metanorma-plugin-glossarist 0.2.1 → 0.2.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8f950715c1c6a808aa2b24e8cb855ef674799465bd7a57b1f04ac73d1e899a91
-  data.tar.gz: 4fb0bdfc76982a98749a03f8f89600cff06073fbe5a555d6ef2004d1fffda0b8
+  metadata.gz: daa2bea1e77bfd5f4bf2ec230949e1b553f977df76078f98cad24d24b471cd2c
+  data.tar.gz: 4f2387682fcca5c77a82a377ce7f739632adfa92656c289a69cd6b5e7e947724
 SHA512:
-  metadata.gz: 02f4df920b7fb3b3b67f647764b49bbfa4f276bb6aa8c8c803c3f55227d257d4cd5bcd1b77a88996ce6d96d1d453626bf2ab088e75c8629520b8976ff7fecf9d
-  data.tar.gz: 5c1a118c591f17783f7f54f9c66c316c3dd60b9034b8a23b904c22335a338a61d526c73d68d0f16a1c43c7d0520ac5469f32fc5da049004141cc42241769d968
+  metadata.gz: 2106de661a748c6f9026889236f3e6b53bc5cb1cbbad24ef0870210173f09ae0ba730afe2469bb7d55b57845ee5693efdee68c14efec608ae3c45961bcaac27b
+  data.tar.gz: c1280bae939146a0bda8c7c1904c9187dd1e922e9087cc14b5b1cd7b5a7fe43bc8b47fb5e87a921b1637f635a023de83824bc0896339b760efdc5fa4ed10c293

data/Gemfile

@@ -4,6 +4,11 @@ Encoding.default_internal = Encoding::UTF_8
 source "https://rubygems.org"
 git_source(:github) { |repo| "https://github.com/#{repo}" }
 
+gem "metanorma-standoc"
+gem "pry"
+gem "rake", "~> 12.0"
+gem "rspec", "~> 3.0"
+
 gemspec
 
 eval_gemfile("Gemfile.devel") rescue nil

data/README.md

@@ -138,6 +138,194 @@ The term bar is used as metasyntactic variables and placeholder names in compute
 The term foo is used as metasyntactic variables and placeholder names in computer programming
 ```
 
+The full concept model (a concept can have multiple localized concepts) is available through the block as follows:
+
+```ruby
+{
+  # UUIDs for localized concept mappings
+  "localized_concepts" => {
+    "eng" => "<uuid>",  # English concept UUID
+    "fre" => "<uuid>"   # French concept UUID
+  },
+
+  # Main concept term
+  "term" => "<string>",
+  
+  # Language-specific content (structure repeated for each language code)
+  "<language_code>" => {
+    "dates" => [],      # Array of relevant dates
+    "definition" => [   # Array of definition objects
+      {
+        "content" => "<string>"  # Definition text
+      }
+    ],
+    "examples" => [],   # Array of example objects
+    "id" => "<string>", # Concept ID
+    
+    "notes" => [        # Array of note objects
+      {
+        "content" => "<string>"  # Note text
+      }
+    ],
+    
+    "sources" => [      # Array of source objects
+      {
+        "origin" => {
+          "ref" => "<string>"    # Reference citation
+        },
+        "type" => "<string>",    # Source type (e.g. "lineage")
+        "status" => "<string>"   # Status (e.g. "identical")
+      }
+    ],
+    
+    "terms" => [        # Array of term objects
+      {
+        "type" => "<string>",              # Term type (e.g. "expression")
+        "normative_status" => "<string>",  # Status (e.g. "preferred")
+        "designation" => "<string>",       # Term text
+        "grammar_info" => [                # Array of grammar objects
+          {
+            "preposition" => boolean,
+            "participle" => boolean,
+            "adj" => boolean,
+            "verb" => boolean,
+            "adverb" => boolean,
+            "noun" => boolean,
+            "gender" => ["<string>"],      # Array of grammatical genders
+            "number" => ["<string>"]       # Array of grammatical numbers
+          }
+        ]
+      }
+    ],
+    
+    "language_code" => "<string>"  # ISO language code
+  }
+}
+```
+
+The language codes used are ISO 639-* 3-character codes, as described in the
+[Glossarist Concept model](https://github.com/glossarist/concept-model).
+
+An example of the full model (from ISO/IEC 2382:2015):
+
+```ruby
+{
+  "localized_concepts" => {
+    "eng" => "01134f51-b88c-5214-8909-5d271ea619cf",
+    "fre" => "f290a3af-f1b3-527a-9045-a2dfcc0caf5a"
+  },
+  "term" => "concept description",
+  "eng" => {
+    "dates" => [],
+    "definition" => [
+      {
+        "content" => "data structure describing the class of all known instances of a concept"
+      }
+    ],
+    "examples" => [],
+    "id" => "2122978",
+    "notes" => [
+      {
+        "content" => "concept description: term and definition standardized by ISO/IEC [ISO/IEC 2382-31:1997]."
+      },
+      {
+        "content" => "31.02.02 (2382)"
+      }
+    ],
+    "sources" => [
+      {
+        "origin" => {
+          "ref" => "ISO/IEC 2382-31:1997"
+        },
+        "type" => "lineage",
+        "status" => "identical"
+      },
+      {
+        "origin" => {
+          "ref" => "Ranger, Natalie * 2006 * Bureau de la traduction / Translation Bureau * Services linguistiques / Linguistic Services * Bur. dir. Centre de traduction et de terminologie / Dir's Office Translation and Terminology Centre * Div. Citoyenneté et Protection civile / Citizen. & Emergency preparedness Div. * Normalisation terminologique / Terminology Standardization"
+        },
+        "type" => "lineage",
+        "status" => "identical"
+      }
+    ],
+    "terms" => [
+      {
+        "type" => "expression",
+        "normative_status" => "preferred",
+        "designation" => "concept description",
+        "grammar_info" => [
+          {
+            "preposition" => false,
+            "participle" => false,
+            "adj" => false,
+            "verb" => false,
+            "adverb" => false,
+            "noun" => false,
+            "gender" => [],
+            "number" => ["singular"]
+          }
+        ]
+      }
+    ],
+    "language_code" => "eng"
+  },
+  "fre" => {
+    "dates" => [],
+    "definition" => [
+      {
+        "content" => "structure de données qui décrit la classe des instances connues d'un concept"
+      }
+    ],
+    "examples" => [],
+    "id" => "2122978",
+    "notes" => [
+      {
+        "content" => "description de concept : terme et définition normalisés par l'ISO/CEI [ISO/IEC 2382-31:1997]."
+      },
+      {
+        "content" => "31.02.02 (2382)"
+      }
+    ],
+    "sources" => [
+      {
+        "origin" => {
+          "ref" => "ISO/IEC 2382-31:1997"
+        },
+        "type" => "lineage",
+        "status" => "identical"
+      },
+      {
+        "origin" => {
+          "ref" => "Ranger, Natalie * 2006 * Bureau de la traduction / Translation Bureau * Services linguistiques / Linguistic Services * Bur. dir. Centre de traduction et de terminologie / Dir's Office Translation and Terminology Centre * Div. Citoyenneté et Protection civile / Citizen. & Emergency preparedness Div. * Normalisation terminologique / Terminology Standardization"
+        },
+        "type" => "lineage",
+        "status" => "identical"
+      }
+    ],
+    "terms" => [
+      {
+        "type" => "expression",
+        "normative_status" => "preferred",
+        "designation" => "description de concept",
+        "grammar_info" => [
+          {
+            "preposition" => false,
+            "participle" => false,
+            "adj" => false,
+            "verb" => false,
+            "adverb" => false,
+            "noun" => false,
+            "gender" => ["f"],
+            "number" => ["singular"]
+          }
+        ]
+      }
+    ],
+    "language_code" => "fre"
+  }
+}
+```
+
 ### Rendering a single term from loaded dataset
 
 This can be used to render a single concept from the loaded dataset. See [Loading Dataset](#loading-dataset) to load a dataset.

data/lib/liquid/drops/concepts_drop.rb

@@ -8,24 +8,20 @@ module Liquid
       # rubocop:disable Lint/MissingSuper
       def initialize(managed_concept_collection, filters = {})
         @concepts_collection = managed_concept_collection
-        @concepts_map = {}
-
-        filtered_concepts(@concepts_collection, filters).each do |concept|
-          @concepts_map[concept["term"]] = concept
-        end
+        @concepts = filtered_concepts(@concepts_collection, filters)
       end
       # rubocop:enable Lint/MissingSuper
 
       def concepts
-        @concepts_map
+        @concepts
       end
 
       def [](concept_name)
-        @concepts_map[concept_name]
+        @concepts.find { |c| c["term"] == concept_name }
       end
 
       def each(&block)
-        @concepts_map.values.each(&block)
+        @concepts.each(&block)
       end
 
       private

data/lib/metanorma/plugin/glossarist/dataset_preprocessor.rb

@@ -60,7 +60,9 @@ module Metanorma
           super
           @config = config
           @datasets = {}
+          @title_depth = { value: 2 }
           @rendered_bibliographies = {}
+          @seen_glossarist = []
         end
 
         def process(document, reader)
@@ -73,10 +75,19 @@ module Metanorma
           @config[:file_system] = file_system
 
           processed_doc = prepare_document(document, input_lines)
+          log(document, processed_doc.to_s) unless @seen_glossarist.empty?
           Asciidoctor::PreprocessorReader.new(document,
                                               processed_doc.to_s.split("\n"))
         end
 
+        protected
+
+        def log(doc, text)
+          File.open("#{doc.attr('docfile')}.glossarist.log.txt", "w:UTF-8") do |f|
+            f.write(text)
+          end
+        end
+
         private
 
         def prepare_document(document, input_lines, end_mark = nil)
@@ -109,11 +120,13 @@ module Metanorma
           elsif match = current_line.match(GLOSSARIST_BLOCK_REGEX)
             process_glossarist_block(document, liquid_doc, input_lines, match)
           else
+            @title_depth[:value] = current_line.sub(/ .*$/, "").size if /^==+ \S/.match?(current_line)
             liquid_doc.add_content(current_line)
           end
         end
 
         def process_dataset_tag(document, input_lines, liquid_doc, match)
+          @seen_glossarist << "x"
           context_names = prepare_dataset_contexts(document, match[1])
 
           dataset_section = <<~TEMPLATE
@@ -129,6 +142,7 @@ module Metanorma
         end
 
         def process_glossarist_block(document, liquid_doc, input_lines, match)
+          @seen_glossarist << "x"
           end_mark = input_lines.next
           path = relative_file_path(document, match[1])
 
@@ -160,6 +174,7 @@ module Metanorma
         end
 
         def process_render_tag(liquid_doc, match)
+          @seen_glossarist << "x"
           context_name, concept_name = match[1].split(",")
 
           liquid_doc.add_content(
@@ -168,6 +183,7 @@ module Metanorma
         end
 
         def process_import_tag(liquid_doc, match)
+          @seen_glossarist << "x"
           context_name = match[1].strip
           dataset = @datasets[context_name.strip]
 
@@ -179,17 +195,19 @@ module Metanorma
         end
 
         def process_bibliography(liquid_doc, match)
+          @seen_glossarist << "x"
           dataset_name = match[1].strip
           dataset = @datasets[dataset_name]
 
           liquid_doc.add_content(
-            dataset.concepts.map do |_concept_name, concept|
+            dataset.concepts.map do |concept|
               concept_bibliography(concept)
             end.compact.sort.join("\n"),
           )
         end
 
         def process_bibliography_entry(liquid_doc, match)
+          @seen_glossarist << "x"
           dataset_name, concept_name = match[1].split(",").map(&:strip)
           concept = @datasets[dataset_name][concept_name]
 
@@ -202,7 +220,7 @@ module Metanorma
           bibliography = concept["eng"]["sources"].map do |source|
             ref = source["origin"]["ref"]
 
-            next if @rendered_bibliographies[ref]
+            next if @rendered_bibliographies[ref] || ref.nil? || ref.empty?
 
             @rendered_bibliographies[ref] = ref.gsub(/[ \/:]/, "_")
             "* [[[#{@rendered_bibliographies[ref]},#{ref}]]]"
@@ -237,7 +255,7 @@ module Metanorma
 
         def concept_template(dataset_name, concept_name)
           <<~CONCEPT_TEMPLATE
-            ==== {{ #{dataset_name}['#{concept_name}'].term }}
+            #{"=" * (@title_depth[:value] + 1)} {{ #{dataset_name}['#{concept_name}'].term }}
             #{alt_terms(dataset_name, concept_name)}
 
             {{ #{dataset_name}['#{concept_name}']['eng'].definition[0].content }}
@@ -289,6 +307,7 @@ module Metanorma
         def sources(dataset_name, concept_name)
           <<~SOURCES
             {% for source in #{dataset_name}['#{concept_name}']['eng'].sources %}
+            {%- if source.origin.ref == nil or source.origin.ref == '' %}{% continue %}{% endif %}
             [.source]
             <<{{ source.origin.ref | replace: ' ', '_' | replace: '/', '_' | replace: ':', '_' }},{{ source.origin.clause }}>>
 

data/lib/metanorma/plugin/glossarist/version.rb

@@ -1,7 +1,7 @@
 module Metanorma
   module Plugin
     module Glossarist
-      VERSION = "0.2.1".freeze
+      VERSION = "0.2.3".freeze
     end
   end
 end

data/metanorma-plugin-glossarist.gemspec

@@ -24,10 +24,6 @@ Gem::Specification.new do |spec|
   spec.require_paths = ["lib"]
 
   spec.add_dependency "asciidoctor", "~> 2.0.0"
-  spec.add_dependency "glossarist", "~> 2.0.1"
+  spec.add_dependency "glossarist", "~> 2.0"
   spec.add_dependency "liquid", "~> 5"
-
-  spec.add_development_dependency "pry"
-  spec.add_development_dependency "rake", "~> 12.0"
-  spec.add_development_dependency "rspec", "~> 3.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: metanorma-plugin-glossarist
 version: !ruby/object:Gem::Version
-  version: 0.2.1
+  version: 0.2.3
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2024-05-30 00:00:00.000000000 Z
+date: 2024-11-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: asciidoctor
@@ -30,14 +30,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.0.1
+        version: '2.0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.0.1
+        version: '2.0'
 - !ruby/object:Gem::Dependency
   name: liquid
   requirement: !ruby/object:Gem::Requirement
@@ -52,48 +52,6 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '5'
-- !ruby/object:Gem::Dependency
-  name: pry
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: rake
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '12.0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '12.0'
-- !ruby/object:Gem::Dependency
-  name: rspec
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '3.0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '3.0'
 description: Metanorma plugin for glossarist
 email:
 - open.source@ribose.com