Lib/base64.py

   1 #! /usr/bin/env python
   2
   3 """RFC 3548: Base16, Base32, Base64 Data Encodings"""
   4
   5 # Modified 04-Oct-1995 by Jack Jansen to use binascii module
   6 # Modified 30-Dec-2003 by Barry Warsaw to add full RFC 3548 support
   7
   8 import re
   9 import struct
  10 import binascii
  11
  12
  13 __all__ = [
  14     # Legacy interface exports traditional RFC 1521 Base64 encodings
  15     'encode', 'decode', 'encodestring', 'decodestring',
  16     # Generalized interface for other encodings
  17     'b64encode', 'b64decode', 'b32encode', 'b32decode',
  18     'b16encode', 'b16decode',
  19     # Standard Base64 encoding
  20     'standard_b64encode', 'standard_b64decode',
  21     # Some common Base64 alternatives.  As referenced by RFC 3458, see thread
  22     # starting at:
  23     #
  24     # http://zgp.org/pipermail/p2p-hackers/2001-September/000316.html
  25     'urlsafe_b64encode', 'urlsafe_b64decode',
  26     ]
  27
  28 _translation = [chr(_x) for _x in range(256)]
  29 EMPTYSTRING = ''
  30
  31
  32 def _translate(s, altchars):
  33     translation = _translation[:]
  34     for k, v in altchars.items():
  35         translation[ord(k)] = v
  36     return s.translate(''.join(translation))
  37
  38
  39 \f
  40 # Base64 encoding/decoding uses binascii
  41
  42 def b64encode(s, altchars=None):
  43     """Encode a string using Base64.
  44
  45     s is the string to encode.  Optional altchars must be a string of at least
  46     length 2 (additional characters are ignored) which specifies an
  47     alternative alphabet for the '+' and '/' characters.  This allows an
  48     application to e.g. generate url or filesystem safe Base64 strings.
  49
  50     The encoded string is returned.
  51     """
  52     # Strip off the trailing newline
  53     encoded = binascii.b2a_base64(s)[:-1]
  54     if altchars is not None:
  55         return _translate(encoded, {'+': altchars[0], '/': altchars[1]})
  56     return encoded
  57
  58
  59 def b64decode(s, altchars=None):
  60     """Decode a Base64 encoded string.
  61
  62     s is the string to decode.  Optional altchars must be a string of at least
  63     length 2 (additional characters are ignored) which specifies the
  64     alternative alphabet used instead of the '+' and '/' characters.
  65
  66     The decoded string is returned.  A TypeError is raised if s were
  67     incorrectly padded or if there are non-alphabet characters present in the
  68     string.
  69     """
  70     if altchars is not None:
  71         s = _translate(s, {altchars[0]: '+', altchars[1]: '/'})
  72     try:
  73         return binascii.a2b_base64(s)
  74     except binascii.Error, msg:
  75         # Transform this exception for consistency
  76         raise TypeError(msg)
  77
  78
  79 def standard_b64encode(s):
  80     """Encode a string using the standard Base64 alphabet.
  81
  82     s is the string to encode.  The encoded string is returned.
  83     """
  84     return b64encode(s)
  85
  86 def standard_b64decode(s):
  87     """Decode a string encoded with the standard Base64 alphabet.
  88
  89     s is the string to decode.  The decoded string is returned.  A TypeError
  90     is raised if the string is incorrectly padded or if there are non-alphabet
  91     characters present in the string.
  92     """
  93     return b64decode(s)
  94
  95 def urlsafe_b64encode(s):
  96     """Encode a string using a url-safe Base64 alphabet.
  97
  98     s is the string to encode.  The encoded string is returned.  The alphabet
  99     uses '-' instead of '+' and '_' instead of '/'.
 100     """
 101     return b64encode(s, '-_')
 102
 103 def urlsafe_b64decode(s):
 104     """Decode a string encoded with the standard Base64 alphabet.
 105
 106     s is the string to decode.  The decoded string is returned.  A TypeError
 107     is raised if the string is incorrectly padded or if there are non-alphabet
 108     characters present in the string.
 109
 110     The alphabet uses '-' instead of '+' and '_' instead of '/'.
 111     """
 112     return b64decode(s, '-_')
 113
 114
 115 \f
 116 # Base32 encoding/decoding must be done in Python
 117 _b32alphabet = {
 118     0: 'A',  9: 'J', 18: 'S', 27: '3',
 119     1: 'B', 10: 'K', 19: 'T', 28: '4',
 120     2: 'C', 11: 'L', 20: 'U', 29: '5',
 121     3: 'D', 12: 'M', 21: 'V', 30: '6',
 122     4: 'E', 13: 'N', 22: 'W', 31: '7',
 123     5: 'F', 14: 'O', 23: 'X',
 124     6: 'G', 15: 'P', 24: 'Y',
 125     7: 'H', 16: 'Q', 25: 'Z',
 126     8: 'I', 17: 'R', 26: '2',
 127     }
 128
 129 _b32tab = [v for v in _b32alphabet.values()]
 130 _b32rev = dict([(v, long(k)) for k, v in _b32alphabet.items()])
 131
 132
 133 def b32encode(s):
 134     """Encode a string using Base32.
 135
 136     s is the string to encode.  The encoded string is returned.
 137     """
 138     parts = []
 139     quanta, leftover = divmod(len(s), 5)
 140     # Pad the last quantum with zero bits if necessary
 141     if leftover:
 142         s += ('\0' * (5 - leftover))
 143         quanta += 1
 144     for i in range(quanta):
 145         # c1 and c2 are 16 bits wide, c3 is 8 bits wide.  The intent of this
 146         # code is to process the 40 bits in units of 5 bits.  So we take the 1
 147         # leftover bit of c1 and tack it onto c2.  Then we take the 2 leftover
 148         # bits of c2 and tack them onto c3.  The shifts and masks are intended
 149         # to give us values of exactly 5 bits in width.
 150         c1, c2, c3 = struct.unpack('!HHB', s[i*5:(i+1)*5])
 151         c2 += (c1 & 1) << 16 # 17 bits wide
 152         c3 += (c2 & 3) << 8  # 10 bits wide
 153         parts.extend([_b32tab[c1 >> 11],         # bits 1 - 5
 154                       _b32tab[(c1 >> 6) & 0x1f], # bits 6 - 10
 155                       _b32tab[(c1 >> 1) & 0x1f], # bits 11 - 15
 156                       _b32tab[c2 >> 12],         # bits 16 - 20 (1 - 5)
 157                       _b32tab[(c2 >> 7) & 0x1f], # bits 21 - 25 (6 - 10)
 158                       _b32tab[(c2 >> 2) & 0x1f], # bits 26 - 30 (11 - 15)
 159                       _b32tab[c3 >> 5],          # bits 31 - 35 (1 - 5)
 160                       _b32tab[c3 & 0x1f],        # bits 36 - 40 (1 - 5)
 161                       ])
 162     encoded = EMPTYSTRING.join(parts)
 163     # Adjust for any leftover partial quanta
 164     if leftover == 1:
 165         return encoded[:-6] + '======'
 166     elif leftover == 2:
 167         return encoded[:-4] + '===='
 168     elif leftover == 3:
 169         return encoded[:-3] + '==='
 170     elif leftover == 4:
 171         return encoded[:-1] + '='
 172     return encoded
 173
 174
 175 def b32decode(s, casefold=False, map01=None):
 176     """Decode a Base32 encoded string.
 177
 178     s is the string to decode.  Optional casefold is a flag specifying whether
 179     a lowercase alphabet is acceptable as input.  For security purposes, the
 180     default is False.
 181
 182     RFC 3548 allows for optional mapping of the digit 0 (zero) to the letter O
 183     (oh), and for optional mapping of the digit 1 (one) to either the letter I
 184     (eye) or letter L (el).  The optional argument map01 when not None,
 185     specifies which letter the digit 1 should be mapped to (when map01 is not
 186     None, the digit 0 is always mapped to the letter O).  For security
 187     purposes the default is None, so that 0 and 1 are not allowed in the
 188     input.
 189
 190     The decoded string is returned.  A TypeError is raised if s were
 191     incorrectly padded or if there are non-alphabet characters present in the
 192     string.
 193     """
 194     quanta, leftover = divmod(len(s), 8)
 195     if leftover:
 196         raise TypeError('Incorrect padding')
 197     # Handle section 2.4 zero and one mapping.  The flag map01 will be either
 198     # False, or the character to map the digit 1 (one) to.  It should be
 199     # either L (el) or I (eye).
 200     if map01:
 201         s = _translate(s, {'0': 'O', '1': map01})
 202     if casefold:
 203         s = s.upper()
 204     # Strip off pad characters from the right.  We need to count the pad
 205     # characters because this will tell us how many null bytes to remove from
 206     # the end of the decoded string.
 207     padchars = 0
 208     mo = re.search('(?P<pad>[=]*)$', s)
 209     if mo:
 210         padchars = len(mo.group('pad'))
 211         if padchars > 0:
 212             s = s[:-padchars]
 213     # Now decode the full quanta
 214     parts = []
 215     acc = 0
 216     shift = 35
 217     for c in s:
 218         val = _b32rev.get(c)
 219         if val is None:
 220             raise TypeError('Non-base32 digit found')
 221         acc += _b32rev[c] << shift
 222         shift -= 5
 223         if shift < 0:
 224             parts.append(binascii.unhexlify('%010x' % acc))
 225             acc = 0
 226             shift = 35
 227     # Process the last, partial quanta
 228     last = binascii.unhexlify('%010x' % acc)
 229     if padchars == 0:
 230         last = ''                       # No characters
 231     elif padchars == 1:
 232         last = last[:-1]
 233     elif padchars == 3:
 234         last = last[:-2]
 235     elif padchars == 4:
 236         last = last[:-3]
 237     elif padchars == 6:
 238         last = last[:-4]
 239     else:
 240         raise TypeError('Incorrect padding')
 241     parts.append(last)
 242     return EMPTYSTRING.join(parts)
 243
 244
 245 \f
 246 # RFC 3548, Base 16 Alphabet specifies uppercase, but hexlify() returns
 247 # lowercase.  The RFC also recommends against accepting input case
 248 # insensitively.
 249 def b16encode(s):
 250     """Encode a string using Base16.
 251
 252     s is the string to encode.  The encoded string is returned.
 253     """
 254     return binascii.hexlify(s).upper()
 255
 256
 257 def b16decode(s, casefold=False):
 258     """Decode a Base16 encoded string.
 259
 260     s is the string to decode.  Optional casefold is a flag specifying whether
 261     a lowercase alphabet is acceptable as input.  For security purposes, the
 262     default is False.
 263
 264     The decoded string is returned.  A TypeError is raised if s were
 265     incorrectly padded or if there are non-alphabet characters present in the
 266     string.
 267     """
 268     if casefold:
 269         s = s.upper()
 270     if re.search('[^0-9A-F]', s):
 271         raise TypeError('Non-base16 digit found')
 272     return binascii.unhexlify(s)
 273
 274
 275 \f
 276 # Legacy interface.  This code could be cleaned up since I don't believe
 277 # binascii has any line length limitations.  It just doesn't seem worth it
 278 # though.
 279
 280 MAXLINESIZE = 76 # Excluding the CRLF
 281 MAXBINSIZE = (MAXLINESIZE//4)*3
 282
 283 def encode(input, output):
 284     """Encode a file."""
 285     while True:
 286         s = input.read(MAXBINSIZE)
 287         if not s:
 288             break
 289         while len(s) < MAXBINSIZE:
 290             ns = input.read(MAXBINSIZE-len(s))
 291             if not ns:
 292                 break
 293             s += ns
 294         line = binascii.b2a_base64(s)
 295         output.write(line)
 296
 297
 298 def decode(input, output):
 299     """Decode a file."""
 300     while True:
 301         line = input.readline()
 302         if not line:
 303             break
 304         s = binascii.a2b_base64(line)
 305         output.write(s)
 306
 307
 308 def encodestring(s):
 309     """Encode a string."""
 310     pieces = []
 311     for i in range(0, len(s), MAXBINSIZE):
 312         chunk = s[i : i + MAXBINSIZE]
 313         pieces.append(binascii.b2a_base64(chunk))
 314     return "".join(pieces)
 315
 316
 317 def decodestring(s):
 318     """Decode a string."""
 319     return binascii.a2b_base64(s)
 320
 321
 322 \f
 323 # Useable as a script...
 324 def test():
 325     """Small test program"""
 326     import sys, getopt
 327     try:
 328         opts, args = getopt.getopt(sys.argv[1:], 'deut')
 329     except getopt.error, msg:
 330         sys.stdout = sys.stderr
 331         print msg
 332         print """usage: %s [-d|-e|-u|-t] [file|-]
 333         -d, -u: decode
 334         -e: encode (default)
 335         -t: encode and decode string 'Aladdin:open sesame'"""%sys.argv[0]
 336         sys.exit(2)
 337     func = encode
 338     for o, a in opts:
 339         if o == '-e': func = encode
 340         if o == '-d': func = decode
 341         if o == '-u': func = decode
 342         if o == '-t': test1(); return
 343     if args and args[0] != '-':
 344         func(open(args[0], 'rb'), sys.stdout)
 345     else:
 346         func(sys.stdin, sys.stdout)
 347
 348
 349 def test1():
 350     s0 = "Aladdin:open sesame"
 351     s1 = encodestring(s0)
 352     s2 = decodestring(s1)
 353     print s0, repr(s1), s2
 354
 355
 356 if __name__ == '__main__':
 357     test()