Decode base64 string and write to file

Question

I'm trying to read file which contains encoded base64 string and write decoded output into another file. My Input.txt contains a base64 string, something like:

PD94bWwgdmVyc2lvbj0iMS4wIiBlbmNvZGluZz0iVVRGLTgiPz48cmV2aWV3LWNhc2UgY3JlYXRl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\r\nIG5hbWU9IlBob25lTnVtYmVyIiB2YWx1

This string is created on server side with Java apache codec.binary.Base64 library. This string is captured with Fiddler when two different web services communicates with each other. Sometimes I have no access to the another web-service, that is why I sniff messages between services. In addition I use Ruby to automate some routine tasks and decided this time to use Ruby again. For encoding captured base64 string I use next snippet of code:

require "base64"

content = File.read('Input.txt')
decode_base64_content = Base64.decode64(content) 
File.open("Output.txt", "wb") do |f|
  f.write(decode_base64_content)
end 

But output looks malformed, like <?xml version="1.0" encoding="UTF-8"?><review-case create®vFFSТ#2фЦ"у#B“ЈCЈS"7—7FVУТ%G&f–wW&хFVЧЖFUфЦзnagement_v5.1" ba and so on. Can you please advise on what I'm doing wrong? I use Ruby 1.9.3 on Windows 7 and Ubuntu 12.04.

Answer 1

I do not know how you manage to do this, but the line endings \r\n in your string seem to be there as 4-byte character sequences, not as 2-byte escaped CRLF. If I copy your file into a ruby string with single ticks:

unescaped='PD94bWwgdmVyc2lvbj0iMS4wIiBlbmNvZGluZz0iVVRGLTgiPz48cmV2aWV3LWNhc2UgY3JlYXRl\r\nZGF0ZT0iMTMvTWFyLzIwMTQgMDk6MDQ6NTEiIHN5c3RlbT0iVHJhZmlndXJhX1RlbXBsYXRlX01h\r\nbmFnZW1lbnRfdjUuMSIgYmF0Y2hpZD0iMCIgdHJhbnNhY3Rpb25ubz0iMSIgYmF0Y2huYW1lPSJH'
Base64.decode64(unescaped)
#=> garbled text for every second line

if I do the same with double quotes (which respect the escape sequences):

escaped="PD94bWwgdmVyc2lvbj0iMS4wIiBlbmNvZGluZz0iVVRGLTgiPz48cmV2aWV3LWNhc2UgY3JlYXRl\r\nZGF0ZT0iMTMvTWFyLzIwMTQgMDk6MDQ6NTEiIHN5c3RlbT0iVHJhZmlndXJhX1RlbXBsYXRlX01h\r\nbmFnZW1lbnRfdjUuMSIgYmF0Y2hpZD0iMCIgdHJhbnNhY3Rpb25ubz0iMSIgYmF0Y2huYW1lPSJH"
Base64.decode64(escaped)
#=> all is well that ends well

Therefore the problem seems to occur when you write the file. It can be amended in Ruby though:

unescaped='PD94bWwgdmVyc2lvbj0iMS4wIiBlbmNvZGluZz0iVVRGLTgiPz48cmV2aWV3LWNhc2UgY3JlYXRl\r\nZGF0ZT0iMTMvTWFyLzIwMTQgMDk6MDQ6NTEiIHN5c3RlbT0iVHJhZmlndXJhX1RlbXBsYXRlX01h\r\nbmFnZW1lbnRfdjUuMSIgYmF0Y2hpZD0iMCIgdHJhbnNhY3Rpb25ubz0iMSIgYmF0Y2huYW1lPSJH'
Base64.decode64(unescaped)
escaped=unescaped.gsub('\\r', "\r").gsub('\\n', "\n")
Base64.decode64(escaped)
#=> now you should be fine again

but of course the correct solution would be to store the file correctly.

Given your current file the following should work:

require "base64"

content = File.read('Input.txt')
content.gsub!('\\r', "\r")
content.gsub!('\\n', "\n")
decode_base64_content = Base64.decode64(content) 
File.open("Output.txt", "wb") do |f|
  f.write(decode_base64_content)
end

Please do post some output if it does not.