b64.c

   1 /* Based on:
   2  *
   3  * from http://base64.sourceforge.net/
   4  * ********************************************************************\
   5
   6 MODULE NAME:    b64.c
   7
   8 AUTHOR:         Bob Trower 08/04/01
   9
  10 PROJECT:        Crypt Data Packaging
  11
  12 COPYRIGHT:      Copyright (c) Trantor Standard Systems Inc., 2001
  13
  14 NOTE:           This source code may be used as you wish, subject to
  15                 the MIT license.  See the LICENCE section below.
  16
  17 DESCRIPTION:
  18                 This little utility implements the Base64
  19                 Content-Transfer-Encoding standard described in
  20                 RFC1113 (http://www.faqs.org/rfcs/rfc1113.html).
  21
  22                 This is the coding scheme used by MIME to allow
  23                 binary data to be transferred by SMTP mail.
  24
  25                 Groups of 3 bytes from a binary stream are coded as
  26                 groups of 4 bytes in a text stream.
  27
  28                 The input stream is 'padded' with zeros to create
  29                 an input that is an even multiple of 3.
  30
  31                 A special character ('=') is used to denote padding so
  32                 that the stream can be decoded back to its exact size.
  33
  34                 Encoded output is formatted in lines which should
  35                 be a maximum of 72 characters to conform to the
  36                 specification.  This program defaults to 72 characters,
  37                 but will allow more or less through the use of a
  38                 switch.  The program enforces a minimum line size
  39                 of 4 characters.
  40
  41                 Example encoding:
  42
  43                 The stream 'ABCD' is 32 bits long.  It is mapped as
  44                 follows:
  45
  46                 ABCD
  47
  48                  A (65)     B (66)     C (67)     D (68)   (None) (None)
  49                 01000001   01000010   01000011   01000100
  50
  51                 16 (Q)  20 (U)  9 (J)   3 (D)    17 (R) 0 (A)  NA (=) NA (=)
  52                 010000  010100  001001  000011   010001 000000 000000 000000
  53
  54
  55                 QUJDRA==
  56
  57                 Decoding is the process in reverse.  A 'decode' lookup
  58                 table has been created to avoid string scans.
  59
  60 DESIGN GOALS:   Specifically:
  61                 Code is a stand-alone utility to perform base64
  62                 encoding/decoding. It should be genuinely useful
  63                 when the need arises and it meets a need that is
  64                 likely to occur for some users.
  65                 Code acts as sample code to show the author's
  66                 design and coding style.
  67
  68                 Generally:
  69                 This program is designed to survive:
  70                 Everything you need is in a single source file.
  71                 It compiles cleanly using a vanilla ANSI C compiler.
  72                 It does its job correctly with a minimum of fuss.
  73                 The code is not overly clever, not overly simplistic
  74                 and not overly verbose.
  75                 Access is 'cut and paste' from a web page.
  76                 Terms of use are reasonable.
  77
  78 VALIDATION:     Non-trivial code is never without errors.  This
  79                 file likely has some problems, since it has only
  80                 been tested by the author.  It is expected with most
  81                 source code that there is a period of 'burn-in' when
  82                 problems are identified and corrected.  That being
  83                 said, it is possible to have 'reasonably correct'
  84                 code by following a regime of unit test that covers
  85                 the most likely cases and regression testing prior
  86                 to release.  This has been done with this code and
  87                 it has a good probability of performing as expected.
  88
  89                 Unit Test Cases:
  90
  91                 case 0:empty file:
  92                     CASE0.DAT  ->  ->
  93                     (Zero length target file created
  94                     on both encode and decode.)
  95
  96                 case 1:One input character:
  97                     CASE1.DAT A -> QQ== -> A
  98
  99                 case 2:Two input characters:
 100                     CASE2.DAT AB -> QUJD -> AB
 101
 102                 case 3:Three input characters:
 103                     CASE3.DAT ABC -> QUJD -> ABC
 104
 105                 case 4:Four input characters:
 106                     case4.dat ABCD -> QUJDRA== -> ABCD
 107
 108                 case 5:All chars from 0 to ff, LINE_SIZE set to 50:
 109
 110                     AAECAwQFBgcICQoLDA0ODxAREhMUFRYXGBkaGxwdHh8gISIj
 111                     JCUmJygpKissLS4vMDEyMzQ1Njc4OTo7PD0+P0BBQkNERUZH
 112                     SElKS0xNTk9QUVJTVFVWV1hZWltcXV5fYGFiY2RlZmdoaWpr
 113                     bG1ub3BxcnN0dXZ3eHl6e3x9fn+AgYKDhIWGh4iJiouMjY6P
 114                     kJGSk5SVlpeYmZqbnJ2en6ChoqOkpaanqKmqq6ytrq+wsbKz
 115                     tLW2t7i5uru8vb6/wMHCw8TFxsfIycrLzM3Oz9DR0tPU1dbX
 116                     2Nna29zd3t/g4eLj5OXm5+jp6uvs7e7v8PHy8/T19vf4+fr7
 117                     /P3+/w==
 118
 119                 case 6:Mime Block from e-mail:
 120                     (Data same as test case 5)
 121
 122                 case 7: Large files:
 123                     Tested 28 MB file in/out.
 124
 125                 case 8: Random Binary Integrity:
 126                     This binary program (b64.exe) was encoded to base64,
 127                     back to binary and then executed.
 128
 129                 case 9 Stress:
 130                     All files in a working directory encoded/decoded
 131                     and compared with file comparison utility to
 132                     ensure that multiple runs do not cause problems
 133                     such as exhausting file handles, tmp storage, etc.
 134
 135                 -------------
 136
 137                 Syntax, operation and failure:
 138                     All options/switches tested.  Performs as
 139                     expected.
 140
 141                 case 10:
 142                     No Args -- Shows Usage Screen
 143                     Return Code 1 (Invalid Syntax)
 144                 case 11:
 145                     One Arg (invalid) -- Shows Usage Screen
 146                     Return Code 1 (Invalid Syntax)
 147                 case 12:
 148                     One Arg Help (-?) -- Shows detailed Usage Screen.
 149                     Return Code 0 (Success -- help request is valid).
 150                 case 13:
 151                     One Arg Help (-h) -- Shows detailed Usage Screen.
 152                     Return Code 0 (Success -- help request is valid).
 153                 case 14:
 154                     One Arg (valid) -- Uses stdin/stdout (filter)
 155                     Return Code 0 (Sucess)
 156                 case 15:
 157                     Two Args (invalid file) -- shows system error.
 158                     Return Code 2 (File Error)
 159                 case 16:
 160                     Encode non-existent file -- shows system error.
 161                     Return Code 2 (File Error)
 162                 case 17:
 163                     Out of disk space -- shows system error.
 164                     Return Code 3 (File I/O Error)
 165
 166                 -------------
 167
 168                 Compile/Regression test:
 169                     gcc compiled binary under Cygwin
 170                     Microsoft Visual Studio under Windows 2000
 171                     Microsoft Version 6.0 C under Windows 2000
 172
 173 DEPENDENCIES:   None
 174
 175 LICENCE:        Copyright (c) 2001 Bob Trower, Trantor Standard Systems Inc.
 176
 177                 Permission is hereby granted, free of charge, to any person
 178                 obtaining a copy of this software and associated
 179                 documentation files (the "Software"), to deal in the
 180                 Software without restriction, including without limitation
 181                 the rights to use, copy, modify, merge, publish, distribute,
 182                 sublicense, and/or sell copies of the Software, and to
 183                 permit persons to whom the Software is furnished to do so,
 184                 subject to the following conditions:
 185
 186                 The above copyright notice and this permission notice shall
 187                 be included in all copies or substantial portions of the
 188                 Software.
 189
 190                 THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY
 191                 KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE
 192                 WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR
 193                 PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS
 194                 OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
 195                 OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
 196                 OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
 197                 SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 198
 199 VERSION HISTORY:
 200                 Bob Trower 08/04/01 -- Create Version 0.00.00B
 201
 202 \******************************************************************* */
 203
 204 #include <stdio.h>
 205 #include <stdlib.h>
 206 #include <string.h>
 207 #include <sysexits.h>
 208
 209 #define LINE_SIZE   72
 210
 211 /*
 212 ** Translation Table as described in RFC1113
 213 */
 214 static const char cb64[]="ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
 215
 216 /*
 217 ** Translation Table to decode (created by author)
 218 */
 219 static const char cd64[]="|$$$}rstuvwxyz{$$$$$$$>?@ABCDEFGHIJKLMNOPQRSTUVW$$$$$$XYZ[\\]^_`abcdefghijklmnopq";
 220
 221 /*
 222 ** encodeblock
 223 **
 224 ** encode 3 8-bit binary bytes as 4 '6-bit' characters
 225 */
 226 void encodeblock( unsigned char in[3], unsigned char out[4], int len )
 227 {
 228     out[0] = cb64[ in[0] >> 2 ];
 229     out[1] = cb64[ ((in[0] & 0x03) << 4) | ((in[1] & 0xf0) >> 4) ];
 230     out[2] = (unsigned char) (len > 1 ? cb64[ ((in[1] & 0x0f) << 2) | ((in[2] & 0xc0) >> 6) ] : '=');
 231     out[3] = (unsigned char) (len > 2 ? cb64[ in[2] & 0x3f ] : '=');
 232 }
 233
 234 /*
 235 ** decodeblock
 236 **
 237 ** decode 4 '6-bit' characters into 3 8-bit binary bytes
 238 */
 239 void decodeblock( unsigned char in[4], unsigned char out[3] )
 240 {
 241     out[ 0 ] = (unsigned char ) (in[0] << 2 | in[1] >> 4);
 242     out[ 1 ] = (unsigned char ) (in[1] << 4 | in[2] >> 2);
 243     out[ 2 ] = (unsigned char ) (((in[2] << 6) & 0xc0) | in[3]);
 244 }
 245
 246 /* Decode base64. Caller frees. */
 247 char *b64_decode(char *in)
 248 {
 249     char *out;
 250     unsigned int at, out_at;
 251
 252
 253     /* Get some space for the output. Needs less, but keep it simple. */
 254     out = malloc(strlen(in));
 255     if (!out) {
 256         perror("malloc");
 257         exit(EX_UNAVAILABLE);
 258     }
 259
 260     memset(out, 0, strlen(in));
 261
 262     at = out_at = 0;
 263     do {
 264         unsigned int i;
 265         unsigned char in4[4];
 266
 267         /* Read 4 base64-encoded characters */
 268         i = 0;
 269         do
 270         {
 271             char c;
 272
 273             c = in[at++];
 274             if (isalpha(c) || isdigit(c))
 275             {
 276                 printf("in4 = %c\n", c);
 277                 in4[i++] = c;
 278             }
 279         } while (i < 4);
 280         decodeblock(in4, (unsigned char *)(out + out_at));
 281         out_at += 3;
 282         printf("\n");
 283     } while(at < strlen(in));
 284
 285     return out;
 286 }
 287
 288 int main()
 289 {
 290     char *s;
 291     int i;
 292
 293     /* TODO: fix this! */
 294     printf("|%s|\n", b64_decode("aXQgd29ya3M="));
 295 }
 296