Generate wide character detection code

full-width.rb

#!/usr/bin/ruby

require 'pp'

# http://unicode.org/Public/UNIDATA/EastAsianWidth.txt
data = File.read('EastAsianWidth.txt')
data.gsub!(/\s*#.*/, '')
data = data.split
data.reject! {|x| x !~ /;/}
data = data.select {|x| x =~ /;[FW]/}
data = data.collect do |line|
  if line =~ /^([\w]+);/
    [$1.hex]
  elsif line =~ /^([\w]+)\.\.([\w]+);/
    $1.hex.upto($2.hex).to_a
  end
end

current_range = nil
ranges = []
data.each do |range|
  range.each do |codepoint|
    if current_range && codepoint == (current_range[1] + 1)
      current_range[1] = codepoint
    else
      current_range = [codepoint, codepoint]
      ranges << current_range
    end
  end
end

ranges.each do |range|
  puts "c >= 0x%04X && c <= 0x%04X ||" % range
end
puts
ranges.each do |range|
  print("[\\u%04X-\\u%04X]|" % range)
end
puts