base64.c

   1 /*
   2  * Copyright (c) 1996-1999 by Internet Software Consortium.
   3  *
   4  * Permission to use, copy, modify, and distribute this software for any
   5  * purpose with or without fee is hereby granted, provided that the above
   6  * copyright notice and this permission notice appear in all copies.
   7  *
   8  * THE SOFTWARE IS PROVIDED "AS IS" AND INTERNET SOFTWARE CONSORTIUM DISCLAIMS
   9  * ALL WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES
  10  * OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL INTERNET SOFTWARE
  11  * CONSORTIUM BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
  12  * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
  13  * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
  14  * ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
  15  * SOFTWARE.
  16  */
  17
  18 /*
  19  * Portions Copyright (c) 1995 by International Business Machines, Inc.
  20  *
  21  * International Business Machines, Inc. (hereinafter called IBM) grants
  22  * permission under its copyrights to use, copy, modify, and distribute this
  23  * Software with or without fee, provided that the above copyright notice and
  24  * all paragraphs of this notice appear in all copies, and that the name of IBM
  25  * not be used in connection with the marketing of any product incorporating
  26  * the Software or modifications thereof, without specific, written prior
  27  * permission.
  28  *
  29  * To the extent it has a right to do so, IBM grants an immunity from suit
  30  * under its patents, if any, for the use, sale or manufacture of products to
  31  * the extent that such products are used for performing Domain Name System
  32  * dynamic updates in TCP/IP networks by means of the Software.  No immunity is
  33  * granted for any product per se or for any other function of any product.
  34  *
  35  * THE SOFTWARE IS PROVIDED "AS IS", AND IBM DISCLAIMS ALL WARRANTIES,
  36  * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
  37  * PARTICULAR PURPOSE.  IN NO EVENT SHALL IBM BE LIABLE FOR ANY SPECIAL,
  38  * DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER ARISING
  39  * OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE, EVEN
  40  * IF IBM IS APPRISED OF THE POSSIBILITY OF SUCH DAMAGES.
  41  */
  42
  43 #ifndef __APPLE__
  44 #if !defined(LINT) && !defined(CODECENTER)
  45 static const char rcsid[] = "$Id: base64.c,v 1.1 2006/03/01 19:01:34 majka Exp $";
  46 #endif /* not lint */
  47 #endif
  48
  49 #ifndef __APPLE__
  50 #include "port_before.h"
  51 #endif
  52
  53 #include <sys/types.h>
  54 #include <sys/param.h>
  55 #include <sys/socket.h>
  56
  57 #include <netinet/in.h>
  58 #include <arpa/inet.h>
  59 #include <arpa/nameser.h>
  60
  61 #include <ctype.h>
  62 #include <resolv.h>
  63 #include <stdio.h>
  64 #include <stdlib.h>
  65 #include <string.h>
  66
  67 #ifndef __APPLE__
  68 #include "port_after.h"
  69 #endif
  70
  71 #define Assert(Cond) if (!(Cond)) abort()
  72
  73 static const char Base64[] =
  74         "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
  75 static const char Pad64 = '=';
  76
  77 /* (From RFC1521 and draft-ietf-dnssec-secext-03.txt)
  78    The following encoding technique is taken from RFC 1521 by Borenstein
  79    and Freed.  It is reproduced here in a slightly edited form for
  80    convenience.
  81
  82    A 65-character subset of US-ASCII is used, enabling 6 bits to be
  83    represented per printable character. (The extra 65th character, "=",
  84    is used to signify a special processing function.)
  85
  86    The encoding process represents 24-bit groups of input bits as output
  87    strings of 4 encoded characters. Proceeding from left to right, a
  88    24-bit input group is formed by concatenating 3 8-bit input groups.
  89    These 24 bits are then treated as 4 concatenated 6-bit groups, each
  90    of which is translated into a single digit in the base64 alphabet.
  91
  92    Each 6-bit group is used as an index into an array of 64 printable
  93    characters. The character referenced by the index is placed in the
  94    output string.
  95
  96                          Table 1: The Base64 Alphabet
  97
  98       Value Encoding  Value Encoding  Value Encoding  Value Encoding
  99           0 A            17 R            34 i            51 z
 100           1 B            18 S            35 j            52 0
 101           2 C            19 T            36 k            53 1
 102           3 D            20 U            37 l            54 2
 103           4 E            21 V            38 m            55 3
 104           5 F            22 W            39 n            56 4
 105           6 G            23 X            40 o            57 5
 106           7 H            24 Y            41 p            58 6
 107           8 I            25 Z            42 q            59 7
 108           9 J            26 a            43 r            60 8
 109          10 K            27 b            44 s            61 9
 110          11 L            28 c            45 t            62 +
 111          12 M            29 d            46 u            63 /
 112          13 N            30 e            47 v
 113          14 O            31 f            48 w         (pad) =
 114          15 P            32 g            49 x
 115          16 Q            33 h            50 y
 116
 117    Special processing is performed if fewer than 24 bits are available
 118    at the end of the data being encoded.  A full encoding quantum is
 119    always completed at the end of a quantity.  When fewer than 24 input
 120    bits are available in an input group, zero bits are added (on the
 121    right) to form an integral number of 6-bit groups.  Padding at the
 122    end of the data is performed using the '=' character.
 123
 124    Since all base64 input is an integral number of octets, only the
 125          -------------------------------------------------
 126    following cases can arise:
 127
 128        (1) the final quantum of encoding input is an integral
 129            multiple of 24 bits; here, the final unit of encoded
 130            output will be an integral multiple of 4 characters
 131            with no "=" padding,
 132        (2) the final quantum of encoding input is exactly 8 bits;
 133            here, the final unit of encoded output will be two
 134            characters followed by two "=" padding characters, or
 135        (3) the final quantum of encoding input is exactly 16 bits;
 136            here, the final unit of encoded output will be three
 137            characters followed by one "=" padding character.
 138    */
 139
 140 int
 141 b64_ntop(u_char const *src, size_t srclength, char *target, size_t targsize) {
 142         size_t datalength = 0;
 143         u_char input[3];
 144         u_char output[4];
 145         size_t i;
 146
 147         while (2 < srclength) {
 148                 input[0] = *src++;
 149                 input[1] = *src++;
 150                 input[2] = *src++;
 151                 srclength -= 3;
 152
 153                 output[0] = input[0] >> 2;
 154                 output[1] = ((input[0] & 0x03) << 4) + (input[1] >> 4);
 155                 output[2] = ((input[1] & 0x0f) << 2) + (input[2] >> 6);
 156                 output[3] = input[2] & 0x3f;
 157                 Assert(output[0] < 64);
 158                 Assert(output[1] < 64);
 159                 Assert(output[2] < 64);
 160                 Assert(output[3] < 64);
 161
 162                 if (datalength + 4 > targsize)
 163                         return (-1);
 164                 target[datalength++] = Base64[output[0]];
 165                 target[datalength++] = Base64[output[1]];
 166                 target[datalength++] = Base64[output[2]];
 167                 target[datalength++] = Base64[output[3]];
 168         }
 169
 170         /* Now we worry about padding. */
 171         if (0 != srclength) {
 172                 /* Get what's left. */
 173                 input[0] = input[1] = input[2] = '\0';
 174                 for (i = 0; i < srclength; i++)
 175                         input[i] = *src++;
 176
 177                 output[0] = input[0] >> 2;
 178                 output[1] = ((input[0] & 0x03) << 4) + (input[1] >> 4);
 179                 output[2] = ((input[1] & 0x0f) << 2) + (input[2] >> 6);
 180                 Assert(output[0] < 64);
 181                 Assert(output[1] < 64);
 182                 Assert(output[2] < 64);
 183
 184                 if (datalength + 4 > targsize)
 185                         return (-1);
 186                 target[datalength++] = Base64[output[0]];
 187                 target[datalength++] = Base64[output[1]];
 188                 if (srclength == 1)
 189                         target[datalength++] = Pad64;
 190                 else
 191                         target[datalength++] = Base64[output[2]];
 192                 target[datalength++] = Pad64;
 193         }
 194         if (datalength >= targsize)
 195                 return (-1);
 196         target[datalength] = '\0';      /* Returned value doesn't count \0. */
 197         return (datalength);
 198 }
 199
 200 /* skips all whitespace anywhere.
 201    converts characters, four at a time, starting at (or after)
 202    src from base - 64 numbers into three 8 bit bytes in the target area.
 203    it returns the number of data bytes stored at the target, or -1 on error.
 204  */
 205
 206 int
 207 b64_pton(src, target, targsize)
 208         char const *src;
 209         u_char *target;
 210         size_t targsize;
 211 {
 212         int tarindex, state, ch;
 213         char *pos;
 214
 215         state = 0;
 216         tarindex = 0;
 217
 218         while ((ch = *src++) != '\0') {
 219                 if (isspace(ch))        /* Skip whitespace anywhere. */
 220                         continue;
 221
 222                 if (ch == Pad64)
 223                         break;
 224
 225                 pos = strchr(Base64, ch);
 226                 if (pos == 0)           /* A non-base64 character. */
 227                         return (-1);
 228
 229                 switch (state) {
 230                 case 0:
 231                         if (target) {
 232                                 if ((size_t)tarindex >= targsize)
 233                                         return (-1);
 234                                 target[tarindex] = (pos - Base64) << 2;
 235                         }
 236                         state = 1;
 237                         break;
 238                 case 1:
 239                         if (target) {
 240                                 if ((size_t)tarindex + 1 >= targsize)
 241                                         return (-1);
 242                                 target[tarindex]   |=  (pos - Base64) >> 4;
 243                                 target[tarindex+1]  = ((pos - Base64) & 0x0f)
 244                                                         << 4 ;
 245                         }
 246                         tarindex++;
 247                         state = 2;
 248                         break;
 249                 case 2:
 250                         if (target) {
 251                                 if ((size_t)tarindex + 1 >= targsize)
 252                                         return (-1);
 253                                 target[tarindex]   |=  (pos - Base64) >> 2;
 254                                 target[tarindex+1]  = ((pos - Base64) & 0x03)
 255                                                         << 6;
 256                         }
 257                         tarindex++;
 258                         state = 3;
 259                         break;
 260                 case 3:
 261                         if (target) {
 262                                 if ((size_t)tarindex >= targsize)
 263                                         return (-1);
 264                                 target[tarindex] |= (pos - Base64);
 265                         }
 266                         tarindex++;
 267                         state = 0;
 268                         break;
 269                 default:
 270                         abort();
 271                 }
 272         }
 273
 274         /*
 275          * We are done decoding Base-64 chars.  Let's see if we ended
 276          * on a byte boundary, and/or with erroneous trailing characters.
 277          */
 278
 279         if (ch == Pad64) {              /* We got a pad char. */
 280                 ch = *src++;            /* Skip it, get next. */
 281                 switch (state) {
 282                 case 0:         /* Invalid = in first position */
 283                 case 1:         /* Invalid = in second position */
 284                         return (-1);
 285
 286                 case 2:         /* Valid, means one byte of info */
 287                         /* Skip any number of spaces. */
 288                         for ((void)NULL; ch != '\0'; ch = *src++)
 289                                 if (!isspace(ch))
 290                                         break;
 291                         /* Make sure there is another trailing = sign. */
 292                         if (ch != Pad64)
 293                                 return (-1);
 294                         ch = *src++;            /* Skip the = */
 295                         /* Fall through to "single trailing =" case. */
 296                         /* FALLTHROUGH */
 297
 298                 case 3:         /* Valid, means two bytes of info */
 299                         /*
 300                          * We know this char is an =.  Is there anything but
 301                          * whitespace after it?
 302                          */
 303                         for ((void)NULL; ch != '\0'; ch = *src++)
 304                                 if (!isspace(ch))
 305                                         return (-1);
 306
 307                         /*
 308                          * Now make sure for cases 2 and 3 that the "extra"
 309                          * bits that slopped past the last full byte were
 310                          * zeros.  If we don't check them, they become a
 311                          * subliminal channel.
 312                          */
 313                         if (target && target[tarindex] != 0)
 314                                 return (-1);
 315                 }
 316         } else {
 317                 /*
 318                  * We ended by seeing the end of the string.  Make sure we
 319                  * have no partial bytes lying around.
 320                  */
 321                 if (state != 0)
 322                         return (-1);
 323         }
 324
 325         return (tarindex);
 326 }