base64.c

   1 /*      $OpenBSD: base64.c,v 1.7 2013/12/31 02:32:56 tedu Exp $ */
   2
   3 /*
   4  * Copyright (c) 1996 by Internet Software Consortium.
   5  *
   6  * Permission to use, copy, modify, and distribute this software for any
   7  * purpose with or without fee is hereby granted, provided that the above
   8  * copyright notice and this permission notice appear in all copies.
   9  *
  10  * THE SOFTWARE IS PROVIDED "AS IS" AND INTERNET SOFTWARE CONSORTIUM DISCLAIMS
  11  * ALL WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES
  12  * OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL INTERNET SOFTWARE
  13  * CONSORTIUM BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
  14  * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
  15  * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
  16  * ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
  17  * SOFTWARE.
  18  */
  19
  20 /*
  21  * Portions Copyright (c) 1995 by International Business Machines, Inc.
  22  *
  23  * International Business Machines, Inc. (hereinafter called IBM) grants
  24  * permission under its copyrights to use, copy, modify, and distribute this
  25  * Software with or without fee, provided that the above copyright notice and
  26  * all paragraphs of this notice appear in all copies, and that the name of IBM
  27  * not be used in connection with the marketing of any product incorporating
  28  * the Software or modifications thereof, without specific, written prior
  29  * permission.
  30  *
  31  * To the extent it has a right to do so, IBM grants an immunity from suit
  32  * under its patents, if any, for the use, sale or manufacture of products to
  33  * the extent that such products are used for performing Domain Name System
  34  * dynamic updates in TCP/IP networks by means of the Software.  No immunity is
  35  * granted for any product per se or for any other function of any product.
  36  *
  37  * THE SOFTWARE IS PROVIDED "AS IS", AND IBM DISCLAIMS ALL WARRANTIES,
  38  * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
  39  * PARTICULAR PURPOSE.  IN NO EVENT SHALL IBM BE LIABLE FOR ANY SPECIAL,
  40  * DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER ARISING
  41  * OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE, EVEN
  42  * IF IBM IS APPRISED OF THE POSSIBILITY OF SUCH DAMAGES.
  43  */
  44
  45 #include <sys/types.h>
  46 #include <ctype.h>
  47 #include <stdio.h>
  48 #include <stdlib.h>
  49 #include <string.h>
  50 #include "base64.h"
  51
  52 static const char Base64[] =
  53         "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
  54 static const char Pad64 = '=';
  55
  56 /* (From RFC1521 and draft-ietf-dnssec-secext-03.txt)
  57    The following encoding technique is taken from RFC 1521 by Borenstein
  58    and Freed.  It is reproduced here in a slightly edited form for
  59    convenience.
  60
  61    A 65-character subset of US-ASCII is used, enabling 6 bits to be
  62    represented per printable character. (The extra 65th character, "=",
  63    is used to signify a special processing function.)
  64
  65    The encoding process represents 24-bit groups of input bits as output
  66    strings of 4 encoded characters. Proceeding from left to right, a
  67    24-bit input group is formed by concatenating 3 8-bit input groups.
  68    These 24 bits are then treated as 4 concatenated 6-bit groups, each
  69    of which is translated into a single digit in the base64 alphabet.
  70
  71    Each 6-bit group is used as an index into an array of 64 printable
  72    characters. The character referenced by the index is placed in the
  73    output string.
  74
  75                          Table 1: The Base64 Alphabet
  76
  77       Value Encoding  Value Encoding  Value Encoding  Value Encoding
  78           0 A            17 R            34 i            51 z
  79           1 B            18 S            35 j            52 0
  80           2 C            19 T            36 k            53 1
  81           3 D            20 U            37 l            54 2
  82           4 E            21 V            38 m            55 3
  83           5 F            22 W            39 n            56 4
  84           6 G            23 X            40 o            57 5
  85           7 H            24 Y            41 p            58 6
  86           8 I            25 Z            42 q            59 7
  87           9 J            26 a            43 r            60 8
  88          10 K            27 b            44 s            61 9
  89          11 L            28 c            45 t            62 +
  90          12 M            29 d            46 u            63 /
  91          13 N            30 e            47 v
  92          14 O            31 f            48 w         (pad) =
  93          15 P            32 g            49 x
  94          16 Q            33 h            50 y
  95
  96    Special processing is performed if fewer than 24 bits are available
  97    at the end of the data being encoded.  A full encoding quantum is
  98    always completed at the end of a quantity.  When fewer than 24 input
  99    bits are available in an input group, zero bits are added (on the
 100    right) to form an integral number of 6-bit groups.  Padding at the
 101    end of the data is performed using the '=' character.
 102
 103    Since all base64 input is an integral number of octets, only the
 104          -------------------------------------------------
 105    following cases can arise:
 106
 107        (1) the final quantum of encoding input is an integral
 108            multiple of 24 bits; here, the final unit of encoded
 109            output will be an integral multiple of 4 characters
 110            with no "=" padding,
 111        (2) the final quantum of encoding input is exactly 8 bits;
 112            here, the final unit of encoded output will be two
 113            characters followed by two "=" padding characters, or
 114        (3) the final quantum of encoding input is exactly 16 bits;
 115            here, the final unit of encoded output will be three
 116            characters followed by one "=" padding character.
 117    */
 118
 119 int b64_ntop(const void *_src, size_t srclength,
 120              void *dest, size_t targsize)
 121 {
 122         const unsigned char *src = _src;
 123         char *target = dest;
 124         size_t datalength = 0;
 125         u_char input[3];
 126         u_char output[4];
 127         int i;
 128
 129         while (2 < srclength) {
 130                 input[0] = *src++;
 131                 input[1] = *src++;
 132                 input[2] = *src++;
 133                 srclength -= 3;
 134
 135                 output[0] = input[0] >> 2;
 136                 output[1] = ((input[0] & 0x03) << 4) + (input[1] >> 4);
 137                 output[2] = ((input[1] & 0x0f) << 2) + (input[2] >> 6);
 138                 output[3] = input[2] & 0x3f;
 139
 140                 if (datalength + 4 > targsize)
 141                         return (-1);
 142                 target[datalength++] = Base64[output[0]];
 143                 target[datalength++] = Base64[output[1]];
 144                 target[datalength++] = Base64[output[2]];
 145                 target[datalength++] = Base64[output[3]];
 146         }
 147
 148         /* Now we worry about padding. */
 149         if (0 != srclength) {
 150                 /* Get what's left. */
 151                 input[0] = input[1] = input[2] = '\0';
 152                 for (i = 0; i < srclength; i++)
 153                         input[i] = *src++;
 154
 155                 output[0] = input[0] >> 2;
 156                 output[1] = ((input[0] & 0x03) << 4) + (input[1] >> 4);
 157                 output[2] = ((input[1] & 0x0f) << 2) + (input[2] >> 6);
 158
 159                 if (datalength + 4 > targsize)
 160                         return (-1);
 161                 target[datalength++] = Base64[output[0]];
 162                 target[datalength++] = Base64[output[1]];
 163                 if (srclength == 1)
 164                         target[datalength++] = Pad64;
 165                 else
 166                         target[datalength++] = Base64[output[2]];
 167                 target[datalength++] = Pad64;
 168         }
 169         if (datalength >= targsize)
 170                 return (-1);
 171         target[datalength] = '\0';      /* Returned value doesn't count \0. */
 172         return (datalength);
 173 }
 174
 175 /* skips all whitespace anywhere.
 176    converts characters, four at a time, starting at (or after)
 177    src from base - 64 numbers into three 8 bit bytes in the target area.
 178    it returns the number of data bytes stored at the target, or -1 on error.
 179  */
 180
 181 int b64_pton(const void *_src, void *dest, size_t targsize)
 182 {
 183         const char *src = _src;
 184         unsigned char *target = dest;
 185         int tarindex, state, ch;
 186         u_char nextbyte;
 187         char *pos;
 188
 189         state = 0;
 190         tarindex = 0;
 191
 192         while ((ch = (unsigned char)*src++) != '\0') {
 193                 if (isspace(ch))        /* Skip whitespace anywhere. */
 194                         continue;
 195
 196                 if (ch == Pad64)
 197                         break;
 198
 199                 pos = strchr(Base64, ch);
 200                 if (pos == 0)           /* A non-base64 character. */
 201                         return (-1);
 202
 203                 switch (state) {
 204                 case 0:
 205                         if (target) {
 206                                 if (tarindex >= targsize)
 207                                         return (-1);
 208                                 target[tarindex] = (pos - Base64) << 2;
 209                         }
 210                         state = 1;
 211                         break;
 212                 case 1:
 213                         if (target) {
 214                                 if (tarindex >= targsize)
 215                                         return (-1);
 216                                 target[tarindex]   |=  (pos - Base64) >> 4;
 217                                 nextbyte = ((pos - Base64) & 0x0f) << 4;
 218                                 if (tarindex + 1 < targsize)
 219                                         target[tarindex+1] = nextbyte;
 220                                 else if (nextbyte)
 221                                         return (-1);
 222                         }
 223                         tarindex++;
 224                         state = 2;
 225                         break;
 226                 case 2:
 227                         if (target) {
 228                                 if (tarindex >= targsize)
 229                                         return (-1);
 230                                 target[tarindex]   |=  (pos - Base64) >> 2;
 231                                 nextbyte = ((pos - Base64) & 0x03) << 6;
 232                                 if (tarindex + 1 < targsize)
 233                                         target[tarindex+1] = nextbyte;
 234                                 else if (nextbyte)
 235                                         return (-1);
 236                         }
 237                         tarindex++;
 238                         state = 3;
 239                         break;
 240                 case 3:
 241                         if (target) {
 242                                 if (tarindex >= targsize)
 243                                         return (-1);
 244                                 target[tarindex] |= (pos - Base64);
 245                         }
 246                         tarindex++;
 247                         state = 0;
 248                         break;
 249                 }
 250         }
 251
 252         /*
 253          * We are done decoding Base-64 chars.  Let's see if we ended
 254          * on a byte boundary, and/or with erroneous trailing characters.
 255          */
 256
 257         if (ch == Pad64) {                      /* We got a pad char. */
 258                 ch = (unsigned char)*src++;     /* Skip it, get next. */
 259                 switch (state) {
 260                 case 0:         /* Invalid = in first position */
 261                 case 1:         /* Invalid = in second position */
 262                         return (-1);
 263
 264                 case 2:         /* Valid, means one byte of info */
 265                         /* Skip any number of spaces. */
 266                         for (; ch != '\0'; ch = (unsigned char)*src++)
 267                                 if (!isspace(ch))
 268                                         break;
 269                         /* Make sure there is another trailing = sign. */
 270                         if (ch != Pad64)
 271                                 return (-1);
 272                         ch = (unsigned char)*src++;             /* Skip the = */
 273                         /* Fall through to "single trailing =" case. */
 274                         /* FALLTHROUGH */
 275
 276                 case 3:         /* Valid, means two bytes of info */
 277                         /*
 278                          * We know this char is an =.  Is there anything but
 279                          * whitespace after it?
 280                          */
 281                         for (; ch != '\0'; ch = (unsigned char)*src++)
 282                                 if (!isspace(ch))
 283                                         return (-1);
 284
 285                         /*
 286                          * Now make sure for cases 2 and 3 that the "extra"
 287                          * bits that slopped past the last full byte were
 288                          * zeros.  If we don't check them, they become a
 289                          * subliminal channel.
 290                          */
 291                         if (target && tarindex < targsize &&
 292                             target[tarindex] != 0)
 293                                 return (-1);
 294                 }
 295         } else {
 296                 /*
 297                  * We ended by seeing the end of the string.  Make sure we
 298                  * have no partial bytes lying around.
 299                  */
 300                 if (state != 0)
 301                         return (-1);
 302         }
 303
 304         return (tarindex);
 305 }