I need to get the €uro sign converted to IKw=, and it is supposed to be a base64 encoding

Question

I have to send the €uro sign into an SMS. I was given some steps to do so, and they are:

• Convert the € sign to hexadecimal, which is: 20AC
• Encode the 20AC into base64, which should be: IKw=

But when I do so, with any online tool I find, I always get MjBBQw==, which is the same python returns.

So I supose I am missing some kind of character encoding between the hexadecimal and the base64.

The Python code I have is as follow:

def encodeGSM7Message( text ):
     text = unicode( text, 'UTF-8' )
     hex_text = ''.join( [ hex( ord( c ) ).rstrip('L').lstrip('0x').upper() for c in text ] )
     return  base64.b64encode( hex_text )

print encodeGSM7Message( '€' ), 'IKw='

This thing should print IKw= IKw=, but it gets to MjBBQw== IKw=.

As another example, they added Ñ to the string, so I also have an extra code line as follows:

print encodeGSM7Message( '€ÑÑ' ), 'IKwA0QDR'

But instead of printing IKwA0QDR IKwA0QDR, which should be the spected behavior, it ends printing MjBBQ0QxRDE= IKwA0QDR

Any idea about what I am missing, or what kind of unicode conversion should be done to get the expected result?

Answer 1

Try this:

# -*- coding: utf-8 -*-

def encodeGSM7Message(s):
  return base64.b64encode( s.decode('utf8').encode('utf-16-be') )

euro = '€'

print encodeGSM7Message(euro)

Note the coding: utf-8 makes the euro variable utf-8 encoded, which is why we have to .decode('utf8') in the encodeGSM7Message routine.

Answer 2

Working backwards from the result, they want utf-16 big-endian encoding.

>>> base64.encodebytes('€'.encode('utf-16be'))
b'IKw=\n'

Answer 3

You don't need the explicit conversion to hexadecimal. However, you do need to reinterpret the 16-bit unicode character as 2 8-bit characters:

def encodeGSM7Message( text ):
    text = unicode(text, 'UTF-8').encode('UTF-16be')
    return  base64.b64encode(text)

print encodeGSM7Message('€'), 'IKw='