ceba1986 |
1 | /* -*-c-*- |
2 | * |
d5679ed0 |
3 | * $Id: base64.c,v 1.6 2003/05/16 00:22:58 mdw Exp $ |
ceba1986 |
4 | * |
5 | * Base64 encoding and decoding. |
6 | * |
7 | * (c) 1997 Straylight/Edgeware |
8 | */ |
9 | |
10 | /*----- Licensing notice --------------------------------------------------* |
11 | * |
12 | * This file is part of the mLib utilities library. |
13 | * |
14 | * mLib is free software; you can redistribute it and/or modify |
15 | * it under the terms of the GNU Library General Public License as |
16 | * published by the Free Software Foundation; either version 2 of the |
17 | * License, or (at your option) any later version. |
18 | * |
19 | * mLib is distributed in the hope that it will be useful, |
20 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
21 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
22 | * GNU Library General Public License for more details. |
23 | * |
24 | * You should have received a copy of the GNU Library General Public |
25 | * License along with mLib; if not, write to the Free |
26 | * Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, |
27 | * MA 02111-1307, USA. |
28 | */ |
29 | |
30 | /*----- Revision history --------------------------------------------------* |
31 | * |
32 | * $Log: base64.c,v $ |
d5679ed0 |
33 | * Revision 1.6 2003/05/16 00:22:58 mdw |
34 | * Test base64 and hex encoding. |
35 | * |
fde30071 |
36 | * Revision 1.5 2002/01/13 13:26:16 mdw |
37 | * Change names for internal tables. |
38 | * |
a226592f |
39 | * Revision 1.4 1999/10/15 21:08:46 mdw |
40 | * Change support for erroneous Base64 streams with length 1 mod 4. |
41 | * |
0bfb1431 |
42 | * Revision 1.3 1999/05/21 22:14:30 mdw |
43 | * Take advantage of the new dynamic string macros. |
44 | * |
c601c4f1 |
45 | * Revision 1.2 1999/05/18 21:45:27 mdw |
46 | * Allow Base64 encode and decode of arbitrary rubbish. |
47 | * |
ceba1986 |
48 | * Revision 1.1 1999/05/17 20:35:00 mdw |
49 | * Base64 encoding and decoding support. |
50 | * |
51 | */ |
52 | |
53 | /*----- Header files ------------------------------------------------------*/ |
54 | |
55 | #include <stdio.h> |
56 | #include <stdlib.h> |
57 | #include <string.h> |
58 | |
59 | #include "base64.h" |
60 | #include "dstr.h" |
61 | |
62 | /*----- Important tables --------------------------------------------------*/ |
63 | |
fde30071 |
64 | static const char encodemap[] = { "ABCDEFGHIJKLMNOPQRSTUVWXYZ" |
65 | "abcdefghijklmnopqrstuvwxyz" |
66 | "0123456789+/" }; |
ceba1986 |
67 | |
fde30071 |
68 | static const signed char decodemap[] = { |
ceba1986 |
69 | -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, /* 0x */ |
70 | -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, /* 1x */ |
71 | -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 62, -1, -1, -1, 63, /* 2x */ |
72 | 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, -1, -1, -1, -1, -1, -1, /* 3x */ |
73 | -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, /* 4x */ |
74 | 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, -1, -1, -1, -1, -1, /* 5x */ |
75 | -1, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36 ,37, 38, 39, 40, /* 6x */ |
76 | 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, -1, -1, -1, -1, -1 /* 7x */ |
77 | }; |
78 | |
79 | /*----- Main code ---------------------------------------------------------*/ |
80 | |
81 | /* --- @base64_encode@ --- * |
82 | * |
83 | * Arguments: @base64_ctx *ctx@ = pointer to a context block |
c601c4f1 |
84 | * @const void *p@ = pointer to a source buffer |
ceba1986 |
85 | * @size_t sz@ = size of the source buffer |
86 | * @dstr *d@ = pointer to destination string |
87 | * |
88 | * Returns: --- |
89 | * |
90 | * Use: Encodes a binary string in base64. To flush out the final |
91 | * few characters (if necessary), pass a null source pointer. |
92 | */ |
93 | |
94 | void base64_encode(base64_ctx *ctx, |
c601c4f1 |
95 | const void *p, size_t sz, |
ceba1986 |
96 | dstr *d) |
97 | { |
c601c4f1 |
98 | if (p) { |
ceba1986 |
99 | unsigned long acc = ctx->acc; |
100 | unsigned qsz = ctx->qsz; |
c601c4f1 |
101 | const unsigned char *src = p; |
ceba1986 |
102 | |
103 | while (sz) { |
104 | acc = (acc << 8) | *src++; |
105 | qsz++; |
106 | sz--; |
107 | if (qsz == 3) { |
fde30071 |
108 | DPUTC(d, encodemap[(acc >> 18) & 0x3f]); |
109 | DPUTC(d, encodemap[(acc >> 12) & 0x3f]); |
110 | DPUTC(d, encodemap[(acc >> 6) & 0x3f]); |
111 | DPUTC(d, encodemap[(acc >> 0) & 0x3f]); |
ceba1986 |
112 | ctx->lnlen += 4; |
113 | if (ctx->maxline && ctx->lnlen >= ctx->maxline) { |
114 | dstr_puts(d, ctx->indent); |
115 | ctx->lnlen = 0; |
116 | } |
117 | qsz = 0; |
118 | acc = 0; |
119 | } |
120 | } |
121 | |
122 | ctx->acc = acc; |
123 | ctx->qsz = qsz; |
124 | } else { |
125 | unsigned long acc = ctx->acc; |
126 | unsigned qsz = ctx->qsz; |
127 | |
128 | switch (qsz) { |
129 | case 0: |
130 | break; |
131 | case 1: |
132 | acc <<= 16; |
fde30071 |
133 | DPUTC(d, encodemap[(acc >> 18) & 0x3f]); |
134 | DPUTC(d, encodemap[(acc >> 12) & 0x3f]); |
c601c4f1 |
135 | DPUTC(d, '='); |
136 | DPUTC(d, '='); |
ceba1986 |
137 | ctx->lnlen += 4; |
138 | break; |
139 | case 2: |
140 | acc <<= 8; |
fde30071 |
141 | DPUTC(d, encodemap[(acc >> 18) & 0x3f]); |
142 | DPUTC(d, encodemap[(acc >> 12) & 0x3f]); |
143 | DPUTC(d, encodemap[(acc >> 6) & 0x3f]); |
c601c4f1 |
144 | DPUTC(d, '='); |
ceba1986 |
145 | ctx->lnlen += 4; |
146 | break; |
147 | } |
148 | ctx->qsz = 0; |
149 | ctx->acc = 0; |
150 | } |
151 | } |
152 | |
153 | /* --- @base64_decode@ --- * |
154 | * |
155 | * Arguments: @base64_ctx *ctx@ = pointer to a context block |
c601c4f1 |
156 | * @const void *p@ = pointer to a source buffer |
ceba1986 |
157 | * @size_t sz@ = size of the source buffer |
158 | * @dstr *d@ = pointer to destination string |
159 | * |
160 | * Returns: --- |
161 | * |
162 | * Use: Decodes a binary string in base64. To flush out the final |
163 | * few characters (if necessary), pass a null source pointer. |
164 | */ |
165 | |
166 | void base64_decode(base64_ctx *ctx, |
c601c4f1 |
167 | const void *p, size_t sz, |
ceba1986 |
168 | dstr *d) |
169 | { |
c601c4f1 |
170 | if (p) { |
ceba1986 |
171 | unsigned long acc = ctx->acc; |
172 | unsigned qsz = ctx->qsz; |
c601c4f1 |
173 | const char *src = p; |
ceba1986 |
174 | int ch; |
175 | |
176 | while (sz) { |
177 | |
178 | /* --- Get the next character and convert it --- */ |
179 | |
180 | ch = *src++; |
181 | if (ch >= 128 || ch < 0) |
182 | ch = -1; |
183 | else |
fde30071 |
184 | ch = decodemap[ch]; |
ceba1986 |
185 | sz--; |
186 | if (ch == -1) |
187 | continue; |
188 | |
189 | /* --- Bung it in the accumulator --- */ |
190 | |
191 | acc = (acc << 6) | ch; |
192 | qsz++; |
193 | |
194 | /* --- Maybe write out a completed triplet --- */ |
195 | |
196 | if (qsz == 4) { |
c601c4f1 |
197 | DPUTC(d, (acc >> 16) & 0xff); |
198 | DPUTC(d, (acc >> 8) & 0xff); |
199 | DPUTC(d, (acc >> 0) & 0xff); |
ceba1986 |
200 | acc = 0; |
201 | qsz = 0; |
202 | } |
203 | } |
204 | |
205 | ctx->acc = acc; |
206 | ctx->qsz = qsz; |
207 | } else { |
208 | |
209 | /* --- Notes about the tail-end bits --- * |
210 | * |
211 | * Ending Base64 decoding is messy. The reference I'm using to define |
212 | * the encoding, RFC1521 section 5.2, is a little hazy on exactly what to |
213 | * do at the end. It explains that I'm meant to ignore spurious `=' |
214 | * characters, and points out that I'm not guaranteed to see anything |
215 | * interesting at the end. I'll play safe here, and ignore all `=' |
216 | * characters, relying on my client to work out when to stop feeding me |
217 | * data. I'll use the queue size to work out how many tail-end bytes |
218 | * I ought to write. |
219 | */ |
220 | |
221 | unsigned long acc = ctx->acc; |
222 | unsigned qsz = ctx->qsz; |
223 | |
a226592f |
224 | /* --- Now fiddle with everything else --- * |
225 | * |
226 | * There's a bodge here for invalid encodings which have only one hextet |
227 | * in the final group. I'm not sure this is really worth having, but it |
228 | * might save some unexpected behaviour. (Not that you won't still get |
229 | * unexpected behaviour if the stream is completely empty, of course.) |
230 | */ |
ceba1986 |
231 | |
a226592f |
232 | if (qsz) { |
233 | acc <<= 6 * (4 - qsz); |
234 | qsz *= 6; |
235 | if (qsz < 8) |
236 | qsz = 8; |
237 | while (qsz >= 8) { |
238 | DPUTC(d, (acc >> 16) & 0xff); |
239 | acc <<= 8; |
240 | qsz -= 8; |
241 | } |
ceba1986 |
242 | } |
243 | |
244 | /* --- That seems to be good enough --- */ |
245 | |
246 | ctx->qsz = 0; |
247 | ctx->acc = 0; |
248 | } |
249 | } |
250 | |
251 | /* --- @base64_init@ --- * |
252 | * |
253 | * Arguments: @base64_ctx *ctx@ = pointer to context block to initialize |
254 | * |
255 | * Returns: --- |
256 | * |
257 | * Use: Initializes a base64 context properly. |
258 | */ |
259 | |
260 | void base64_init(base64_ctx *ctx) |
261 | { |
262 | ctx->acc = 0; |
263 | ctx->qsz = 0; |
264 | ctx->lnlen = 0; |
265 | ctx->indent = "\n"; |
266 | ctx->maxline = 72; |
267 | } |
268 | |
269 | /*----- Test driver code --------------------------------------------------*/ |
270 | |
271 | #ifdef TEST_RIG |
272 | |
273 | int main(int argc, char *argv[]) |
274 | { |
275 | unsigned char buf[BUFSIZ]; |
0bfb1431 |
276 | dstr d = DSTR_INIT; |
ceba1986 |
277 | base64_ctx ctx; |
d5679ed0 |
278 | void (*proc)(base64_ctx *, const void *, size_t, dstr *); |
ceba1986 |
279 | size_t sz; |
280 | |
ceba1986 |
281 | base64_init(&ctx); |
282 | |
283 | if (argc > 1 && strcmp(argv[1], "-d") == 0) |
284 | proc = base64_decode; |
285 | else { |
286 | proc = base64_encode; |
287 | putchar('\t'); |
288 | ctx.indent = "\n\t"; |
289 | ctx.maxline = 64; |
290 | } |
291 | |
292 | do { |
293 | sz = fread(buf, 1, sizeof(buf), stdin); |
294 | if (sz) { |
295 | proc(&ctx, buf, sz, &d); |
296 | dstr_write(&d, stdout); |
297 | dstr_destroy(&d); |
298 | } |
299 | } while (sz == sizeof(buf)); |
300 | |
301 | proc(&ctx, 0, 0, &d); |
302 | dstr_write(&d, stdout); |
303 | |
304 | if (proc == base64_encode) |
305 | putchar('\n'); |
306 | |
307 | return (0); |
308 | } |
309 | |
310 | #endif |
311 | |
312 | /*----- That's all, folks -------------------------------------------------*/ |