unicode-display_width 0.1.0 → 0.1.1

data/.gemspec

@@ -7,8 +7,8 @@ Gem::Specification.new do |s|
   s.version     = Unicode::DisplayWidth::VERSION
   s.authors     = ["Jan Lelis"]
   s.email       = "mail@janlelis.de"
-  s.homepage    = "http://github.com/janlelis/unicode-display_size"
-  s.summary = "Support for east_asian_width String sizes."
+  s.homepage    = "http://github.com/janlelis/unicode-display_width"
+  s.summary = "Support for east_asian_width string widths."
   s.description =  "This gem adds String#display_size to get the display size of a string using EastAsianWidth.txt."
   s.required_rubygems_version = ">= 1.3.6"
   s.files = Dir.glob(%w[{lib,test}/**/*.rb bin/* [A-Z]*.{txt,rdoc} data/* ext/**/*.{rb,c} **/deps.rip]) + %w{Rakefile .gemspec}

data/README.rdoc

@@ -1,26 +1,36 @@
 == Description
-A early draft of a way to determine the size of the characters using EastAsianWidth.txt, based on the very early draft of a [Ruby interface to UnicodeData.txt]{https://github.com/runpaint/unicode-data} by runpaint.
+
+An early draft of a way to determine the size of the characters using <tt>EastAsianWidth.txt</tt>, based on the very early draft of a {Ruby interface to UnicodeData.txt}[https://github.com/runpaint/unicode-data] by runpaint.
 
 == Install
+
 Install the gem with
 
   gem install unicode-display_width
 
 == Usage
+
   require 'unicode/display_width'
 
 The easy way is to use the String#display_size method:
   "⚀".display_width #=> 1
+  '一'.display_width #=> 2
 
-To obtain more detailed data, you can use the Codepoint struct (c is the codepoint in integer format):
-  Unicode::DiplayWidth.codepoint( c ).width
+To obtain more detailed data, you can use the following syntax:
+  Unicode::DiplayWidth.codepoint( c )
 
 == Bugs / TODO
-It does not work correctly for all characters ;). For example, there is a class of unicode characters, which is not exactly defined to be narrow or full-width ("ambiguous"). This gem currently gives them a size of 1, which is not always correct.
+
+Since it is not tested extensively, it does probably not work correctly for all characters... For example, there is a class of unicode characters, which is not exactly defined to be narrow or full-width ("ambiguous"). This gem currently gives them a size of 1, which is not always correct.
+
+Furthermore, the index data structure can be improved.
 
 == Copyright
+
 Copyright (c) 2011 Jan Lelis, http://rbjl.net, released under the MIT license.
-Copyright (c) 2009 Run Paint Run Run
-EastAsianWidth.txt: Copyright (c) 1991-2010 Unicode, Inc.
+
+Contains code by runpaint:   Copyright (c) 2009 Run Paint Run Run
+
+Contains EastAsianWidth.txt: Copyright (c) 1991-2010 Unicode, Inc.
 
 J-_-L

data/lib/unicode/display_width.rb

@@ -1,68 +1,62 @@
 # encoding: utf-8
 module Unicode; end
-
 module Unicode::DisplayWidth
-  VERSION = '0.1.0'
+  VERSION = '0.1.1'
+end
 
+class << Unicode::DisplayWidth
   DATA_DIR = File.join(File.dirname(__FILE__), '../../data/')
-  FIELDS = [:codepoint, :width, :name]
-  INDEX_FILE = DATA_DIR + 'EastAsianWidth.index'
+  TABLE_FILE = DATA_DIR + 'EastAsianWidth.index'
   DATA_FILE  = DATA_DIR + 'EastAsianWidth.txt'
 
-  def self.data
-    @@data ||= File.open DATA_FILE
-  end
-
-  def self.offsets
-    @@offsets ||= Marshal.load File.respond_to?(:binread) ? File.binread(INDEX_FILE) : File.read(INDEX_FILE)
+  # only needed for building the index
+  def data
+    @data ||= File.open DATA_FILE
   end
 
-  class Codepoint < Struct.new(*FIELDS)
-    def initialize(*args)
-      super
-      self.codepoint = self.codepoint.to_i(16) if self.codepoint && self.codepoint !~ /\.\./
-                                                  # TODO cleaner    # FIXME ranges
+  def table
+    if @table
+      @table
+    else
+      build_table unless File.file?(TABLE_FILE)
+      @table = Marshal.load File.respond_to?(:binread) ? File.binread(TABLE_FILE) : File.read(TABLE_FILE)
     end
-
-    def self.from_line(line)
-      line =~ /(.*);(.*) # (.*)$/
-        raise 'BUG' unless line
-      new $1,$2,$3
-    end
-  end
-
-  def self.line(n)
-    data.rewind
-    offset = offsets[n] or raise ArgumentError
-    data.seek offset
-    data.readline.chomp
-  end
-
-  def self.codepoint(n)
-    Codepoint.from_line line(n)
   end
 
-  def self.valid_index?
-    !!offsets rescue false
+  def codepoint(n)
+    n = n.to_s.unpack('U')[0] unless n.is_a? Integer
+    table[n] or raise ArgumentError
   end
+  alias width codepoint
+  alias of    codepoint
 
-  def self.build_index
+  def build_table
     data.rewind
-    offsets = {}
-    dir = File.dirname INDEX_FILE
+    table = {}
+    dir = File.dirname TABLE_FILE
     Dir.mkdir(dir) unless Dir.exists?(dir)
-    data.lines.map do |line|
-      offsets[Codepoint.from_line(line).codepoint] = data.pos - line.size
-    end
-    File.open(INDEX_FILE, 'wb') { |f| Marshal.dump(offsets, f) }
+    data.lines.each{ |line|
+      line =~ /^(.*);(.*) # .*$/
+      if $1 && $2
+        cps, width = $1, $2
+        if cps['..']
+          range = Range.new *cps.split('..').map{ |cp| cp.to_i(16) }
+          range.each{ |cp| table[ cp ] = width.to_sym }
+        else
+          table[ cps.to_i(16) ] = width.to_sym
+        end
+      end
+
+    }
+    File.open(TABLE_FILE, 'wb') { |f| Marshal.dump(table, f) }
   end
 end
 
 class String
-  def display_width(ambiguous=1)
+  def display_width(ambiguous = 1)
     #codepoints.inject(0){ |a,c|
     unpack('U*').inject(0){ |a,c|
-      width = case Unicode::DisplayWidth.codepoint(c).width
+      width = case Unicode::DisplayWidth.codepoint(c).to_s
               when *%w[F W]
                 2
               when *%w[N Na H]

metadata

@@ -1,13 +1,12 @@
 --- !ruby/object:Gem::Specification 
 name: unicode-display_width
 version: !ruby/object:Gem::Version 
-  hash: 27
   prerelease: false
   segments: 
   - 0
   - 1
-  - 0
-  version: 0.1.0
+  - 1
+  version: 0.1.1
 platform: ruby
 authors: 
 - Jan Lelis
@@ -15,7 +14,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-01-03 00:00:00 +01:00
+date: 2011-01-05 00:00:00 +01:00
 default_executable: 
 dependencies: []
 
@@ -38,7 +37,7 @@ files:
 - Rakefile
 - .gemspec
 has_rdoc: true
-homepage: http://github.com/janlelis/unicode-display_size
+homepage: http://github.com/janlelis/unicode-display_width
 licenses: 
 - MIT
 post_install_message: 
@@ -51,7 +50,6 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
-      hash: 3
       segments: 
       - 0
       version: "0"
@@ -60,7 +58,6 @@ required_rubygems_version: !ruby/object:Gem::Requirement
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
-      hash: 23
       segments: 
       - 1
       - 3
@@ -72,6 +69,6 @@ rubyforge_project:
 rubygems_version: 1.3.7
 signing_key: 
 specification_version: 3
-summary: Support for east_asian_width String sizes.
+summary: Support for east_asian_width string widths.
 test_files: []