/src/libwebsockets/lib/misc/base64-decode.c
Line | Count | Source |
1 | | /* |
2 | | * This code originally came from here |
3 | | * |
4 | | * http://base64.sourceforge.net/b64.c |
5 | | * |
6 | | * already with MIT license, which is retained. |
7 | | * |
8 | | * LICENCE: Copyright (c) 2001 Bob Trower, Trantor Standard Systems Inc. |
9 | | * |
10 | | * Permission is hereby granted, free of charge, to any person |
11 | | * obtaining a copy of this software and associated |
12 | | * documentation files (the "Software"), to deal in the |
13 | | * Software without restriction, including without limitation |
14 | | * the rights to use, copy, modify, merge, publish, distribute, |
15 | | * sublicense, and/or sell copies of the Software, and to |
16 | | * permit persons to whom the Software is furnished to do so, |
17 | | * subject to the following conditions: |
18 | | * |
19 | | * The above copyright notice and this permission notice shall |
20 | | * be included in all copies or substantial portions of the |
21 | | * Software. |
22 | | * |
23 | | * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY |
24 | | * KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE |
25 | | * WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR |
26 | | * PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS |
27 | | * OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR |
28 | | * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR |
29 | | * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE |
30 | | * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. |
31 | | * |
32 | | * VERSION HISTORY: |
33 | | * Bob Trower 08/04/01 -- Create Version 0.00.00B |
34 | | * |
35 | | * I cleaned it up quite a bit to match the (linux kernel) style of the rest |
36 | | * of libwebsockets |
37 | | */ |
38 | | |
39 | | #include "private-lib-core.h" |
40 | | |
41 | | #include <stdio.h> |
42 | | #include <string.h> |
43 | | |
44 | | static const char encode_orig[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ" |
45 | | "abcdefghijklmnopqrstuvwxyz0123456789+/"; |
46 | | static const char encode_url[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ" |
47 | | "abcdefghijklmnopqrstuvwxyz0123456789-_"; |
48 | | static const char decode[] = "|$$$}rstuvwxyz{$$$$$$$>?@ABCDEFGHIJKLMNOPQRSTUVW" |
49 | | "$$$$$$XYZ[\\]^_`abcdefghijklmnopq"; |
50 | | |
51 | | static int |
52 | | _lws_b64_encode_string(const char *encode, const char *in, int in_len, |
53 | | char *out, int out_size) |
54 | 0 | { |
55 | 0 | unsigned char triple[3]; |
56 | 0 | int i, done = 0; |
57 | |
|
58 | 0 | while (in_len) { |
59 | 0 | int len = 0; |
60 | 0 | for (i = 0; i < 3; i++) { |
61 | 0 | if (in_len) { |
62 | 0 | triple[i] = (unsigned char)*in++; |
63 | 0 | len++; |
64 | 0 | in_len--; |
65 | 0 | } else |
66 | 0 | triple[i] = 0; |
67 | 0 | } |
68 | |
|
69 | 0 | if (done + 4 >= out_size) |
70 | 0 | return -1; |
71 | | |
72 | 0 | *out++ = encode[triple[0] >> 2]; |
73 | 0 | *out++ = encode[(((triple[0] & 0x03) << 4) & 0x30) | |
74 | 0 | (((triple[1] & 0xf0) >> 4) & 0x0f)]; |
75 | 0 | *out++ = (char)(len > 1 ? encode[(((triple[1] & 0x0f) << 2) & 0x3c) | |
76 | 0 | (((triple[2] & 0xc0) >> 6) & 3)] : '='); |
77 | 0 | *out++ = (char)(len > 2 ? encode[triple[2] & 0x3f] : '='); |
78 | |
|
79 | 0 | done += 4; |
80 | 0 | } |
81 | | |
82 | 0 | if (done + 1 >= out_size) |
83 | 0 | return -1; |
84 | | |
85 | 0 | *out++ = '\0'; |
86 | |
|
87 | 0 | return done; |
88 | 0 | } |
89 | | |
90 | | int |
91 | | lws_b64_encode_string(const char *in, int in_len, char *out, int out_size) |
92 | 0 | { |
93 | 0 | return _lws_b64_encode_string(encode_orig, in, in_len, out, out_size); |
94 | 0 | } |
95 | | |
96 | | int |
97 | | lws_b64_encode_string_url(const char *in, int in_len, char *out, int out_size) |
98 | 0 | { |
99 | 0 | return _lws_b64_encode_string(encode_url, in, in_len, out, out_size); |
100 | 0 | } |
101 | | |
102 | | |
103 | | void |
104 | | lws_b64_decode_state_init(struct lws_b64state *state) |
105 | 0 | { |
106 | 0 | memset(state, 0, sizeof(*state)); |
107 | 0 | } |
108 | | |
109 | | int |
110 | | lws_b64_decode_stateful(struct lws_b64state *s, const char *in, size_t *in_len, |
111 | | uint8_t *out, size_t *out_size, int final) |
112 | 0 | { |
113 | 0 | const char *orig_in = in, *end_in = in + *in_len; |
114 | 0 | uint8_t *orig_out = out, *end_out = out + *out_size; |
115 | 0 | int equals = 0; |
116 | |
|
117 | 0 | while (in < end_in && *in && out + 3 <= end_out) { |
118 | |
|
119 | 0 | for (; s->i < 4 && in < end_in && *in; s->i++) { |
120 | 0 | uint8_t v; |
121 | |
|
122 | 0 | v = 0; |
123 | 0 | s->c = 0; |
124 | 0 | while (in < end_in && *in && !v) { |
125 | 0 | s->c = v = (unsigned char)*in++; |
126 | |
|
127 | 0 | if (v == '\x0a' || v == '\x0d') { |
128 | 0 | v = 0; |
129 | 0 | continue; |
130 | 0 | } |
131 | | |
132 | 0 | if (v == '=') { |
133 | 0 | equals++; |
134 | 0 | v = 0; |
135 | 0 | continue; |
136 | 0 | } |
137 | | |
138 | | /* Sanity check this is part of the charset */ |
139 | | |
140 | 0 | if ((v < '0' || v > '9') && |
141 | 0 | (v < 'A' || v > 'Z') && |
142 | 0 | (v < 'a' || v > 'z') && |
143 | 0 | v != '-' && v != '+' && v != '_' && v != '/') { |
144 | 0 | lwsl_err("%s: bad base64 0x%02X '%c' @+%d\n", __func__, v, v, lws_ptr_diff(in, orig_in)); |
145 | 0 | return -1; |
146 | 0 | } |
147 | | |
148 | 0 | if (equals) { |
149 | 0 | lwsl_err("%s: non = after =\n", __func__); |
150 | 0 | return -1; |
151 | 0 | } |
152 | | |
153 | | /* support the url base64 variant too */ |
154 | 0 | if (v == '-') |
155 | 0 | s->c = v = '+'; |
156 | 0 | if (v == '_') |
157 | 0 | s->c = v = '/'; |
158 | 0 | v = (uint8_t)decode[v - 43]; |
159 | 0 | if (v) |
160 | 0 | v = (uint8_t)((v == '$') ? 0 : v - 61); |
161 | 0 | } |
162 | 0 | if (s->c) { |
163 | 0 | s->len++; |
164 | 0 | if (v) |
165 | 0 | s->quad[s->i] = (uint8_t)(v - 1); |
166 | 0 | } else |
167 | 0 | s->quad[s->i] = 0; |
168 | 0 | } |
169 | | |
170 | 0 | if (s->i != 4 && !final) |
171 | 0 | continue; |
172 | | |
173 | 0 | s->i = 0; |
174 | | |
175 | | /* |
176 | | * Normally we convert a group of 4 incoming symbols into 3 bytes. |
177 | | * |
178 | | * "The 'XX==' sequence indicates that the last group contained |
179 | | * only one byte, and 'XXX=' indicates that it contained two |
180 | | * bytes." (wikipedia) |
181 | | * |
182 | | */ |
183 | |
|
184 | 0 | if (s->len >= 2) |
185 | 0 | *out++ = (uint8_t)(s->quad[0] << 2 | s->quad[1] >> 4); |
186 | |
|
187 | 0 | if (s->len >= 3 && equals != 2) |
188 | 0 | *out++ = (uint8_t)(s->quad[1] << 4 | s->quad[2] >> 2); |
189 | |
|
190 | 0 | if (s->len >= 4 && equals != 1) |
191 | 0 | *out++ = (uint8_t)(((s->quad[2] << 6) & 0xc0) | s->quad[3]); |
192 | |
|
193 | 0 | s->done += s->len - 1; |
194 | 0 | s->len = 0; |
195 | 0 | } |
196 | | |
197 | 0 | if (out < end_out) |
198 | 0 | *out = '\0'; |
199 | |
|
200 | 0 | *in_len = (unsigned int)(in - orig_in); |
201 | 0 | *out_size = (unsigned int)(out - orig_out); |
202 | |
|
203 | 0 | return 0; |
204 | 0 | } |
205 | | |
206 | | |
207 | | /* |
208 | | * returns length of decoded string in out, or -1 if out was too small |
209 | | * according to out_size |
210 | | * |
211 | | * Only reads up to in_len chars, otherwise if in_len is -1 on entry reads until |
212 | | * the first NUL in the input. |
213 | | */ |
214 | | |
215 | | static size_t |
216 | | _lws_b64_decode_string(const char *in, int in_len, char *out, size_t out_size) |
217 | 0 | { |
218 | 0 | struct lws_b64state state; |
219 | 0 | size_t il = (size_t)in_len, ol = out_size; |
220 | |
|
221 | 0 | if (in_len == -1) { |
222 | 0 | il = strlen(in); |
223 | 0 | in_len = (int)il; |
224 | 0 | } |
225 | |
|
226 | 0 | lws_b64_decode_state_init(&state); |
227 | 0 | if (lws_b64_decode_stateful(&state, in, &il, (uint8_t *)out, &ol, 1) < 0) |
228 | | /* pass on the failure */ |
229 | 0 | return 0; |
230 | | |
231 | 0 | if ((int)il != in_len) { |
232 | 0 | lwsl_err("%s: base64 must end at end of input\n", __func__); |
233 | 0 | return 0; |
234 | 0 | } |
235 | | |
236 | 0 | return ol; |
237 | 0 | } |
238 | | |
239 | | int |
240 | | lws_b64_decode_string(const char *in, char *out, int out_size) |
241 | 0 | { |
242 | 0 | return (int)_lws_b64_decode_string(in, -1, out, (unsigned int)out_size); |
243 | 0 | } |
244 | | |
245 | | int |
246 | | lws_b64_decode_string_len(const char *in, int in_len, char *out, int out_size) |
247 | 0 | { |
248 | 0 | size_t s = _lws_b64_decode_string(in, in_len, out, (unsigned int)out_size); |
249 | |
|
250 | 0 | return !s ? -1 : (int)s; |
251 | 0 | } |
252 | | |
253 | | static const char encode_b32[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZ234567"; |
254 | | |
255 | | int |
256 | | lws_b32_encode_string(const char *in, int in_len, char *out, int out_size) |
257 | 0 | { |
258 | 0 | unsigned char buf[5]; |
259 | 0 | int i, done = 0; |
260 | |
|
261 | 0 | while (in_len) { |
262 | 0 | int len = 0; |
263 | 0 | for (i = 0; i < 5; i++) { |
264 | 0 | if (in_len) { |
265 | 0 | buf[i] = (unsigned char)*in++; |
266 | 0 | len++; |
267 | 0 | in_len--; |
268 | 0 | } else |
269 | 0 | buf[i] = 0; |
270 | 0 | } |
271 | |
|
272 | 0 | if (done + 8 >= out_size) |
273 | 0 | return -1; |
274 | | |
275 | 0 | out[0] = encode_b32[buf[0] >> 3]; |
276 | 0 | out[1] = encode_b32[((buf[0] & 0x07) << 2) | (buf[1] >> 6)]; |
277 | 0 | out[2] = len > 1 ? encode_b32[((buf[1] & 0x3e) >> 1)] : '='; |
278 | 0 | out[3] = len > 1 ? encode_b32[((buf[1] & 0x01) << 4) | (buf[2] >> 4)] : '='; |
279 | 0 | out[4] = len > 2 ? encode_b32[((buf[2] & 0x0f) << 1) | (buf[3] >> 7)] : '='; |
280 | 0 | out[5] = len > 3 ? encode_b32[((buf[3] & 0x7c) >> 2)] : '='; |
281 | 0 | out[6] = len > 3 ? encode_b32[((buf[3] & 0x03) << 3) | (buf[4] >> 5)] : '='; |
282 | 0 | out[7] = len > 4 ? encode_b32[(buf[4] & 0x1f)] : '='; |
283 | |
|
284 | 0 | out += 8; |
285 | 0 | done += 8; |
286 | 0 | } |
287 | | |
288 | 0 | if (done + 1 >= out_size) |
289 | 0 | return -1; |
290 | | |
291 | 0 | *out++ = '\0'; |
292 | |
|
293 | 0 | return done; |
294 | 0 | } |
295 | | |
296 | | static const int8_t decode_b32[256] = { |
297 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
298 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
299 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
300 | | -1,-1,26,27,28,29,30,31,-1,-1,-1,-1,-1,-1,-1,-1, |
301 | | -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14, |
302 | | 15,16,17,18,19,20,21,22,23,24,25,-1,-1,-1,-1,-1, |
303 | | -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14, |
304 | | 15,16,17,18,19,20,21,22,23,24,25,-1,-1,-1,-1,-1, |
305 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
306 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
307 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
308 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
309 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
310 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
311 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
312 | | -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, |
313 | | }; |
314 | | |
315 | | int |
316 | | lws_b32_decode_string_len(const char *in, int in_len, char *out, int out_size) |
317 | 0 | { |
318 | 0 | int done = 0; |
319 | 0 | int buf[8] = {0}; |
320 | 0 | int i; |
321 | |
|
322 | 0 | if (in_len == -1) |
323 | 0 | in_len = (int)strlen(in); |
324 | |
|
325 | 0 | while (in_len > 0) { |
326 | 0 | int len = 0; |
327 | 0 | for (i = 0; i < 8; i++) { |
328 | 0 | while (in_len > 0 && (*in == ' ' || *in == '\t' || *in == '\n' || *in == '\r')) { |
329 | 0 | in++; |
330 | 0 | in_len--; |
331 | 0 | } |
332 | 0 | if (in_len > 0) { |
333 | 0 | char c = *in++; |
334 | 0 | in_len--; |
335 | 0 | if (c == '=') { |
336 | 0 | buf[i] = 0; |
337 | 0 | } else { |
338 | 0 | int8_t v = decode_b32[(unsigned char)c]; |
339 | 0 | if (v < 0) return -1; |
340 | 0 | buf[i] = v; |
341 | 0 | len++; |
342 | 0 | } |
343 | 0 | } else { |
344 | 0 | buf[i] = 0; |
345 | 0 | } |
346 | 0 | } |
347 | | |
348 | 0 | if (len == 0) break; |
349 | | |
350 | 0 | if (done + 5 > out_size) return -1; |
351 | | |
352 | 0 | out[0] = (char)((buf[0] << 3) | (buf[1] >> 2)); |
353 | 0 | out[1] = (char)(((buf[1] & 0x03) << 6) | (buf[2] << 1) | (buf[3] >> 4)); |
354 | 0 | out[2] = (char)(((buf[3] & 0x0f) << 4) | (buf[4] >> 1)); |
355 | 0 | out[3] = (char)(((buf[4] & 0x01) << 7) | (buf[5] << 2) | (buf[6] >> 3)); |
356 | 0 | out[4] = (char)(((buf[6] & 0x07) << 5) | buf[7]); |
357 | |
|
358 | 0 | if (len == 2) done += 1; |
359 | 0 | else if (len == 4) done += 2; |
360 | 0 | else if (len == 5) done += 3; |
361 | 0 | else if (len == 7) done += 4; |
362 | 0 | else if (len == 8) done += 5; |
363 | 0 | else return -1; /* invalid base32 chunk */ |
364 | | |
365 | 0 | out += 5; |
366 | 0 | } |
367 | | |
368 | 0 | if (done < out_size) |
369 | 0 | *out = '\0'; |
370 | |
|
371 | 0 | return done; |
372 | 0 | } |
373 | | |
374 | | int |
375 | | lws_b32_decode_string(const char *in, char *out, int out_size) |
376 | 0 | { |
377 | 0 | return lws_b32_decode_string_len(in, -1, out, out_size); |
378 | 0 | } |
379 | | |
380 | | #if 0 |
381 | | static const char * const plaintext[] = { |
382 | | "any carnal pleasure.", |
383 | | "any carnal pleasure", |
384 | | "any carnal pleasur", |
385 | | "any carnal pleasu", |
386 | | "any carnal pleas", |
387 | | "Admin:kloikloi" |
388 | | }; |
389 | | static const char * const coded[] = { |
390 | | "YW55IGNhcm5hbCBwbGVhc3VyZS4=", |
391 | | "YW55IGNhcm5hbCBwbGVhc3VyZQ==", |
392 | | "YW55IGNhcm5hbCBwbGVhc3Vy", |
393 | | "YW55IGNhcm5hbCBwbGVhc3U=", |
394 | | "YW55IGNhcm5hbCBwbGVhcw==", |
395 | | "QWRtaW46a2xvaWtsb2k=" |
396 | | }; |
397 | | |
398 | | int |
399 | | lws_b64_selftest(void) |
400 | | { |
401 | | char buf[64]; |
402 | | unsigned int n, r = 0; |
403 | | unsigned int test; |
404 | | |
405 | | lwsl_notice("%s\n", __func__); |
406 | | |
407 | | /* examples from https://en.wikipedia.org/wiki/Base64 */ |
408 | | |
409 | | for (test = 0; test < (int)LWS_ARRAY_SIZE(plaintext); test++) { |
410 | | |
411 | | buf[sizeof(buf) - 1] = '\0'; |
412 | | n = lws_b64_encode_string(plaintext[test], |
413 | | strlen(plaintext[test]), buf, sizeof buf); |
414 | | if (n != strlen(coded[test]) || strcmp(buf, coded[test])) { |
415 | | lwsl_err("Failed lws_b64 encode selftest " |
416 | | "%d result '%s' %d\n", test, buf, n); |
417 | | r = -1; |
418 | | } |
419 | | |
420 | | buf[sizeof(buf) - 1] = '\0'; |
421 | | n = lws_b64_decode_string(coded[test], buf, sizeof buf); |
422 | | if (n != strlen(plaintext[test]) || |
423 | | strcmp(buf, plaintext[test])) { |
424 | | lwsl_err("Failed lws_b64 decode selftest " |
425 | | "%d result '%s' / '%s', %d / %zu\n", |
426 | | test, buf, plaintext[test], n, |
427 | | strlen(plaintext[test])); |
428 | | lwsl_hexdump_err(buf, n); |
429 | | r = -1; |
430 | | } |
431 | | } |
432 | | |
433 | | if (!r) |
434 | | lwsl_notice("Base 64 selftests passed\n"); |
435 | | else |
436 | | lwsl_notice("Base64 selftests failed\n"); |
437 | | |
438 | | return r; |
439 | | } |
440 | | #endif |