possible_fix.rb

  def self.binary?(file)
    s = File.read(file, 1024) or return false
    set_encoding(s)
    begin
      if s[0, 2] == Marshal.dump('')[0, 2] then
        true
      elsif file =~ /erb\.rb$/ then
        false
      elsif s.scan(/<%|%>/).length >= 4 || s.index("\x00") then
        true
      elsif 0.respond_to? :fdiv then
        s.count("^ -~\t\r\n").fdiv(s.size) > 0.3
      else # HACK 1.8.6
        (s.count("^ -~\t\r\n").to_f / s.size) > 0.3
      end
    rescue ArgumentError => e
      # puts file
      # puts e
      # puts caller
      true
    end
  end

show_parsing_error.rb

file = "/Users/stephen/.rvm/rubies/ruby-1.9.2-p0/lib/ruby/1.9.1/irb/lc/ja/help-message"
size = File::Stat.new(file).size
puts "file: #{File.basename(file)}, size (bytes) #{size}"
puts "reading 1024 bytes into var: f"
f = File.read(file, 1024)
puts "f.encoding: #{f.encoding}"
puts "f.valid_encoding: #{f.valid_encoding?}"
if /coding[=:]\s*([^\s;]+)/i =~ f[%r"\A(?:#!.*\n)?.*\n"]
  if enc = ::Encoding.find($1)
    puts "f.force_encoding(#{enc})"
    f.force_encoding(enc)
  end
end
puts "f.encoding: #{f.encoding}"
puts "f.valid_encoding: #{f.valid_encoding?}"
puts
f2 = ""
f2.force_encoding(enc)
puts "f2.length (chars): #{f2.length}"
puts "f2.encoding: #{f2.encoding}"
puts "f2.valid_encoding: #{f2.valid_encoding?}"
begin
  f.each_char { |c| f2 << c }
rescue ArgumentError
end
puts "f2.length (chars): #{f2.length}"
puts "f2.encoding: #{f2.encoding}"
puts "f2.valid_encoding: #{f2.valid_encoding?}"