chiark / gitweb /
Infrastructure: Split the files into subdirectories.
[mLib] / codec / base64.c
CommitLineData
ceba1986 1/* -*-c-*-
2 *
ceba1986 3 * Base64 encoding and decoding.
4 *
5 * (c) 1997 Straylight/Edgeware
6 */
7
d4efbcd9 8/*----- Licensing notice --------------------------------------------------*
ceba1986 9 *
10 * This file is part of the mLib utilities library.
11 *
12 * mLib is free software; you can redistribute it and/or modify
13 * it under the terms of the GNU Library General Public License as
14 * published by the Free Software Foundation; either version 2 of the
15 * License, or (at your option) any later version.
d4efbcd9 16 *
ceba1986 17 * mLib is distributed in the hope that it will be useful,
18 * but WITHOUT ANY WARRANTY; without even the implied warranty of
19 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
20 * GNU Library General Public License for more details.
d4efbcd9 21 *
ceba1986 22 * You should have received a copy of the GNU Library General Public
23 * License along with mLib; if not, write to the Free
24 * Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
25 * MA 02111-1307, USA.
26 */
27
ceba1986 28/*----- Header files ------------------------------------------------------*/
29
30#include <stdio.h>
31#include <stdlib.h>
32#include <string.h>
33
34#include "base64.h"
35#include "dstr.h"
36
37/*----- Important tables --------------------------------------------------*/
38
fde30071 39static const char encodemap[] = { "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
d4efbcd9 40 "abcdefghijklmnopqrstuvwxyz"
fde30071 41 "0123456789+/" };
ceba1986 42
fde30071 43static const signed char decodemap[] = {
ceba1986 44 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, /* 0x */
45 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, /* 1x */
46 -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 62, -1, -1, -1, 63, /* 2x */
47 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, -1, -1, -1, -1, -1, -1, /* 3x */
48 -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, /* 4x */
49 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, -1, -1, -1, -1, -1, /* 5x */
50 -1, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36 ,37, 38, 39, 40, /* 6x */
51 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, -1, -1, -1, -1, -1 /* 7x */
d4efbcd9 52};
ceba1986 53
54/*----- Main code ---------------------------------------------------------*/
55
56/* --- @base64_encode@ --- *
57 *
58 * Arguments: @base64_ctx *ctx@ = pointer to a context block
c601c4f1 59 * @const void *p@ = pointer to a source buffer
ceba1986 60 * @size_t sz@ = size of the source buffer
61 * @dstr *d@ = pointer to destination string
62 *
63 * Returns: ---
64 *
65 * Use: Encodes a binary string in base64. To flush out the final
66 * few characters (if necessary), pass a null source pointer.
67 */
68
69void base64_encode(base64_ctx *ctx,
c601c4f1 70 const void *p, size_t sz,
ceba1986 71 dstr *d)
72{
c601c4f1 73 if (p) {
ceba1986 74 unsigned long acc = ctx->acc;
75 unsigned qsz = ctx->qsz;
c601c4f1 76 const unsigned char *src = p;
ceba1986 77
78 while (sz) {
79 acc = (acc << 8) | *src++;
80 qsz++;
81 sz--;
82 if (qsz == 3) {
fde30071 83 DPUTC(d, encodemap[(acc >> 18) & 0x3f]);
84 DPUTC(d, encodemap[(acc >> 12) & 0x3f]);
85 DPUTC(d, encodemap[(acc >> 6) & 0x3f]);
86 DPUTC(d, encodemap[(acc >> 0) & 0x3f]);
ceba1986 87 ctx->lnlen += 4;
88 if (ctx->maxline && ctx->lnlen >= ctx->maxline) {
89 dstr_puts(d, ctx->indent);
90 ctx->lnlen = 0;
91 }
92 qsz = 0;
93 acc = 0;
94 }
95 }
96
97 ctx->acc = acc;
98 ctx->qsz = qsz;
99 } else {
100 unsigned long acc = ctx->acc;
101 unsigned qsz = ctx->qsz;
102
103 switch (qsz) {
104 case 0:
105 break;
106 case 1:
107 acc <<= 16;
fde30071 108 DPUTC(d, encodemap[(acc >> 18) & 0x3f]);
109 DPUTC(d, encodemap[(acc >> 12) & 0x3f]);
c601c4f1 110 DPUTC(d, '=');
111 DPUTC(d, '=');
ceba1986 112 ctx->lnlen += 4;
113 break;
114 case 2:
115 acc <<= 8;
fde30071 116 DPUTC(d, encodemap[(acc >> 18) & 0x3f]);
117 DPUTC(d, encodemap[(acc >> 12) & 0x3f]);
118 DPUTC(d, encodemap[(acc >> 6) & 0x3f]);
c601c4f1 119 DPUTC(d, '=');
ceba1986 120 ctx->lnlen += 4;
121 break;
122 }
123 ctx->qsz = 0;
124 ctx->acc = 0;
125 }
126}
127
128/* --- @base64_decode@ --- *
129 *
130 * Arguments: @base64_ctx *ctx@ = pointer to a context block
c601c4f1 131 * @const void *p@ = pointer to a source buffer
ceba1986 132 * @size_t sz@ = size of the source buffer
133 * @dstr *d@ = pointer to destination string
134 *
135 * Returns: ---
136 *
137 * Use: Decodes a binary string in base64. To flush out the final
138 * few characters (if necessary), pass a null source pointer.
139 */
140
141void base64_decode(base64_ctx *ctx,
c601c4f1 142 const void *p, size_t sz,
ceba1986 143 dstr *d)
144{
c601c4f1 145 if (p) {
ceba1986 146 unsigned long acc = ctx->acc;
147 unsigned qsz = ctx->qsz;
c601c4f1 148 const char *src = p;
ceba1986 149 int ch;
150
151 while (sz) {
152
153 /* --- Get the next character and convert it --- */
154
155 ch = *src++;
156 if (ch >= 128 || ch < 0)
157 ch = -1;
158 else
fde30071 159 ch = decodemap[ch];
ceba1986 160 sz--;
161 if (ch == -1)
162 continue;
163
164 /* --- Bung it in the accumulator --- */
165
166 acc = (acc << 6) | ch;
167 qsz++;
168
169 /* --- Maybe write out a completed triplet --- */
170
171 if (qsz == 4) {
c601c4f1 172 DPUTC(d, (acc >> 16) & 0xff);
173 DPUTC(d, (acc >> 8) & 0xff);
174 DPUTC(d, (acc >> 0) & 0xff);
ceba1986 175 acc = 0;
176 qsz = 0;
177 }
178 }
179
180 ctx->acc = acc;
181 ctx->qsz = qsz;
182 } else {
183
184 /* --- Notes about the tail-end bits --- *
185 *
186 * Ending Base64 decoding is messy. The reference I'm using to define
187 * the encoding, RFC1521 section 5.2, is a little hazy on exactly what to
188 * do at the end. It explains that I'm meant to ignore spurious `='
189 * characters, and points out that I'm not guaranteed to see anything
190 * interesting at the end. I'll play safe here, and ignore all `='
191 * characters, relying on my client to work out when to stop feeding me
192 * data. I'll use the queue size to work out how many tail-end bytes
193 * I ought to write.
194 */
195
196 unsigned long acc = ctx->acc;
197 unsigned qsz = ctx->qsz;
198
a226592f 199 /* --- Now fiddle with everything else --- *
200 *
201 * There's a bodge here for invalid encodings which have only one hextet
202 * in the final group. I'm not sure this is really worth having, but it
203 * might save some unexpected behaviour. (Not that you won't still get
204 * unexpected behaviour if the stream is completely empty, of course.)
205 */
ceba1986 206
a226592f 207 if (qsz) {
208 acc <<= 6 * (4 - qsz);
209 qsz *= 6;
210 if (qsz < 8)
211 qsz = 8;
212 while (qsz >= 8) {
213 DPUTC(d, (acc >> 16) & 0xff);
214 acc <<= 8;
215 qsz -= 8;
216 }
ceba1986 217 }
218
219 /* --- That seems to be good enough --- */
220
221 ctx->qsz = 0;
222 ctx->acc = 0;
223 }
224}
225
226/* --- @base64_init@ --- *
227 *
228 * Arguments: @base64_ctx *ctx@ = pointer to context block to initialize
229 *
230 * Returns: ---
231 *
232 * Use: Initializes a base64 context properly.
233 */
234
235void base64_init(base64_ctx *ctx)
236{
237 ctx->acc = 0;
238 ctx->qsz = 0;
239 ctx->lnlen = 0;
240 ctx->indent = "\n";
241 ctx->maxline = 72;
242}
243
244/*----- Test driver code --------------------------------------------------*/
245
246#ifdef TEST_RIG
247
248int main(int argc, char *argv[])
249{
250 unsigned char buf[BUFSIZ];
0bfb1431 251 dstr d = DSTR_INIT;
ceba1986 252 base64_ctx ctx;
d5679ed0 253 void (*proc)(base64_ctx *, const void *, size_t, dstr *);
ceba1986 254 size_t sz;
255
ceba1986 256 base64_init(&ctx);
257
258 if (argc > 1 && strcmp(argv[1], "-d") == 0)
259 proc = base64_decode;
260 else {
261 proc = base64_encode;
262 putchar('\t');
263 ctx.indent = "\n\t";
264 ctx.maxline = 64;
265 }
266
267 do {
268 sz = fread(buf, 1, sizeof(buf), stdin);
269 if (sz) {
270 proc(&ctx, buf, sz, &d);
271 dstr_write(&d, stdout);
272 dstr_destroy(&d);
273 }
274 } while (sz == sizeof(buf));
275
276 proc(&ctx, 0, 0, &d);
277 dstr_write(&d, stdout);
278
279 if (proc == base64_encode)
280 putchar('\n');
281
282 return (0);
283}
284
285#endif
286
287/*----- That's all, folks -------------------------------------------------*/