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 = _b32alphabet.items()
 130 _b32tab.sort()
 131 _b32tab = [v for k, v in _b32tab]
 132 _b32rev = dict([(v, long(k)) for k, v in _b32alphabet.items()])
 133
 134
 135 def b32encode(s):
 136     """Encode a string using Base32.
 137
 138     s is the string to encode.  The encoded string is returned.
 139     """
 140     parts = []
 141     quanta, leftover = divmod(len(s), 5)
 142     # Pad the last quantum with zero bits if necessary
 143     if leftover:
 144         s += ('\0' * (5 - leftover))
 145         quanta += 1
 146     for i in range(quanta):
 147         # c1 and c2 are 16 bits wide, c3 is 8 bits wide.  The intent of this
 148         # code is to process the 40 bits in units of 5 bits.  So we take the 1
 149         # leftover bit of c1 and tack it onto c2.  Then we take the 2 leftover
 150         # bits of c2 and tack them onto c3.  The shifts and masks are intended
 151         # to give us values of exactly 5 bits in width.
 152         c1, c2, c3 = struct.unpack('!HHB', s[i*5:(i+1)*5])
 153         c2 += (c1 & 1) << 16 # 17 bits wide
 154         c3 += (c2 & 3) << 8  # 10 bits wide
 155         parts.extend([_b32tab[c1 >> 11],         # bits 1 - 5
 156                       _b32tab[(c1 >> 6) & 0x1f], # bits 6 - 10
 157                       _b32tab[(c1 >> 1) & 0x1f], # bits 11 - 15
 158                       _b32tab[c2 >> 12],         # bits 16 - 20 (1 - 5)
 159                       _b32tab[(c2 >> 7) & 0x1f], # bits 21 - 25 (6 - 10)
 160                       _b32tab[(c2 >> 2) & 0x1f], # bits 26 - 30 (11 - 15)
 161                       _b32tab[c3 >> 5],          # bits 31 - 35 (1 - 5)
 162                       _b32tab[c3 & 0x1f],        # bits 36 - 40 (1 - 5)
 163                       ])
 164     encoded = EMPTYSTRING.join(parts)
 165     # Adjust for any leftover partial quanta
 166     if leftover == 1:
 167         return encoded[:-6] + '======'
 168     elif leftover == 2:
 169         return encoded[:-4] + '===='
 170     elif leftover == 3:
 171         return encoded[:-3] + '==='
 172     elif leftover == 4:
 173         return encoded[:-1] + '='
 174     return encoded
 175
 176
 177 def b32decode(s, casefold=False, map01=None):
 178     """Decode a Base32 encoded string.
 179
 180     s is the string to decode.  Optional casefold is a flag specifying whether
 181     a lowercase alphabet is acceptable as input.  For security purposes, the
 182     default is False.
 183
 184     RFC 3548 allows for optional mapping of the digit 0 (zero) to the letter O
 185     (oh), and for optional mapping of the digit 1 (one) to either the letter I
 186     (eye) or letter L (el).  The optional argument map01 when not None,
 187     specifies which letter the digit 1 should be mapped to (when map01 is not
 188     None, the digit 0 is always mapped to the letter O).  For security
 189     purposes the default is None, so that 0 and 1 are not allowed in the
 190     input.
 191
 192     The decoded string is returned.  A TypeError is raised if s were
 193     incorrectly padded or if there are non-alphabet characters present in the
 194     string.
 195     """
 196     quanta, leftover = divmod(len(s), 8)
 197     if leftover:
 198         raise TypeError('Incorrect padding')
 199     # Handle section 2.4 zero and one mapping.  The flag map01 will be either
 200     # False, or the character to map the digit 1 (one) to.  It should be
 201     # either L (el) or I (eye).
 202     if map01:
 203         s = _translate(s, {'0': 'O', '1': map01})
 204     if casefold:
 205         s = s.upper()
 206     # Strip off pad characters from the right.  We need to count the pad
 207     # characters because this will tell us how many null bytes to remove from
 208     # the end of the decoded string.
 209     padchars = 0
 210     mo = re.search('(?P<pad>[=]*)$', s)
 211     if mo:
 212         padchars = len(mo.group('pad'))
 213         if padchars > 0:
 214             s = s[:-padchars]
 215     # Now decode the full quanta
 216     parts = []
 217     acc = 0
 218     shift = 35
 219     for c in s:
 220         val = _b32rev.get(c)
 221         if val is None:
 222             raise TypeError('Non-base32 digit found')
 223         acc += _b32rev[c] << shift
 224         shift -= 5
 225         if shift < 0:
 226             parts.append(binascii.unhexlify('%010x' % acc))
 227             acc = 0
 228             shift = 35
 229     # Process the last, partial quanta
 230     last = binascii.unhexlify('%010x' % acc)
 231     if padchars == 0:
 232         last = ''                       # No characters
 233     elif padchars == 1:
 234         last = last[:-1]
 235     elif padchars == 3:
 236         last = last[:-2]
 237     elif padchars == 4:
 238         last = last[:-3]
 239     elif padchars == 6:
 240         last = last[:-4]
 241     else:
 242         raise TypeError('Incorrect padding')
 243     parts.append(last)
 244     return EMPTYSTRING.join(parts)
 245
 246
 247 \f
 248 # RFC 3548, Base 16 Alphabet specifies uppercase, but hexlify() returns
 249 # lowercase.  The RFC also recommends against accepting input case
 250 # insensitively.
 251 def b16encode(s):
 252     """Encode a string using Base16.
 253
 254     s is the string to encode.  The encoded string is returned.
 255     """
 256     return binascii.hexlify(s).upper()
 257
 258
 259 def b16decode(s, casefold=False):
 260     """Decode a Base16 encoded string.
 261
 262     s is the string to decode.  Optional casefold is a flag specifying whether
 263     a lowercase alphabet is acceptable as input.  For security purposes, the
 264     default is False.
 265
 266     The decoded string is returned.  A TypeError is raised if s were
 267     incorrectly padded or if there are non-alphabet characters present in the
 268     string.
 269     """
 270     if casefold:
 271         s = s.upper()
 272     if re.search('[^0-9A-F]', s):
 273         raise TypeError('Non-base16 digit found')
 274     return binascii.unhexlify(s)
 275
 276
 277 \f
 278 # Legacy interface.  This code could be cleaned up since I don't believe
 279 # binascii has any line length limitations.  It just doesn't seem worth it
 280 # though.
 281
 282 MAXLINESIZE = 76 # Excluding the CRLF
 283 MAXBINSIZE = (MAXLINESIZE//4)*3
 284
 285 def encode(input, output):
 286     """Encode a file."""
 287     while True:
 288         s = input.read(MAXBINSIZE)
 289         if not s:
 290             break
 291         while len(s) < MAXBINSIZE:
 292             ns = input.read(MAXBINSIZE-len(s))
 293             if not ns:
 294                 break
 295             s += ns
 296         line = binascii.b2a_base64(s)
 297         output.write(line)
 298
 299
 300 def decode(input, output):
 301     """Decode a file."""
 302     while True:
 303         line = input.readline()
 304         if not line:
 305             break
 306         s = binascii.a2b_base64(line)
 307         output.write(s)
 308
 309
 310 def encodestring(s):
 311     """Encode a string into multiple lines of base-64 data."""
 312     pieces = []
 313     for i in range(0, len(s), MAXBINSIZE):
 314         chunk = s[i : i + MAXBINSIZE]
 315         pieces.append(binascii.b2a_base64(chunk))
 316     return "".join(pieces)
 317
 318
 319 def decodestring(s):
 320     """Decode a string."""
 321     return binascii.a2b_base64(s)
 322
 323
 324 \f
 325 # Useable as a script...
 326 def test():
 327     """Small test program"""
 328     import sys, getopt
 329     try:
 330         opts, args = getopt.getopt(sys.argv[1:], 'deut')
 331     except getopt.error, msg:
 332         sys.stdout = sys.stderr
 333         print msg
 334         print """usage: %s [-d|-e|-u|-t] [file|-]
 335         -d, -u: decode
 336         -e: encode (default)
 337         -t: encode and decode string 'Aladdin:open sesame'"""%sys.argv[0]
 338         sys.exit(2)
 339     func = encode
 340     for o, a in opts:
 341         if o == '-e': func = encode
 342         if o == '-d': func = decode
 343         if o == '-u': func = decode
 344         if o == '-t': test1(); return
 345     if args and args[0] != '-':
 346         func(open(args[0], 'rb'), sys.stdout)
 347     else:
 348         func(sys.stdin, sys.stdout)
 349
 350
 351 def test1():
 352     s0 = "Aladdin:open sesame"
 353     s1 = encodestring(s0)
 354     s2 = decodestring(s1)
 355     print s0, repr(s1), s2
 356
 357
 358 if __name__ == '__main__':
 359     test()