uc3-dmp-citation 0.0.1

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: 6f168bafa37250b52f74453f1e9f044fd5ba07eab51632b650207514d7c9decd
+  data.tar.gz: b1f37ab91666d6356f29b6d5da6242947310df4fe5f7e337d75310b6ea78ec7c
+SHA512:
+  metadata.gz: d6fbdb819e93454d608a3f281e0be0839e2b7785b8072035b3a36192c5ae2be8fe79a6433174f37bdbaec8f99b836bce0a993d8189e966bd78f40a1a22f174dc
+  data.tar.gz: c56061f1ce9e204c97d05ddd03d385b0dcc4044e4f3bfc26e3f4ed436e26550059e8d44c0ba5b64e14ad941f4ffae74fb61d6bdaab6aaaa35cddbfc3a974b8c2

data/README.md

@@ -0,0 +1,3 @@
+# uc3-dmp-logger
+
+Logger service to write CloudWatch logs in a specific format

data/lib/uc3-dmp-citation/version.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+module Uc3DmpCitation
+  VERSION = '0.0.1'
+end

data/lib/uc3-dmp-citation.rb

@@ -0,0 +1,106 @@
+# rubocop:disable Naming/FileName
+# frozen_string_literal: true
+
+require 'bibtex'
+require 'citeproc'
+require 'csl/styles'
+
+require 'uc3-dmp-external-api'
+
+module Uc3DmpCitation
+  class CiterError < StandardError; end
+
+  class Citer
+    DEFAULT_CITATION_STYLE = 'chicago-author-date'
+    DEFAULT_DOI_URL = 'http://doi.org'
+    DEFAULT_WORK_TYPE = 'Dataset'
+
+    DOI_REGEX = %r{[0-9]{2}\.[0-9]{4,}/[a-zA-Z0-9/_.-]+}
+
+    MSG_BIBTEX_FAILURE = 'Unable to fetch Bibtex for the specified DOI.'
+    MSG_UNABLE_TO_UPDATE = 'Unable to update the citations on the DMP ID.'
+
+    def fetch_citation(doi:, logger: nil)
+      uri = _doi_to_uri(doi: doi)
+      return nil if uri.nil? || uri.blank?
+
+      logger.debug(message: "Fetching BibTeX from: #{uri}") if logger.respond_to?(:debug)
+      resp = Uc3DmpExternalApi::Client.call(url: uri, method: :get, additional_headers: headers, logger: logger)
+      return nil if resp.nil? || resp.to_s.strip.empty?
+
+      bibtex = BibTeX.parse(_cleanse_bibtex(text: resp))
+      work_type = identifier['work_type'].nil? ? determine_work_type(bibtex: bibtex) :  identifier['work_type']
+      _bibtex_to_citation(uri: uri, work_type: work_type, bibtex: bibtex)
+    end
+
+    private
+
+    # Will convert 'doi:10.1234/abcdefg' to 'http://doi.org/10.1234/abcdefg'
+    def _doi_to_uri(doi:)
+      val = doi.match(DOI_REGEX).to_s
+      return nil if val.nil? || val.strip == ''
+
+      doi.start_with?('http') ? doi : "#{DEFAULT_DOI_URL}/#{doi.gsub('doi:', '')}"
+    end
+
+    # If no :work_type was specified we can try to derive it from the BibTeX metadata
+    def _determine_work_type(bibtex:)
+      return '' if bibtex.nil? || bibtex.data.nil? || bibtex.data.first.nil?
+
+      return 'article' unless bibtex.data.first.journal.nil?
+
+      ''
+    end
+
+    def _cleanse_bibtex(text:)
+      return nil if text.nil? || text.to_s.strip == ''
+
+      # Make sure we're working with UTF8
+      utf8 = text.force_encoding('UTF-8')
+
+      # Remove any encoded HTML (e.g. "Regular text $\\lt$strong$\\gt$Bold text$\\lt$/strong$\\gt$")
+      utf8 = utf8.gsub(%r{\$?\\\$?(less|lt|Lt)\$/?[a-zA-Z]+\$?\\\$?(greater|gt|Gt)\$}, '')
+      # Replace any special dash, semicolon and quote characters with a minus sign or single/double quote
+      utf8 = utf8.gsub(%r{\$?\\(T|t)ext[a-zA-Z]+dash\$?}, '-').gsub(%r{\{(T|t)ext[a-zA-Z]+dash\}}, '-')
+                 .gsub(%r{\$?\\(M|m)athsemicolon\$?}, ':').gsub(%r{\{(M|m)semicolon\}}, ':')
+                 .gsub(%r{\$?\\(T|t)extquotesingle\$?}, "'").gsub(%r{\{(T|t)extquotesingle\}}, "'")
+                 .gsub(%r{\$?\\(T|t)extquotedouble\$?}, '"').gsub(%r{\{(T|t)extquotedouble\}}, '"')
+      # Remove any remaining `\v` entries which attempt to construct an accented character
+      utf8.gsub(%r{\\v}, '')
+    end
+
+    # Convert the BibTeX item to a citation
+    def _bibtex_to_citation(uri:, work_type: DEFAULT_WORK_TYPE, bibtex:, style: DEFAULT_CITATION_STYLE)
+      return nil unless uri.is_a?(String) && uri.strip != ''
+      return nil if bibtex.nil? || bibtex.data.nil? || bibtex.data.first.nil?
+
+      cp = CiteProc::Processor.new(style: style, format: 'html')
+      cp.import(bibtex.to_citeproc)
+      citation = cp.render(:bibliography, id: bibtex.data.first.id)
+      return nil unless citation.is_a?(Array) && citation.any?
+
+      # The CiteProc renderer has trouble with some things so fix them here
+      #   - For some reason words in all caps in the title get wrapped in curl brackets
+      citation = citation.first.gsub('{', '').gsub('}', '')
+
+      unless work_type.nil? || work_type.strip == ''
+        # This supports the :apa and :chicago-author-date styles
+        citation = citation.gsub(/\.”\s+/, "\.” [#{work_type.gsub('_', ' ').capitalize}]. ")
+                           .gsub(/<\/i>\.\s+/, "<\/i>\. [#{work_type.gsub('_', ' ').capitalize}]. ")
+      end
+
+      # Convert the URL into a link. Ensure that the trailing period is not a part of
+      # the link!
+      citation.gsub(URI.regexp) do |url|
+        if url.start_with?('http')
+          '<a href="%{url}" target="_blank">%{url}</a>.' % {
+            url: url.end_with?('.') ? uri : "#{uri}."
+          }
+        else
+          url
+        end
+      end
+    end
+  end
+end
+# rubocop:enable Naming/FileName

metadata

@@ -0,0 +1,104 @@
+--- !ruby/object:Gem::Specification
+name: uc3-dmp-citation
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+platform: ruby
+authors:
+- Brian Riley
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2023-08-17 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: byebug
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 11.1.3
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 11.1.3
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.9.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 3.9.0
+- !ruby/object:Gem::Dependency
+  name: rubocop
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 1.50.2
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 1.50.2
+- !ruby/object:Gem::Dependency
+  name: rubocop-rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 2.20.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 2.20.0
+description: Helper for working with DOI citations
+email:
+- brian.riley@ucop.edu
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- README.md
+- lib/uc3-dmp-citation.rb
+- lib/uc3-dmp-citation/version.rb
+homepage: https://github.com/CDLUC3/dmp-hub-cfn/blob/main/src/sam/gems/uc3-dmp-citation
+licenses:
+- MIT
+metadata:
+  rubygems_mfa_required: 'false'
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '2.7'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.1.6
+signing_key: 
+specification_version: 4
+summary: DMPTool gem that provides support for fetching BibTex for a DOI and converting
+  it into a citation
+test_files: []