relaton-nist 1.19.3 → 1.19.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 7a4b4d939bc25c7c8968747b391784b5e0238a6961b5544d6f48cd77848addf9
-  data.tar.gz: f5684a7aa32f04b30fed89c1d43dc0c5247562a94bf97b1d65f6b0c80de2f054
+  metadata.gz: 47c01d46202e813e27019c4beb5e5ccc4aa258a61dd4f6ea3b184009e6ae9e09
+  data.tar.gz: 2dadbcee5b551aefb4f4cefe04620964d3455acb842f005944e2f30e195d94ba
 SHA512:
-  metadata.gz: aa2d613dc42ac689b422b73d0b930459bbfdad99ca8e0e4302a8f586931f89f90ae0189d6d052e8e003659b75df4f5a07941835d4f03f4e98ead50ebcfef738d
-  data.tar.gz: 19b7411e9f15891b7513dd934a678702fcfaccc80512385e2354d69ca66f2cf3617b2f8e88ce47b7141ee885dc933e59931731dec14e1b90e550964184f4ff92
+  metadata.gz: 3e264c34cfea0e5ef97894265d0e6917081d05e7d317142475badb242036eeae4e4caafde6cd2f19c62ae12b99560cef88d92638f40d2e9592a37a704cef8aca
+  data.tar.gz: 262e7bf0461fbd19f9da7346128166ab2d07fdc3e8f3cffff2ec20ec5fad2b6883a8f4f85624675e5e586a29b3a17cfe28eee426b63b23ce8b7c3fc3fb83a426

data/lib/relaton_nist/data_fetcher.rb

@@ -6,7 +6,7 @@ require_relative "mods_parser"
 
 module RelatonNist
   class DataFetcher
-    URL = "https://github.com/usnistgov/NIST-Tech-Pubs/releases/download/May2024/allrecords-MODS.xml"
+    URL = "https://github.com/usnistgov/NIST-Tech-Pubs/releases/download/Oct2024/allrecords-MODS.xml"
 
     def initialize(output, format)
       @output = output

data/lib/relaton_nist/hit_collection.rb

@@ -23,7 +23,7 @@ module RelatonNist
     #
     def initialize(text, year = nil, opts = {})
       super text, year
-      @reference = year && !text.match?(/:\d{4}$/) ? "#{text}:#{year}" : text
+      @reference = text # year && !text.match?(/:\d{4}$/) ? "#{text}-#{year}" : text
       @opts = opts
     end
 
@@ -58,21 +58,29 @@ module RelatonNist
     # @return [Array<RelatonNist::Hit>] hits
     #
     def search_filter # rubocop:disable Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity,Metrics/MethodLength
+      refid = ::Pubid::Nist::Identifier.parse(@reference)
       arr = @array.select do |item|
-        parts = doi_parts(item.hit[:json]) || code_parts(item.hit[:code])
-        iteration = item.hit.dig(:json, "iteration")
-        draft = parts[:draft] || (!iteration.nil? && iteration.match?(/\wpd/i))
-        refparts[:code] && [parts[:series], item.hit[:series]].include?(refparts[:series]) &&
-          refparts[:code].casecmp(parts[:code].upcase).zero? &&
-          refparts[:prt] == parts[:prt] &&
-          (refparts[:vol].nil? || refparts[:vol] == parts[:vol]) &&
-          (refparts[:ver].nil? || refparts[:ver] == parts[:ver]) &&
-          (refparts[:rev].nil? || refparts[:rev] == parts[:rev]) &&
-          refparts[:draft] == draft && refparts[:add] == parts[:add]
+        pubid = ::Pubid::Nist::Identifier.parse(item.hit[:code].sub(/\.$/, ""))
+        pubid == refid
+        # parts = doi_parts(item.hit[:json]) || code_parts(item.hit[:code])
+        # iteration = item.hit.dig(:json, "iteration")
+        # draft = parts[:draft] || (!iteration.nil? && iteration.match?(/\wpd/i))
+        # refparts[:code] && [parts[:series], item.hit[:series]].include?(refparts[:series]) &&
+        #   refparts[:code].casecmp(parts[:code].upcase).zero? &&
+        #   refparts[:prt] == parts[:prt] &&
+        #   (refparts[:vol].nil? || refparts[:vol] == parts[:vol]) &&
+        #   (refparts[:ver].nil? || refparts[:ver] == parts[:ver]) &&
+        #   (refparts[:rev].nil? || refparts[:rev] == parts[:rev]) &&
+        #   refparts[:draft] == draft && refparts[:add] == parts[:add]
+      rescue StandardError
+        item.hit[:code] == text
       end
+    rescue StandardError
+      arr = @array.select { |item| item.hit[:code] == text }
+    ensure
       dup = self.dup
       dup.array = arr
-      dup
+      return dup
     end
 
     private
@@ -182,6 +190,12 @@ module RelatonNist
       self
     end
 
+    def pubid(id = text)
+      Pubid::Nist::Identifier.parse(id).to_s
+    rescue StandardError
+      id
+    end
+
     #
     # Get hit from GitHub repo
     #
@@ -191,8 +205,8 @@ module RelatonNist
     #
     def from_ga # rubocop:disable Metrics/AbcSize, Metrics/MethodLength
       Util.info "Fetching from Relaton repository ...", key: @reference
-      ref = full_ref
-      return [] if ref.empty?
+      ref = pubid
+      # return [] if ref.empty?
 
       # fn = ref.gsub(%r{[/\s:.]}, "_").upcase
       index = Relaton::Index.find_or_create :nist, url: "#{GHNISTDATA}index-v1.zip", file: INDEX_FILE
@@ -227,24 +241,44 @@ module RelatonNist
         /(?<series>(?<=-)\w+$)/ =~ h["series"]
         title = [h["title-main"], h["title-sub"]].compact.join " - "
         release_date = RelatonBib.parse_date h["published-date"], false
-        Hit.new({ code: docidentifier(h), series: series.upcase, title: title, url: h["uri"],
+        Hit.new({ code: pubs_export_id(h), series: series.upcase, title: title, url: h["uri"],
                   status: h["status"], release_date: release_date, json: h }, self)
       end
     end
 
-    def docidentifier(json) # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity, Metrics/MethodLength
-      parts = doi_parts(json) || code_parts(json["docidentifier"])
+    def pubs_export_id(json) # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity, Metrics/MethodLength
+      if json && json["doi"]
+        json["doi"].sub(/^10.6028\//, "")
+      else
+        json["docidentifier"]
+      end => id
+      # parts = doi_parts(json) || code_parts(json["docidentifier"])
       # return json["docidentifier"] unless parts
 
-      id = parts[:code]
-      id = "NIST #{parts[:series]} #{id}" if parts[:series]
-      id += " Part #{parts[:prt]}" if parts[:prt]
-      id += " Vol. #{parts[:vol]}" if parts[:vol]
-      id += " Ver. #{parts[:ver]}" if parts[:ver]
-      id += " Rev. #{parts[:rev]}" if parts[:rev]
-      id += "-Add" if parts[:add]
-      id += " (Draft)" if parts[:draft] || @opts[:stage]
+      id.sub!(/(?:-draft\d*|\.\wpd)$/, "")
+      pid = ::Pubid::Nist::Identifier.parse(id)
+      case json["iteration"]
+      when "final"
+        pid.stage = ::Pubid::Nist::Stage.new id: "f", type: "pd"
+      when "fpd"
+        pid.stage = ::Pubid::Nist::Stage.new id: "f", type: "pd"
+      when /(\w)pd/
+        pid.stage = ::Pubid::Nist::Stage.new id: Regexp.last_match(1), type: "pd"
+      end
+      pid.to_s
+    rescue StandardError
+      id += " #{json["iteration"].sub('final', 'fpd')}" if json["iteration"]
       id
+
+      # id = parts[:code]
+      # id = "NIST #{parts[:series]} #{id}" if parts[:series]
+      # id += " Part #{parts[:prt]}" if parts[:prt]
+      # id += " Vol. #{parts[:vol]}" if parts[:vol]
+      # id += " Ver. #{parts[:ver]}" if parts[:ver]
+      # id += " Rev. #{parts[:rev]}" if parts[:rev]
+      # id += "-Add" if parts[:add]
+      # id += " (Draft)" if parts[:draft] || @opts[:stage]
+      # id
     end
 
     #

data/lib/relaton_nist/mods_parser.rb

@@ -29,33 +29,43 @@ module RelatonNist
     def parse_docid
       [
         { type: "NIST", id: pub_id, primary: true },
-        { type: "DOI", id: doi },
+        { type: "DOI", id: parse_doi },
       ].map { |id| RelatonBib::DocumentIdentifier.new(**id) }
     end
 
     # @return [String]
     def pub_id
-      get_id_from_str doi
+      get_id_from_str parse_doi
     end
 
     def get_id_from_str(str)
-      str.match(/\/((?:NBS|NIST).+)/)[1].gsub(".", " ")
+      ::Pubid::Nist::Identifier.parse(str).to_s
+    rescue ::Pubid::Core::Errors::ParseError
+      str.gsub(".", " ").sub(/^[\D]+/, &:upcase)
     end
 
     # @return [String]
-    def doi
-      url = @doc.location.reduce(nil) { |m, l| m || l.url.detect { |u| u.usage == "primary display" } }
-      id = url.content.match(/10\.6028\/.+/)[0]
+    def replace_wrong_doi(id)
       case id
-      when "10.6028/NBS.CIRC.sup" then "10.6028/NBS.CIRC.24e7sup"
-      when "10.6028/NBS.CIRC.supJun1925-Jun1926" then "10.6028/NBS.CIRC.24e7sup2"
-      when "10.6028/NBS.CIRC.supJun1925-Jun1927" then "10.6028/NBS.CIRC.24e7sup3"
-      when "10.6028/NBS.CIRC.24supJuly1922" then "10.6028/NBS.CIRC.24e6sup"
-      when "10.6028/NBS.CIRC.24supJan1924" then "10.6028/NBS.CIRC.24e6sup2"
+      when "NBS.CIRC.sup" then "NBS.CIRC.24e7sup"
+      when "NBS.CIRC.supJun1925-Jun1926" then "NBS.CIRC.24e7sup2"
+      when "NBS.CIRC.supJun1925-Jun1927" then "NBS.CIRC.24e7sup3"
+      when "NBS.CIRC.24supJuly1922" then "NBS.CIRC.24e6sup"
+      when "NBS.CIRC.24supJan1924" then "NBS.CIRC.24e6sup2"
       else id
       end
     end
 
+    def parse_doi
+      url = @doc.location.reduce(nil) { |m, l| m || l.url.detect { |u| u.usage == "primary display" } }
+      id = remove_doi_prefix(url.content)
+      replace_wrong_doi(id)
+    end
+
+    def remove_doi_prefix(id)
+      id.match(/10\.6028\/(.+)/)[1]
+    end
+
     # @return [Array<RelatonBib::TypedTitleString>]
     def parse_title
       title = @doc.title_info.reduce([]) do |a, ti|
@@ -180,7 +190,8 @@ module RelatonNist
         item.other_type
       else
         item.name[0].name_part[0].content
-      end
+      end => id
+      replace_wrong_doi remove_doi_prefix(id)
     end
 
     def parse_place

data/lib/relaton_nist/version.rb

@@ -1,3 +1,3 @@
 module RelatonNist
-  VERSION = "1.19.3".freeze
+  VERSION = "1.19.5".freeze
 end

data/lib/relaton_nist.rb

@@ -1,8 +1,10 @@
 require "singleton"
+require "pubid"
 require "relaton/index"
 require "relaton_bib"
 require "relaton_nist/version"
 require "relaton_nist/util"
+# require "relaton_nist/pubid"
 require "relaton_nist/nist_bibliography"
 require "relaton_nist/data_fetcher"
 require "relaton_nist/pubs_export"

data/relaton_nist.gemspec

@@ -25,6 +25,7 @@ Gem::Specification.new do |spec|
 
   spec.add_dependency "base64"
   spec.add_dependency "loc_mods", "~> 0.2.0"
+  spec.add_dependency "pubid", "~> 0.1.0"
   spec.add_dependency "relaton-bib", "~> 1.19.2"
   spec.add_dependency "relaton-index", "~> 0.2.0"
   spec.add_dependency "rubyzip"

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: relaton-nist
 version: !ruby/object:Gem::Version
-  version: 1.19.3
+  version: 1.19.5
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2024-09-06 00:00:00.000000000 Z
+date: 2024-10-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: base64
@@ -38,6 +38,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: 0.2.0
+- !ruby/object:Gem::Dependency
+  name: pubid
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.1.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.1.0
 - !ruby/object:Gem::Dependency
   name: relaton-bib
   requirement: !ruby/object:Gem::Requirement