1 /* Prosody IM
2 -- Copyright (C) 2008-2010 Matthew Wild
3 -- Copyright (C) 2008-2010 Waqas Hussain
4 -- Copyright (C) 1994-2015 Lua.org, PUC-Rio.
5 --
6 -- This project is MIT/X11 licensed. Please see the
7 -- COPYING file in the source package for more information.
8 --
9 */
10
11 /*
12 * encodings.c
13 * Lua library for base64, stringprep and idna encodings
14 */
15
16 /* Newer MSVC compilers deprecate strcpy as unsafe, but we use it in a safe way */
17 #define _CRT_SECURE_NO_DEPRECATE
18
19 #include <string.h>
20 #include <stdlib.h>
21 #include "lua.h"
22 #include "lauxlib.h"
23
24 #if (LUA_VERSION_NUM == 501)
25 #define luaL_setfuncs(L, R, N) luaL_register(L, NULL, R)
26 #endif
27
28 /***************** BASE64 *****************/
29
30 static const char code[] =
31 "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
32
base64_encode(luaL_Buffer * b,unsigned int c1,unsigned int c2,unsigned int c3,int n)33 static void base64_encode(luaL_Buffer *b, unsigned int c1, unsigned int c2, unsigned int c3, int n) {
34 unsigned long tuple = c3 + 256UL * (c2 + 256UL * c1);
35 int i;
36 char s[4];
37
38 for(i = 0; i < 4; i++) {
39 s[3 - i] = code[tuple % 64];
40 tuple /= 64;
41 }
42
43 for(i = n + 1; i < 4; i++) {
44 s[i] = '=';
45 }
46
47 luaL_addlstring(b, s, 4);
48 }
49
Lbase64_encode(lua_State * L)50 static int Lbase64_encode(lua_State *L) { /** encode(s) */
51 size_t l;
52 const unsigned char *s = (const unsigned char *)luaL_checklstring(L, 1, &l);
53 luaL_Buffer b;
54 int n;
55 luaL_buffinit(L, &b);
56
57 for(n = l / 3; n--; s += 3) {
58 base64_encode(&b, s[0], s[1], s[2], 3);
59 }
60
61 switch(l % 3) {
62 case 1:
63 base64_encode(&b, s[0], 0, 0, 1);
64 break;
65
66 case 2:
67 base64_encode(&b, s[0], s[1], 0, 2);
68 break;
69 }
70
71 luaL_pushresult(&b);
72 return 1;
73 }
74
base64_decode(luaL_Buffer * b,int c1,int c2,int c3,int c4,int n)75 static void base64_decode(luaL_Buffer *b, int c1, int c2, int c3, int c4, int n) {
76 unsigned long tuple = c4 + 64L * (c3 + 64L * (c2 + 64L * c1));
77 char s[3];
78
79 switch(--n) {
80 case 3:
81 s[2] = (char) tuple;
82 /* Falls through. */
83
84 case 2:
85 s[1] = (char)(tuple >> 8);
86 /* Falls through. */
87
88 case 1:
89 s[0] = (char)(tuple >> 16);
90 }
91
92 luaL_addlstring(b, s, n);
93 }
94
Lbase64_decode(lua_State * L)95 static int Lbase64_decode(lua_State *L) { /** decode(s) */
96 size_t l;
97 const char *s = luaL_checklstring(L, 1, &l);
98 luaL_Buffer b;
99 int n = 0;
100 char t[4];
101 luaL_buffinit(L, &b);
102
103 for(;;) {
104 int c = *s++;
105
106 switch(c) {
107 const char *p;
108
109 default:
110 p = strchr(code, c);
111
112 if(p == NULL) {
113 return 0;
114 }
115
116 t[n++] = (char)(p - code);
117
118 if(n == 4) {
119 base64_decode(&b, t[0], t[1], t[2], t[3], 4);
120 n = 0;
121 }
122
123 break;
124
125 case '=':
126
127 switch(n) {
128 case 1:
129 base64_decode(&b, t[0], 0, 0, 0, 1);
130 break;
131
132 case 2:
133 base64_decode(&b, t[0], t[1], 0, 0, 2);
134 break;
135
136 case 3:
137 base64_decode(&b, t[0], t[1], t[2], 0, 3);
138 break;
139 }
140
141 n = 0;
142 break;
143
144 case 0:
145 luaL_pushresult(&b);
146 return 1;
147
148 case '\n':
149 case '\r':
150 case '\t':
151 case ' ':
152 case '\f':
153 case '\b':
154 break;
155 }
156 }
157 }
158
159 static const luaL_Reg Reg_base64[] = {
160 { "encode", Lbase64_encode },
161 { "decode", Lbase64_decode },
162 { NULL, NULL }
163 };
164
165 /******************* UTF-8 ********************/
166
167 /*
168 * Adapted from Lua 5.3
169 * Needed because libidn does not validate that input is valid UTF-8
170 */
171
172 #define MAXUNICODE 0x10FFFF
173
174 /*
175 * Decode one UTF-8 sequence, returning NULL if byte sequence is invalid.
176 */
utf8_decode(const char * o,int * val)177 static const char *utf8_decode(const char *o, int *val) {
178 static const unsigned int limits[] = {0xFF, 0x7F, 0x7FF, 0xFFFF};
179 const unsigned char *s = (const unsigned char *)o;
180 unsigned int c = s[0];
181 unsigned int res = 0; /* final result */
182
183 if(c < 0x80) { /* ascii? */
184 res = c;
185 } else {
186 int count = 0; /* to count number of continuation bytes */
187
188 while(c & 0x40) { /* still have continuation bytes? */
189 int cc = s[++count]; /* read next byte */
190
191 if((cc & 0xC0) != 0x80) { /* not a continuation byte? */
192 return NULL; /* invalid byte sequence */
193 }
194
195 res = (res << 6) | (cc & 0x3F); /* add lower 6 bits from cont. byte */
196 c <<= 1; /* to test next bit */
197 }
198
199 res |= ((c & 0x7F) << (count * 5)); /* add first byte */
200
201 if(count > 3 || res > MAXUNICODE || res <= limits[count] || (0xd800 <= res && res <= 0xdfff)) {
202 return NULL; /* invalid byte sequence */
203 }
204
205 s += count; /* skip continuation bytes read */
206 }
207
208 if(val) {
209 *val = res;
210 }
211
212 return (const char *)s + 1; /* +1 to include first byte */
213 }
214
215 /*
216 * Check that a string is valid UTF-8
217 * Returns NULL if not
218 */
check_utf8(lua_State * L,int idx,size_t * l)219 const char *check_utf8(lua_State *L, int idx, size_t *l) {
220 size_t pos, len;
221 const char *s = luaL_checklstring(L, idx, &len);
222 pos = 0;
223
224 while(pos <= len) {
225 const char *s1 = utf8_decode(s + pos, NULL);
226
227 if(s1 == NULL) { /* conversion error? */
228 return NULL;
229 }
230
231 pos = s1 - s;
232 }
233
234 if(l != NULL) {
235 *l = len;
236 }
237
238 return s;
239 }
240
Lutf8_valid(lua_State * L)241 static int Lutf8_valid(lua_State *L) {
242 lua_pushboolean(L, check_utf8(L, 1, NULL) != NULL);
243 return 1;
244 }
245
Lutf8_length(lua_State * L)246 static int Lutf8_length(lua_State *L) {
247 size_t len;
248
249 if(!check_utf8(L, 1, &len)) {
250 lua_pushnil(L);
251 lua_pushliteral(L, "invalid utf8");
252 return 2;
253 }
254
255 lua_pushinteger(L, len);
256 return 1;
257 }
258
259 static const luaL_Reg Reg_utf8[] = {
260 { "valid", Lutf8_valid },
261 { "length", Lutf8_length },
262 { NULL, NULL }
263 };
264
265 /***************** STRINGPREP *****************/
266 #ifdef USE_STRINGPREP_ICU
267
268 #include <unicode/usprep.h>
269 #include <unicode/ustring.h>
270 #include <unicode/utrace.h>
271
icu_stringprep_prep(lua_State * L,const UStringPrepProfile * profile)272 static int icu_stringprep_prep(lua_State *L, const UStringPrepProfile *profile) {
273 size_t input_len;
274 int32_t unprepped_len, prepped_len, output_len;
275 const char *input;
276 char output[1024];
277
278 UChar unprepped[1024]; /* Temporary unicode buffer (1024 characters) */
279 UChar prepped[1024];
280
281 UErrorCode err = U_ZERO_ERROR;
282
283 if(!lua_isstring(L, 1)) {
284 lua_pushnil(L);
285 return 1;
286 }
287
288 input = lua_tolstring(L, 1, &input_len);
289
290 if(input_len >= 1024) {
291 lua_pushnil(L);
292 return 1;
293 }
294
295 u_strFromUTF8(unprepped, 1024, &unprepped_len, input, input_len, &err);
296
297 if(U_FAILURE(err)) {
298 lua_pushnil(L);
299 return 1;
300 }
301
302 prepped_len = usprep_prepare(profile, unprepped, unprepped_len, prepped, 1024, USPREP_ALLOW_UNASSIGNED, NULL, &err);
303
304 if(U_FAILURE(err)) {
305 lua_pushnil(L);
306 return 1;
307 } else {
308 u_strToUTF8(output, 1024, &output_len, prepped, prepped_len, &err);
309
310 if(U_SUCCESS(err) && output_len < 1024) {
311 lua_pushlstring(L, output, output_len);
312 } else {
313 lua_pushnil(L);
314 }
315
316 return 1;
317 }
318 }
319
320 UStringPrepProfile *icu_nameprep;
321 UStringPrepProfile *icu_nodeprep;
322 UStringPrepProfile *icu_resourceprep;
323 UStringPrepProfile *icu_saslprep;
324
325 /* initialize global ICU stringprep profiles */
init_icu()326 void init_icu() {
327 UErrorCode err = U_ZERO_ERROR;
328 utrace_setLevel(UTRACE_VERBOSE);
329 icu_nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, &err);
330 icu_nodeprep = usprep_openByType(USPREP_RFC3920_NODEPREP, &err);
331 icu_resourceprep = usprep_openByType(USPREP_RFC3920_RESOURCEPREP, &err);
332 icu_saslprep = usprep_openByType(USPREP_RFC4013_SASLPREP, &err);
333
334 if(U_FAILURE(err)) {
335 fprintf(stderr, "[c] util.encodings: error: %s\n", u_errorName((UErrorCode)err));
336 }
337 }
338
339 #define MAKE_PREP_FUNC(myFunc, prep) \
340 static int myFunc(lua_State *L) { return icu_stringprep_prep(L, prep); }
341
342 MAKE_PREP_FUNC(Lstringprep_nameprep, icu_nameprep) /** stringprep.nameprep(s) */
343 MAKE_PREP_FUNC(Lstringprep_nodeprep, icu_nodeprep) /** stringprep.nodeprep(s) */
344 MAKE_PREP_FUNC(Lstringprep_resourceprep, icu_resourceprep) /** stringprep.resourceprep(s) */
345 MAKE_PREP_FUNC(Lstringprep_saslprep, icu_saslprep) /** stringprep.saslprep(s) */
346
347 static const luaL_Reg Reg_stringprep[] = {
348 { "nameprep", Lstringprep_nameprep },
349 { "nodeprep", Lstringprep_nodeprep },
350 { "resourceprep", Lstringprep_resourceprep },
351 { "saslprep", Lstringprep_saslprep },
352 { NULL, NULL }
353 };
354 #else /* USE_STRINGPREP_ICU */
355
356 /****************** libidn ********************/
357
358 #include <stringprep.h>
359
stringprep_prep(lua_State * L,const Stringprep_profile * profile)360 static int stringprep_prep(lua_State *L, const Stringprep_profile *profile) {
361 size_t len;
362 const char *s;
363 char string[1024];
364 int ret;
365
366 if(!lua_isstring(L, 1)) {
367 lua_pushnil(L);
368 return 1;
369 }
370
371 s = check_utf8(L, 1, &len);
372
373 if(s == NULL || len >= 1024 || len != strlen(s)) {
374 lua_pushnil(L);
375 return 1; /* TODO return error message */
376 }
377
378 strcpy(string, s);
379 ret = stringprep(string, 1024, (Stringprep_profile_flags)0, profile);
380
381 if(ret == STRINGPREP_OK) {
382 lua_pushstring(L, string);
383 return 1;
384 } else {
385 lua_pushnil(L);
386 return 1; /* TODO return error message */
387 }
388 }
389
390 #define MAKE_PREP_FUNC(myFunc, prep) \
391 static int myFunc(lua_State *L) { return stringprep_prep(L, prep); }
392
393 MAKE_PREP_FUNC(Lstringprep_nameprep, stringprep_nameprep) /** stringprep.nameprep(s) */
394 MAKE_PREP_FUNC(Lstringprep_nodeprep, stringprep_xmpp_nodeprep) /** stringprep.nodeprep(s) */
395 MAKE_PREP_FUNC(Lstringprep_resourceprep, stringprep_xmpp_resourceprep) /** stringprep.resourceprep(s) */
396 MAKE_PREP_FUNC(Lstringprep_saslprep, stringprep_saslprep) /** stringprep.saslprep(s) */
397
398 static const luaL_Reg Reg_stringprep[] = {
399 { "nameprep", Lstringprep_nameprep },
400 { "nodeprep", Lstringprep_nodeprep },
401 { "resourceprep", Lstringprep_resourceprep },
402 { "saslprep", Lstringprep_saslprep },
403 { NULL, NULL }
404 };
405 #endif
406
407 /***************** IDNA *****************/
408 #ifdef USE_STRINGPREP_ICU
409 #include <unicode/ustdio.h>
410 #include <unicode/uidna.h>
411 /* IDNA2003 or IDNA2008 ? ? ? */
Lidna_to_ascii(lua_State * L)412 static int Lidna_to_ascii(lua_State *L) { /** idna.to_ascii(s) */
413 size_t len;
414 int32_t ulen, dest_len, output_len;
415 const char *s = luaL_checklstring(L, 1, &len);
416 UChar ustr[1024];
417 UErrorCode err = U_ZERO_ERROR;
418 UChar dest[1024];
419 char output[1024];
420
421 u_strFromUTF8(ustr, 1024, &ulen, s, len, &err);
422
423 if(U_FAILURE(err)) {
424 lua_pushnil(L);
425 return 1;
426 }
427
428 dest_len = uidna_IDNToASCII(ustr, ulen, dest, 1024, UIDNA_USE_STD3_RULES, NULL, &err);
429
430 if(U_FAILURE(err)) {
431 lua_pushnil(L);
432 return 1;
433 } else {
434 u_strToUTF8(output, 1024, &output_len, dest, dest_len, &err);
435
436 if(U_SUCCESS(err) && output_len < 1024) {
437 lua_pushlstring(L, output, output_len);
438 } else {
439 lua_pushnil(L);
440 }
441
442 return 1;
443 }
444 }
445
Lidna_to_unicode(lua_State * L)446 static int Lidna_to_unicode(lua_State *L) { /** idna.to_unicode(s) */
447 size_t len;
448 int32_t ulen, dest_len, output_len;
449 const char *s = luaL_checklstring(L, 1, &len);
450 UChar ustr[1024];
451 UErrorCode err = U_ZERO_ERROR;
452 UChar dest[1024];
453 char output[1024];
454
455 u_strFromUTF8(ustr, 1024, &ulen, s, len, &err);
456
457 if(U_FAILURE(err)) {
458 lua_pushnil(L);
459 return 1;
460 }
461
462 dest_len = uidna_IDNToUnicode(ustr, ulen, dest, 1024, UIDNA_USE_STD3_RULES, NULL, &err);
463
464 if(U_FAILURE(err)) {
465 lua_pushnil(L);
466 return 1;
467 } else {
468 u_strToUTF8(output, 1024, &output_len, dest, dest_len, &err);
469
470 if(U_SUCCESS(err) && output_len < 1024) {
471 lua_pushlstring(L, output, output_len);
472 } else {
473 lua_pushnil(L);
474 }
475
476 return 1;
477 }
478 }
479
480 #else /* USE_STRINGPREP_ICU */
481 /****************** libidn ********************/
482
483 #include <idna.h>
484 #include <idn-free.h>
485
Lidna_to_ascii(lua_State * L)486 static int Lidna_to_ascii(lua_State *L) { /** idna.to_ascii(s) */
487 size_t len;
488 const char *s = check_utf8(L, 1, &len);
489 char *output = NULL;
490 int ret;
491
492 if(s == NULL || len != strlen(s)) {
493 lua_pushnil(L);
494 return 1; /* TODO return error message */
495 }
496
497 ret = idna_to_ascii_8z(s, &output, IDNA_USE_STD3_ASCII_RULES);
498
499 if(ret == IDNA_SUCCESS) {
500 lua_pushstring(L, output);
501 idn_free(output);
502 return 1;
503 } else {
504 lua_pushnil(L);
505 idn_free(output);
506 return 1; /* TODO return error message */
507 }
508 }
509
Lidna_to_unicode(lua_State * L)510 static int Lidna_to_unicode(lua_State *L) { /** idna.to_unicode(s) */
511 size_t len;
512 const char *s = luaL_checklstring(L, 1, &len);
513 char *output = NULL;
514 int ret = idna_to_unicode_8z8z(s, &output, 0);
515
516 if(ret == IDNA_SUCCESS) {
517 lua_pushstring(L, output);
518 idn_free(output);
519 return 1;
520 } else {
521 lua_pushnil(L);
522 idn_free(output);
523 return 1; /* TODO return error message */
524 }
525 }
526 #endif
527
528 static const luaL_Reg Reg_idna[] = {
529 { "to_ascii", Lidna_to_ascii },
530 { "to_unicode", Lidna_to_unicode },
531 { NULL, NULL }
532 };
533
534 /***************** end *****************/
535
luaopen_util_encodings(lua_State * L)536 LUALIB_API int luaopen_util_encodings(lua_State *L) {
537 #if (LUA_VERSION_NUM > 501)
538 luaL_checkversion(L);
539 #endif
540 #ifdef USE_STRINGPREP_ICU
541 init_icu();
542 #endif
543 lua_newtable(L);
544
545 lua_newtable(L);
546 luaL_setfuncs(L, Reg_base64, 0);
547 lua_setfield(L, -2, "base64");
548
549 lua_newtable(L);
550 luaL_setfuncs(L, Reg_stringprep, 0);
551 lua_setfield(L, -2, "stringprep");
552
553 lua_newtable(L);
554 luaL_setfuncs(L, Reg_idna, 0);
555 lua_setfield(L, -2, "idna");
556
557 lua_newtable(L);
558 luaL_setfuncs(L, Reg_utf8, 0);
559 lua_setfield(L, -2, "utf8");
560
561 lua_pushliteral(L, "-3.14");
562 lua_setfield(L, -2, "version");
563 return 1;
564 }
565