Extract stream blocks from a PDF

extract.rb

# Usage: ruby extract.rb PDF...
#
# Extracts all 'stream' blocks from the PDF. This includes all images embedded
# in it. All of the output files have a '.bin' extension. Sometimes you can see
# what they are by doing something like this:
#
#   $ ruby extract.rb example.pdf
#   $ file *.bin

def main
  namer = Namer.new
  ARGV.each do |arg|
    begin
      process(arg, namer)
    rescue => e
      puts "#{arg}: (#{e.class}) #{e}"
    end
  end
end

def process(f, namer)
  all = File.open(f, "rb") { |f| f.read }
  streams = all.split("\nstream\n")
  prev = ""
  streams.drop(1).each do |stream|
    good, rest = stream.split("\nendstream\n", 2)
    if rest.nil?
      prev = prev + good
      next
    end
    good = prev + good
    prev = ""
    name = namer.name("bin")
    puts "#{f} -> #{name}"
    File.write(name, good)
  end
end

class Namer
  def initialize
    @t = Time.now.to_i
    @n = 0
  end

  def name(ext)
    @n += 1
    "#@t-#@n.#{ext}"
  end
end

main