rudachi-rb 1.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: '0136994c8ace08ad3b922acc5599ad8b35d5c1f0387826e24c4d4ad9efd26fe8'
+  data.tar.gz: 3fbbf61bdf9db17a35efad222c211aacc741f7b6b1221edf044fd19984695ff6
+SHA512:
+  metadata.gz: 810dc0f49f4ee9d8dbceb25847c623c0cb84618d74c25542b0632c45a2f021b4ca127d7649e4bb129bcef6b8e7a7efa7e551865410f0b605a079eabb6375c582
+  data.tar.gz: 4135642a21814f82103eaf3c107ae4a5f1b2de3728f06ca9159b9286057fae5631fa36ee77073106ed0b815962a68d5162c94f913884fab57aee754e732feba9

data/README.md

@@ -0,0 +1,88 @@
+# Rudachi-rb
+Ruby wrapper for [Sudachi](https://github.com/WorksApplications/Sudachi).<br>
+([rudachi](https://github.com/SongCastle/rudachi) for Ruby)
+
+#### Text
+```rb
+Rudachi::TextParser.parse('東京都へ行く')
+=> "東京都\t名詞,固有名詞,地名,一般,*,*\t東京都\nへ\t助詞,格助詞,*,*,*,*\tへ\n行く\t動詞,非自立可能,*,*,五段-カ行,終止形-一般\t行く\nEOS\n"
+```
+
+#### File
+```rb
+File.open('input.txt', 'w') { |f| f << '東京都へ行く' }
+Rudachi::FileParser.parse('input.txt')
+=> "東京都\t名詞,固有名詞,地名,一般,*,*\t東京都\nへ\t助詞,格助詞,*,*,*,*\tへ\n行く\t動詞,非自立可能,*,*,五段-カ行,終止形-一般\t行く\nEOS\n"
+```
+
+#### IO
+```rb
+Rudachi::StreamParser.parse(StringIO.new('東京都へ行く'))
+=> "東京都\t名詞,固有名詞,地名,一般,*,*\t東京都\nへ\t助詞,格助詞,*,*,*,*\tへ\n行く\t動詞,非自立可能,*,*,五段-カ行,終止形-一般\t行く\nEOS\n"
+```
+
+#### With [some options](https://github.com/WorksApplications/Sudachi#options)
+```rb
+Rudachi::TextParser.new(o: 'output.txt', m: 'A').parse('東京都へ行く')
+File.read('output.txt')
+=> "東京\t名詞,固有名詞,地名,一般,*,*\t東京\n都\t名詞,普通名詞,一般,*,*,*\t都\nへ\t助詞,格助詞,*,*,*,*\tへ\n行く\t動詞,非自立可能,*,*,五段-カ行,終止形-一般\t行く\nEOS"
+```
+
+## Requirements
+
+- Ruby 2.3.0 or newer
+- [rjb](https://github.com/arton/rjb) 1.1.1 or newer
+- [Sudachi](https://github.com/WorksApplications/Sudachi)
+
+For JRuby, please check [rudachi](https://github.com/SongCastle/rudachi).
+
+## Installation
+
+1. Install JAR and dictionary of Sudachi ([Details](https://github.com/WorksApplications/Sudachi/blob/develop/docs/tutorial.md#linux-%E3%81%AE%E5%A0%B4%E5%90%88))
+
+##### Install the Sudachi JAR file
+```sh
+$ wget https://github.com/WorksApplications/Sudachi/releases/download/v0.5.3/sudachi-0.5.3-executable.zip
+$ unzip sudachi-0.5.3-executable.zip
+$ ls sudachi-0.5.3
+LICENSE-2.0.txt  README.md  javax.json-1.1.jar	jdartsclone-1.2.0.jar  licenses  sudachi-0.5.3.jar  sudachi.json  sudachi_fulldict.json
+```
+
+##### Install the Sudachi dictionary
+```sh
+$ wget http://sudachi.s3-website-ap-northeast-1.amazonaws.com/sudachidict/sudachi-dictionary-latest-full.zip
+$ unzip -j -d sudachi-dictionary-latest-full sudachi-dictionary-latest-full.zip
+$ mv sudachi-dictionary-latest-full/system_full.dic sudachi-dictionary-latest-full/system_core.dic
+$ ls sudachi-dictionary-latest-full
+LEGAL  LICENSE-2.0.txt	system_core.dic
+```
+
+2. Install Rudachi
+
+```rb
+# Gemfile
+gem 'rudachi-rb'
+```
+
+Then run `bundle install`.
+
+3. Initialize Rudachi
+
+```rb
+require 'rudachi/rb'
+
+Rudachi.configure do |config|
+  config.jar_path = 'sudachi-0.5.3/sudachi-0.5.3.jar'
+end
+
+Rudachi::Option.configure do |config|
+  config.p = 'sudachi-dictionary-latest-full'
+end
+```
+
+4. Did it !!
+
+```rb
+Rudachi::TextParser.parse('こんにちは世界')
+=> "こんにちは\t感動詞,一般,*,*,*,*\t今日は\n世界\t名詞,普通名詞,一般,*,*,*\t世界\nEOS\n"
+```

data/lib/rudachi/rb/dependencies.rb

@@ -0,0 +1,41 @@
+require 'rjb'
+require 'rudachi/lazy_load'
+
+Rjb::load(Rudachi.jar_path.to_s)
+
+module Rudachi
+  module RB
+    module Java
+      String                = Rjb::import('java.lang.String')
+      System                = Rjb::import('java.lang.System')
+      PrintStream           = Rjb::import('java.io.PrintStream')
+      ByteArrayInputStream  = Rjb::import('java.io.ByteArrayInputStream')
+      ByteArrayOutputStream = Rjb::import('java.io.ByteArrayOutputStream')
+      UTF_8                 = Rjb::import('java.nio.charset.StandardCharsets').UTF_8
+      SudachiCommandLine    = Rjb::import('com.worksap.nlp.sudachi.SudachiCommandLine')
+
+      Rudachi::LazyLoad.run_load_hooks(:java_rb, self)
+    end
+
+    module StreamProcessor
+      class InvalidInclusion < StandardError; end
+
+      TERM = ?\n
+
+      def self.included(base)
+        raise InvalidInclusion unless base.ancestors.include?(TextParser)
+      end
+
+      def parse(io)
+        ret = []
+        while data = io.gets
+          ret << super(data).strip
+        end
+
+        ret.join(TERM)
+      end
+
+      Rudachi::LazyLoad.run_load_hooks(:stream_processor, self)
+    end
+  end
+end

data/lib/rudachi/rb/loader.rb

@@ -0,0 +1,10 @@
+module Rudachi
+  module RB
+    module Loader
+      def load!
+        raise Rudachi::Loader::UnavailableError, 'ruby_required' if jruby?
+        require 'rudachi/rb/dependencies'
+      end
+    end
+  end
+end

data/lib/rudachi/rb/version.rb

@@ -0,0 +1,5 @@
+module Rudachi
+  module RB
+    VERSION = '1.0.0'
+  end
+end

data/lib/rudachi/rb.rb

@@ -0,0 +1,16 @@
+require 'rudachi'
+require 'rudachi/rb/loader'
+
+module Rudachi
+  module Loader
+    class << self
+      prepend RB::Loader
+    end
+  end
+
+  module Java
+    LazyLoad.on_load(:java_rb) do |mod|
+      include mod
+    end
+  end
+end

data/lib/rudachi-rb.rb

@@ -0,0 +1 @@
+require 'rudachi/rb'

metadata

@@ -0,0 +1,76 @@
+--- !ruby/object:Gem::Specification
+name: rudachi-rb
+version: !ruby/object:Gem::Version
+  version: 1.0.0
+platform: ruby
+authors:
+- SongCastle
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2022-04-30 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: rjb
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.1.1
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.1.1
+- !ruby/object:Gem::Dependency
+  name: rudachi
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.4.0
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 1.4.0
+description: A Ruby wrapper for Sudachi.
+email: "-"
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- README.md
+- lib/rudachi-rb.rb
+- lib/rudachi/rb.rb
+- lib/rudachi/rb/dependencies.rb
+- lib/rudachi/rb/loader.rb
+- lib/rudachi/rb/version.rb
+homepage: https://github.com/SongCastle/rudachi-rb
+licenses:
+- MIT
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '2.3'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.0.3
+signing_key: 
+specification_version: 4
+summary: A Ruby wrapper for Sudachi
+test_files: []