termium 0.1.0 → 0.1.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 9f242a07fdedab976941acc1580ca44d301ab96e3b3c5f70bf17704fabf39f0e
-  data.tar.gz: 9b650e4b6d9786014c09ada236f5d369c5a9138616d77083b14f121b881cb2eb
+  metadata.gz: b6e0e5aefeb1e77b272a066f4e530779c522fa1d58d1e4890beda19db2e2c203
+  data.tar.gz: eccef83149a4c29c751ada19d6d1db07fa961d983322986b49ff40d33954255d
 SHA512:
-  metadata.gz: 98d930ad1aab03405b14561d17cd14b76531d3a65381362b3a2822e1c8f4244bb49a9f4074a3d189fc8049a6bccbd43d40d02f9766ec00bfe78070c1950c0c5c
-  data.tar.gz: 904748c40e59b69efeb8d2c9355ac9523b005f39b89ca03f2fac18384780e2bba4e87d2b5a441b9c725ed2c9aff426a4c588ef4cb72956199fa92aa371e099c0
+  metadata.gz: 2feb9eef3d42a4f9d07e710c2608ab00bb903484a98ba890ed259a7e7cbc7dbcea68b00bf69e493d9921f4362b79bfc314e961ccb147a559c5e835b287942b62
+  data.tar.gz: 70544c5334f85c6b8a7d7087840169a98578d57249d6ac3eb29412c70ae04759f0f28bc488ebba333a1ac3dbeac55122e44f9dd63baa74757baf9f08ddb3f7ee

data/lib/termium/cli.rb

@@ -5,53 +5,59 @@ require_relative "../termium"
 module Termium
   # Command-line interface
   class Cli < Thor
-    desc "convert", "Convert Termium entries into a Glossarist dataset"
+    desc "convert", "Convert TERMIUM entries into a Glossarist dataset"
 
     option :input_file, aliases: :i, required: true, desc: "Path to TERMIUM Plus XML extract"
     option :output_file, aliases: :o, desc: "Output file path"
 
-    def input_file_as_path
-      input_path = Pathname.new(Dir.pwd).join(Pathname.new(options[:input_file]))
+    no_commands do
+      def input_file_as_path(input_file)
+        input_path = Pathname.new(Dir.pwd).join(Pathname.new(input_file))
 
-      unless input_path.exist?
-        error "Input file `#{options[:input_file]}` does not exist."
-        exit 1
-      end
-
-      input_path
-    end
+        unless input_path.exist?
+          error "TERMIUM export file `#{options[:input_file]}` does not exist."
+          exit 1
+        end
 
-    def output_path_ready
-      output_path = options[:output_file]
-      output_path ||= input_path.dirname.join(input_path.basename(input_path.extname))
+        input_path
+      end
 
-      output_path = Pathname.new(Dir.pwd).join(Pathname.new(output_path))
+      def output_dir_as_path(output_path, input_path)
+        output_path ||= input_path.dirname.join(input_path.basename(input_path.extname))
 
-      if output_path.exist?
-        puts "Using existing directory: #{output_path.relative_path_from(Dir.pwd)}"
-      else # and is directory
-        puts "Created directory: #{output_path.relative_path_from(Dir.pwd)}"
-        output_path.mkdir
+        output_path = Pathname.new(Dir.pwd).join(output_path)
+        create_or_use_output_path(output_path)
+        output_path
       end
 
-      output_path
+      def create_or_use_output_path(output_path)
+        output_path_rel = output_path.relative_path_from(Dir.pwd)
+        if output_path.exist?
+          puts "Using existing directory: #{output_path_rel}"
+        else # and is directory
+          puts "Created directory: #{output_path_rel}"
+          output_path.mkdir
+        end
+      end
     end
 
     def convert
-      input_path = input_file_as_path
-      puts "Reading input file: #{input_path.relative_path_from(Dir.pwd)}"
+      input_path = input_file_as_path(options[:input_file])
+
+      puts "Reading TERMIUM export file: #{input_path.relative_path_from(Dir.pwd)}"
       termium_extract = Termium::Extract.from_xml(IO.read(input_path.expand_path))
 
-      puts "Size of dataset: #{termium_extract.core.size}"
+      puts "Size of TERMIUM dataset: #{termium_extract.core.size}"
 
       puts "Converting to Glossarist..."
       glossarist_col = termium_extract.to_concept
       # pp glossarist_col.first
 
-      output_path = output_path_ready
-      puts "Writing Glossarist dataset..."
+      output_path = output_dir_as_path(options[:output_file], input_path)
+      puts "Writing Glossarist dataset to: #{output_path.relative_path_from(Dir.pwd)}"
       glossarist_col.save_to_files(output_path.expand_path)
-      puts "Written Glossarist dataset to: #{output_path.relative_path_from(Dir.pwd)}"
+      puts "Done."
+      exit 0
     end
 
     def method_missing(*args)

data/lib/termium/core.rb

@@ -28,10 +28,17 @@ module Termium
 
     # TODO: In Termium XML, each definition per lang or note can be linked to a
     # particular source via the sourceRef number.
+    # We should utilize "source" order ID in the Glossarist object:
+    # <source order="1" details="ISO-2382-6 * 1987 *  *  * " />
+    # <source order="2"
+    #   details="Ranger, Natalie * 2006 * Bureau de la traduction..." />
     def concept_sources
       source.map(&:to_concept_source)
     end
 
+    # TODO: Utilize "subject" in the Glossarist object:
+    # <subject abbreviation="YBB"
+    # details="Compartment - ISO/IEC JTC 1 Information Technology Vocabulary" />
     def to_concept
       concept = Glossarist::ManagedConcept.new(id: identification_number)
 

data/lib/termium/entry_term.rb

@@ -62,11 +62,17 @@ module Termium
       end
     end
 
+    def normative_status
+      return "deprecated" if deprecated
+
+      order == 1 ? "preferred" : "admitted"
+    end
+
     def to_h
       set = {
         "designation" => value,
         "type" => "expression",
-        "normative_status" => deprecated ? "deprecated" : "preferred"
+        "normative_status" => normative_status
       }
 
       set["geographical_area"] = geographical_area if geographical_area

data/lib/termium/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Termium
-  VERSION = "0.1.0"
+  VERSION = "0.1.2"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: termium
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.2
 platform: ruby
 authors:
 - Ribose
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2023-11-15 00:00:00.000000000 Z
+date: 2023-11-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: glossarist