problems with base64

Karl Pech KarlPech at users.sf.net
Sun Jul 11 23:54:40 CEST 2004


"Byron" <DesertLinux at netscape.net> wrote in
news:29dIc.13450$oD3.9178 at newsread1.news.pas.earthlink.net...

Hi Byron,

> I don't know if this is much help for you, but have you tried using the 
> following:
> 
> import base64
> print base64.decodestring("UGxlYXNlLCBoZWxwIG1lIQ==")
> print base64.encodestring("Please, help me!")

Thank you very much for your answer! I'm glad that somebody answered me,
because my problem slowly but surely is getting worse. :((
I found the mistake in the previous code which was a problem with read()
I think it should be read(1). Anyway I deleted this old version of my
program since it was very slow for base64 - files > 0.5 MB. Now I coded the
program below. This program doesn't work at all. It seems to me that
some variables that read_in_data should use are somehow out of scope.
The idea of the program is: "Read in 5-Blocks of base64-data. Since
5 % 4 =|= 0 we get an IndexError-Exception in the "while len(source) < 4"-loop,
we catch this exception, set the source file pointer 1 byte back, because the
would lose one byte otherwise, and read in the next 4 bytes of the source file.
Unfortunately I don't know how to solve this Unbound...Error-Thing. |((
I think I'm really stuck now. @{

And as I said, since this task is an exercise for this weekend I'm not allowed to
use the standard base64-module. Well, I have to code the base64-decoder myself. :(((

Anyway, Thanks, Byron.

---
import string

def convert_to_8bits(source):
  base64_table = {'A' :  0, 'N' : 13, 'a' : 26, 'n' : 39, '0' : 52,
                  'B' :  1, 'O' : 14, 'b' : 27, 'o' : 40, '1' : 53,
                  'C' :  2, 'P' : 15, 'c' : 28, 'p' : 41, '2' : 54,
                  'D' :  3, 'Q' : 16, 'd' : 29, 'q' : 42, '3' : 55,
                  'E' :  4, 'R' : 17, 'e' : 30, 'r' : 43, '4' : 56,
                  'F' :  5, 'S' : 18, 'f' : 31, 's' : 44, '5' : 57,
                  'G' :  6, 'T' : 19, 'g' : 32, 't' : 45, '6' : 58,
                  'H' :  7, 'U' : 20, 'h' : 33, 'u' : 46, '7' : 59,
                  'I' :  8, 'V' : 21, 'i' : 34, 'v' : 47, '8' : 60,
                  'J' :  9, 'W' : 22, 'j' : 35, 'w' : 48, '9' : 61,
                  'K' : 10, 'X' : 23, 'k' : 36, 'x' : 49, '+' : 62,
                  'L' : 11, 'Y' : 24, 'l' : 37, 'y' : 50, '/' : 63,
                  'M' : 12, 'Z' : 25, 'm' : 38, 'z' : 51, '=' : 0}

  result_ = []

  box_ =  int( (base64_table[source[0]] << 26)\
             + (base64_table[source[1]] << 20)\
             + (base64_table[source[2]] << 14)\
             + (base64_table[source[3]] <<  8) )

  result_ += chr((box_ >> 24) & 255) + chr((box_ >> 16) & 255) + chr((box_ >> 8) & 255)

  return result_

fsource = open(raw_input("Bitte geben Sie die zu dekodierende base64-Datei an: "), "rb")

_1stline = fsource.readline().split()

del _1stline[0:2]

targetname = string.join(_1stline)
ftarget = open(targetname, "wb")

global source_file_data
global target_file_buffer
global filling_counter

source_file_data = fsource.read(5)
target_file_buffer = []
filling_counter = 0

def read_in_data():
  for i in range(0, len(source_file_data)):
    print i
    source = []
    while len(source) < 4:
      source += source_file_data[i]

      if source_file_data[i] == '=':
        filling_counter += 1

      if not(ord('a') <= ord(source_file_data[i]) <= ord('z')) and\
         not(ord('A') <= ord(source_file_data[i]) <= ord('Z')) and\
         not(ord('0') <= ord(source_file_data[i]) <= ord('9'))\
         and not(ord(source_file_data[i]) in [ord('+'), ord('/'), ord('=')]):
        del source[len(source)-1]

      i += 1

    clear_text = convert_to_8bits(source)

    target_file_buffer += clear_text

try:
  read_in_data()
except IndexError:
  print "exception"
  fsource.seek(-1, 1)
  source_file_data = fsource.read(5)
  read_in_data()

ftarget.writelines(target_file_buffer)
ftarget.seek(filling_counter-1, 2)
ftarget.truncate()

ftarget.close()
fsource.close()

print "decoded file "+targetname+" has been written!"
---




More information about the Python-list mailing list