filesanitize.rb

# encoding: utf-8
# clean filenames for OS X, Windows and Linux
 
class Filename
  CHARACTER_FILTER = /[\x00-\x1F\/\\:\*\?\"<>\|]/u
  UNICODE_WHITESPACE = /[[:space:]]+/u
  
  def initialize(filename)
    @raw = filename.to_s.freeze
  end
  
  # strip whitespace on beginning and end
  # collapse intra-string whitespace into single spaces
  def normalize
    @normalized ||= @raw.strip.gsub(UNICODE_WHITESPACE,' ')
  end
  
  # remove characters that aren't allowed cross-OS
  def sanitize
    @sanitized ||= normalize.gsub(CHARACTER_FILTER,'')
  end
  
  # normalize unicode string and cut off at 255 characters
  def truncate
    @truncated ||= sanitize.chars.to_a.slice(0..254).join
  end
  
  # convert back from multibyte string
  def to_s
    truncate
  end
end

x = Filename.new("   ühere\\ x\tis a   totally:unsafeöø.name!!ø ").to_s
puts "YAY!" if x == "ühere x is a totallyunsafeöø.name!!ø"


# rvm 1.8.7,1.9.2,1.9.3,2.0.0 do ruby fn.rb
# YAY!
# YAY!
# YAY!
# YAY!