inno/38235147.pdf",
+         "subject"=>"History and archaeology",
+         "subjectScheme"=>"Fields of Science and Technology (FOS)"},
+        {"classificationCode"=>"6.2",
+         "schemeUri"=>"http://www.oecd.org/science/inno/38235147.pdf",
+         "subject"=>"Languages and literature",
+         "subjectScheme"=>"Fields of Science and Technology (FOS)"},
+        {"classificationCode"=>"1.6",
+         "schemeUri"=>"http://www.oecd.org/science/inno/38235147.pdf",
+         "subject"=>"Biological sciences",
+         "subjectScheme"=>"Fields of Science and Technology (FOS)"},
+        {"classificationCode"=>"3.1",
+         "schemeUri"=>"http://www.oecd.org/science/inno/38235147.pdf",
+         "subject"=>"Basic medicine",
+         "subjectScheme"=>"Fields of Science and Technology (FOS)"}
+      ])
+    end
+
+
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: bolognese
 version: !ruby/object:Gem::Version
-  version: 1.9.18
+  version: 1.10.0
 platform: ruby
 authors:
 - Martin Fenner
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2022-10-11 00:00:00.000000000 Z
+date: 2022-10-14 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: maremma
@@ -567,6 +567,7 @@ extra_rdoc_files:
 files:
 - ".github/workflows/build.yml"
 - ".github/workflows/changelog.yml"
+- ".github/workflows/ci.yml"
 - ".github/workflows/pull-request.yml"
 - ".github/workflows/release.yml"
 - ".gitignore"
@@ -1288,7 +1289,7 @@ require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - "~>"
+  - - ">="
     - !ruby/object:Gem::Version
       version: '2.3'
 required_rubygems_version: !ruby/object:Gem::Requirement