wordcuta 0.2.1 → 0.4.0

data/README.md

@@ -5,11 +5,21 @@
 * Install [wordcutw](https://github.com/veer66/wordcutw)
 * gem install wordcuta
 
+## Prerequisite
+
+### Dictionaries and rules
+
+```Shell
+wget https://codeberg.org/mekong-lang/mekong-lang-data/archive/main.tar.gz -O - | tar -xzvf -
+```
+
 ## Example
 
 ```Ruby
 require 'wordcuta'
-wc = WordcutA::Wordcut.new(WordcutA::DEFAULT_THAI_DICT_PATH)
+
+wc = WordcutA::Wordcut.new("mekong-lang-data/dictionaries/mixed/chamkho-dict.txt",
+                           "mekong-lang-data/cluster-rules/thai_cluster_rules.txt")
 p wc.put_delimiters('กากาก้า', '|')
 p wc.into_strings('กากา')
 p wc.into_ranges('กากา')

data/lib/wordcuta.rb

@@ -10,8 +10,9 @@ module WordcutFFI
   extend FFI::Library
 
   ffi_lib "wordcutw"
-  
+
   attach_function :wordcut_new_with_dict, [:string], :pointer
+  attach_function :wordcut_new_with_dict_and_cluster_rules, [:string, :string], :pointer
   attach_function :wordcut_into_text_ranges, [:pointer, :string, :pointer], :pointer
   attach_function :wordcut_into_strings, [:pointer, :string, :pointer], :pointer
   attach_function :wordcut_put_delimiters, [:pointer, :string, :string], :string
@@ -23,11 +24,17 @@ module WordcutA
   TextRange = Struct.new(:s, :e)
 
   DEFAULT_THAI_DICT_PATH = File.expand_path('../../data/thai-dix.txt', __FILE__)
-  
+
   class Wordcut
-    def initialize(dict_path)
-      @wordcut_p = FFI::AutoPointer.new(WordcutFFI.wordcut_new_with_dict(dict_path),
-                                        WordcutFFI.method(:delete_wordcut))
+    def initialize(dict_path, cluster_rule_path = nil)
+      if cluster_rule_path
+        @wordcut_p = FFI::AutoPointer.new(
+          WordcutFFI.wordcut_new_with_dict_and_cluster_rules(dict_path, cluster_rule_path),
+          WordcutFFI.method(:delete_wordcut))
+      else
+        @wordcut_p = FFI::AutoPointer.new(WordcutFFI.wordcut_new_with_dict(dict_path),
+                                          WordcutFFI.method(:delete_wordcut))
+      end
     end
 
     def into_ranges(text)

data/wordcuta.gemspec

@@ -1,13 +1,13 @@
 Gem::Specification.new do |s|
-  s.name = 'wordcuta'
-  s.version = '0.2.1'
-  s.authors = ['Vee Satayamas']
-  s.email = ['5ssgdxltv@relay.firefox.com']
-  s.licenses = ['LGPL-3.0']
-  s.description = "A word segmentation tools for ASEAN languages wrapper for Ruby"
-  s.homepage = "https://github.com/veer66/wordcuta"
+  s.name = "wordcuta"
+  s.version = "0.4.0"
+  s.authors = ["Vee Satayamas"]
+  s.email = ["vsatayamas@gmail.com"]
+  s.licenses = ["BSD-2-Clause"]
+  s.description = "A Ruby wrapper of wordcut-engine - a word segmentation tools for ASEAN languages, i.e. Khmer, Lao, Myanmar, and Thai"
+  s.homepage = "https://codeberg.org/mekong/wordcuta"
   s.required_ruby_version = Gem::Requirement.new(">= 2.3.0")
   s.summary = "A word segmentation tools for ASEAN languages wrapper for Ruby"
-  s.files = %w(README.md LICENSE Gemfile wordcuta.gemspec data/thai-dix.txt lib/wordcuta.rb)
-  s.add_dependency 'ffi', '1.15.5'
+  s.files = %w(README.md LICENSE wordcuta.gemspec lib/wordcuta.rb)
+  s.add_dependency "ffi", "~> 1.15"
 end

metadata

@@ -1,47 +1,44 @@
 --- !ruby/object:Gem::Specification
 name: wordcuta
 version: !ruby/object:Gem::Version
-  version: 0.2.1
+  version: 0.4.0
 platform: ruby
 authors:
 - Vee Satayamas
-autorequire: 
 bindir: bin
 cert_chain: []
-date: 2022-02-16 00:00:00.000000000 Z
+date: 1980-01-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: ffi
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.15.5
+        version: '1.15'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.15.5
-description: A word segmentation tools for ASEAN languages wrapper for Ruby
+        version: '1.15'
+description: A Ruby wrapper of wordcut-engine - a word segmentation tools for ASEAN
+  languages, i.e. Khmer, Lao, Myanmar, and Thai
 email:
-- 5ssgdxltv@relay.firefox.com
+- vsatayamas@gmail.com
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- Gemfile
 - LICENSE
 - README.md
-- data/thai-dix.txt
 - lib/wordcuta.rb
 - wordcuta.gemspec
-homepage: https://github.com/veer66/wordcuta
+homepage: https://codeberg.org/mekong/wordcuta
 licenses:
-- LGPL-3.0
+- BSD-2-Clause
 metadata: {}
-post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
@@ -56,8 +53,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.2.22
-signing_key: 
+rubygems_version: 3.7.1
 specification_version: 4
 summary: A word segmentation tools for ASEAN languages wrapper for Ruby
 test_files: []

data/Gemfile

@@ -1,5 +0,0 @@
-source :rubygems
-
-gemspec
-
-gem 'ffi'