/src/unbound/sldns/parse.c
Line | Count | Source (jump to first uncovered line) |
1 | | /* |
2 | | * a generic (simple) parser. Use to parse rr's, private key |
3 | | * information and /etc/resolv.conf files |
4 | | * |
5 | | * a Net::DNS like library for C |
6 | | * LibDNS Team @ NLnet Labs |
7 | | * (c) NLnet Labs, 2005-2006 |
8 | | * See the file LICENSE for the license |
9 | | */ |
10 | | #include "config.h" |
11 | | #include "sldns/parse.h" |
12 | | #include "sldns/parseutil.h" |
13 | | #include "sldns/sbuffer.h" |
14 | | |
15 | | #include <limits.h> |
16 | | #include <strings.h> |
17 | | |
18 | | sldns_lookup_table sldns_directive_types[] = { |
19 | | { LDNS_DIR_TTL, "$TTL" }, |
20 | | { LDNS_DIR_ORIGIN, "$ORIGIN" }, |
21 | | { LDNS_DIR_INCLUDE, "$INCLUDE" }, |
22 | | { 0, NULL } |
23 | | }; |
24 | | |
25 | | /* add max_limit here? */ |
26 | | ssize_t |
27 | | sldns_fget_token(FILE *f, char *token, const char *delim, size_t limit) |
28 | 0 | { |
29 | 0 | return sldns_fget_token_l(f, token, delim, limit, NULL); |
30 | 0 | } |
31 | | |
32 | | ssize_t |
33 | | sldns_fget_token_l(FILE *f, char *token, const char *delim, size_t limit, int *line_nr) |
34 | 0 | { |
35 | 0 | int c, prev_c; |
36 | 0 | int p; /* 0 -> no parentheses seen, >0 nr of ( seen */ |
37 | 0 | int com, quoted, only_blank; |
38 | 0 | char *t; |
39 | 0 | size_t i; |
40 | 0 | const char *d; |
41 | 0 | const char *del; |
42 | | |
43 | | /* standard delimiters */ |
44 | 0 | if (!delim) { |
45 | | /* from isspace(3) */ |
46 | 0 | del = LDNS_PARSE_NORMAL; |
47 | 0 | } else { |
48 | 0 | del = delim; |
49 | 0 | } |
50 | |
|
51 | 0 | p = 0; |
52 | 0 | i = 0; |
53 | 0 | com = 0; |
54 | 0 | quoted = 0; |
55 | 0 | prev_c = 0; |
56 | 0 | only_blank = 1; /* Assume we got only <blank> until now */ |
57 | 0 | t = token; |
58 | 0 | if (del[0] == '"') { |
59 | 0 | quoted = 1; |
60 | 0 | } |
61 | 0 | while ((c = getc(f)) != EOF) { |
62 | 0 | if (c == '\r') /* carriage return */ |
63 | 0 | c = ' '; |
64 | 0 | if (c == '(' && prev_c != '\\' && !quoted) { |
65 | | /* this only counts for non-comments */ |
66 | 0 | if (com == 0) { |
67 | 0 | p++; |
68 | 0 | } |
69 | 0 | prev_c = c; |
70 | 0 | continue; |
71 | 0 | } |
72 | | |
73 | 0 | if (c == ')' && prev_c != '\\' && !quoted) { |
74 | | /* this only counts for non-comments */ |
75 | 0 | if (com == 0) { |
76 | 0 | p--; |
77 | 0 | } |
78 | 0 | prev_c = c; |
79 | 0 | continue; |
80 | 0 | } |
81 | | |
82 | 0 | if (p < 0) { |
83 | | /* more ) then ( - close off the string */ |
84 | 0 | *t = '\0'; |
85 | 0 | return 0; |
86 | 0 | } |
87 | | |
88 | | /* do something with comments ; */ |
89 | 0 | if (c == ';' && quoted == 0) { |
90 | 0 | if (prev_c != '\\') { |
91 | 0 | com = 1; |
92 | 0 | } |
93 | 0 | } |
94 | 0 | if (c == '\"' && com == 0 && prev_c != '\\') { |
95 | 0 | quoted = 1 - quoted; |
96 | 0 | } |
97 | |
|
98 | 0 | if (c == '\n' && com != 0) { |
99 | | /* comments */ |
100 | 0 | com = 0; |
101 | 0 | *t = ' '; |
102 | 0 | if (line_nr) { |
103 | 0 | *line_nr = *line_nr + 1; |
104 | 0 | } |
105 | 0 | if (only_blank && i > 0) { |
106 | | /* Got only <blank> so far. Reset and try |
107 | | * again with the next line. |
108 | | */ |
109 | 0 | i = 0; |
110 | 0 | t = token; |
111 | 0 | } |
112 | 0 | if (p == 0) { |
113 | | /* If p != 0 then the next line is a continuation. So |
114 | | * we assume that the next line starts with a blank only |
115 | | * if it is actually a new line. |
116 | | */ |
117 | 0 | only_blank = 1; /* Assume next line starts with |
118 | | * <blank>. |
119 | | */ |
120 | 0 | } |
121 | 0 | if (p == 0 && i > 0) { |
122 | 0 | goto tokenread; |
123 | 0 | } else { |
124 | 0 | prev_c = c; |
125 | 0 | continue; |
126 | 0 | } |
127 | 0 | } |
128 | | |
129 | 0 | if (com == 1) { |
130 | 0 | *t = ' '; |
131 | 0 | prev_c = c; |
132 | 0 | continue; |
133 | 0 | } |
134 | | |
135 | 0 | if (c == '\n' && p != 0 && t > token) { |
136 | | /* in parentheses */ |
137 | 0 | if (line_nr) { |
138 | 0 | *line_nr = *line_nr + 1; |
139 | 0 | } |
140 | 0 | if (limit > 0 && (i+1 >= limit || (size_t)(t-token)+1 >= limit)) { |
141 | 0 | *t = '\0'; |
142 | 0 | return -1; |
143 | 0 | } |
144 | 0 | *t++ = ' '; |
145 | 0 | prev_c = c; |
146 | 0 | continue; |
147 | 0 | } |
148 | | |
149 | | /* check if we hit the delim */ |
150 | 0 | for (d = del; *d; d++) { |
151 | 0 | if (c == *d) |
152 | 0 | break; |
153 | 0 | } |
154 | |
|
155 | 0 | if (c == *d && i > 0 && prev_c != '\\' && p == 0) { |
156 | 0 | if (c == '\n' && line_nr) { |
157 | 0 | *line_nr = *line_nr + 1; |
158 | 0 | } |
159 | 0 | if (only_blank) { |
160 | | /* Got only <blank> so far. Reset and |
161 | | * try again with the next line. |
162 | | */ |
163 | 0 | i = 0; |
164 | 0 | t = token; |
165 | 0 | only_blank = 1; |
166 | 0 | prev_c = c; |
167 | 0 | continue; |
168 | 0 | } |
169 | 0 | goto tokenread; |
170 | 0 | } |
171 | 0 | if (c != ' ' && c != '\t') { |
172 | | /* Found something that is not <blank> */ |
173 | 0 | only_blank= 0; |
174 | 0 | } |
175 | 0 | if (c != '\0' && c != '\n') { |
176 | 0 | i++; |
177 | 0 | } |
178 | | /* is there space for the character and the zero after it */ |
179 | 0 | if (limit > 0 && (i+1 >= limit || (size_t)(t-token)+1 >= limit)) { |
180 | 0 | *t = '\0'; |
181 | 0 | return -1; |
182 | 0 | } |
183 | 0 | if (c != '\0' && c != '\n') { |
184 | 0 | *t++ = c; |
185 | 0 | } |
186 | 0 | if (c == '\n') { |
187 | 0 | if (line_nr) { |
188 | 0 | *line_nr = *line_nr + 1; |
189 | 0 | } |
190 | 0 | only_blank = 1; /* Assume next line starts with |
191 | | * <blank>. |
192 | | */ |
193 | 0 | } |
194 | 0 | if (c == '\\' && prev_c == '\\') |
195 | 0 | prev_c = 0; |
196 | 0 | else prev_c = c; |
197 | 0 | } |
198 | 0 | *t = '\0'; |
199 | 0 | if (c == EOF) { |
200 | 0 | return (ssize_t)i; |
201 | 0 | } |
202 | | |
203 | 0 | if (i == 0) { |
204 | | /* nothing read */ |
205 | 0 | return -1; |
206 | 0 | } |
207 | 0 | if (p != 0) { |
208 | 0 | return -1; |
209 | 0 | } |
210 | 0 | return (ssize_t)i; |
211 | | |
212 | 0 | tokenread: |
213 | 0 | if(*del == '"') |
214 | | /* do not skip over quotes after the string, they are part |
215 | | * of the next string. But skip over whitespace (if needed)*/ |
216 | 0 | sldns_fskipcs_l(f, del+1, line_nr); |
217 | 0 | else sldns_fskipcs_l(f, del, line_nr); |
218 | 0 | *t = '\0'; |
219 | 0 | if (p != 0) { |
220 | 0 | return -1; |
221 | 0 | } |
222 | | |
223 | 0 | return (ssize_t)i; |
224 | 0 | } |
225 | | |
226 | | ssize_t |
227 | | sldns_fget_keyword_data(FILE *f, const char *keyword, const char *k_del, char *data, |
228 | | const char *d_del, size_t data_limit) |
229 | 0 | { |
230 | 0 | return sldns_fget_keyword_data_l(f, keyword, k_del, data, d_del, |
231 | 0 | data_limit, NULL); |
232 | 0 | } |
233 | | |
234 | | ssize_t |
235 | | sldns_fget_keyword_data_l(FILE *f, const char *keyword, const char *k_del, char *data, |
236 | | const char *d_del, size_t data_limit, int *line_nr) |
237 | 0 | { |
238 | | /* we assume: keyword|sep|data */ |
239 | 0 | char *fkeyword; |
240 | 0 | ssize_t i; |
241 | |
|
242 | 0 | if(strlen(keyword) >= LDNS_MAX_KEYWORDLEN) |
243 | 0 | return -1; |
244 | 0 | fkeyword = (char*)malloc(LDNS_MAX_KEYWORDLEN); |
245 | 0 | if(!fkeyword) |
246 | 0 | return -1; |
247 | | |
248 | 0 | i = sldns_fget_token(f, fkeyword, k_del, LDNS_MAX_KEYWORDLEN); |
249 | 0 | if(i==0 || i==-1) { |
250 | 0 | free(fkeyword); |
251 | 0 | return -1; |
252 | 0 | } |
253 | | |
254 | | /* case??? i instead of strlen? */ |
255 | 0 | if (strncmp(fkeyword, keyword, LDNS_MAX_KEYWORDLEN - 1) == 0) { |
256 | | /* whee! */ |
257 | | /* printf("%s\n%s\n", "Matching keyword", fkeyword); */ |
258 | 0 | i = sldns_fget_token_l(f, data, d_del, data_limit, line_nr); |
259 | 0 | free(fkeyword); |
260 | 0 | return i; |
261 | 0 | } else { |
262 | | /*printf("no match for %s (read: %s)\n", keyword, fkeyword);*/ |
263 | 0 | free(fkeyword); |
264 | 0 | return -1; |
265 | 0 | } |
266 | 0 | } |
267 | | |
268 | | int |
269 | | sldns_bgetc(sldns_buffer *buffer) |
270 | 0 | { |
271 | 0 | if (!sldns_buffer_available_at(buffer, buffer->_position, sizeof(uint8_t))) { |
272 | 0 | sldns_buffer_set_position(buffer, sldns_buffer_limit(buffer)); |
273 | | /* sldns_buffer_rewind(buffer);*/ |
274 | 0 | return EOF; |
275 | 0 | } |
276 | 0 | return (int)sldns_buffer_read_u8(buffer); |
277 | 0 | } |
278 | | |
279 | | ssize_t |
280 | | sldns_bget_token(sldns_buffer *b, char *token, const char *delim, size_t limit) |
281 | 0 | { |
282 | 0 | return sldns_bget_token_par(b, token, delim, limit, NULL, NULL); |
283 | 0 | } |
284 | | |
285 | | ssize_t |
286 | | sldns_bget_token_par(sldns_buffer *b, char *token, const char *delim, |
287 | | size_t limit, int* par, const char* skipw) |
288 | 0 | { |
289 | 0 | int c, lc; |
290 | 0 | int p; /* 0 -> no parentheses seen, >0 nr of ( seen */ |
291 | 0 | int com, quoted; |
292 | 0 | char *t; |
293 | 0 | size_t i; |
294 | 0 | const char *d; |
295 | 0 | const char *del; |
296 | | |
297 | | /* standard delimiters */ |
298 | 0 | if (!delim) { |
299 | | /* from isspace(3) */ |
300 | 0 | del = LDNS_PARSE_NORMAL; |
301 | 0 | } else { |
302 | 0 | del = delim; |
303 | 0 | } |
304 | |
|
305 | 0 | p = (par?*par:0); |
306 | 0 | i = 0; |
307 | 0 | com = 0; |
308 | 0 | quoted = 0; |
309 | 0 | t = token; |
310 | 0 | lc = 0; |
311 | 0 | if (del[0] == '"') { |
312 | 0 | quoted = 1; |
313 | 0 | } |
314 | |
|
315 | 0 | while ((c = sldns_bgetc(b)) != EOF) { |
316 | 0 | if (c == '\r') /* carriage return */ |
317 | 0 | c = ' '; |
318 | 0 | if (c == '(' && lc != '\\' && !quoted) { |
319 | | /* this only counts for non-comments */ |
320 | 0 | if (com == 0) { |
321 | 0 | if(par) (*par)++; |
322 | 0 | p++; |
323 | 0 | } |
324 | 0 | lc = c; |
325 | 0 | continue; |
326 | 0 | } |
327 | | |
328 | 0 | if (c == ')' && lc != '\\' && !quoted) { |
329 | | /* this only counts for non-comments */ |
330 | 0 | if (com == 0) { |
331 | 0 | if(par) (*par)--; |
332 | 0 | p--; |
333 | 0 | } |
334 | 0 | lc = c; |
335 | 0 | continue; |
336 | 0 | } |
337 | | |
338 | 0 | if (p < 0) { |
339 | | /* more ) then ( */ |
340 | 0 | *t = '\0'; |
341 | 0 | return 0; |
342 | 0 | } |
343 | | |
344 | | /* do something with comments ; */ |
345 | 0 | if (c == ';' && quoted == 0) { |
346 | 0 | if (lc != '\\') { |
347 | 0 | com = 1; |
348 | 0 | } |
349 | 0 | } |
350 | 0 | if (c == '"' && com == 0 && lc != '\\') { |
351 | 0 | quoted = 1 - quoted; |
352 | 0 | } |
353 | |
|
354 | 0 | if (c == '\n' && com != 0) { |
355 | | /* comments */ |
356 | 0 | com = 0; |
357 | 0 | *t = ' '; |
358 | 0 | lc = c; |
359 | 0 | continue; |
360 | 0 | } |
361 | | |
362 | 0 | if (com == 1) { |
363 | 0 | *t = ' '; |
364 | 0 | lc = c; |
365 | 0 | continue; |
366 | 0 | } |
367 | | |
368 | 0 | if (c == '\n' && p != 0) { |
369 | | /* in parentheses */ |
370 | | /* do not write ' ' if we want to skip spaces */ |
371 | 0 | if(!(skipw && (strchr(skipw, c)||strchr(skipw, ' ')))) { |
372 | | /* check for space for the space character and a zero delimiter after that. */ |
373 | 0 | if (limit > 0 && (i+1 >= limit || (size_t)(t-token)+1 >= limit)) { |
374 | 0 | *t = '\0'; |
375 | 0 | return -1; |
376 | 0 | } |
377 | 0 | *t++ = ' '; |
378 | 0 | } |
379 | 0 | lc = c; |
380 | 0 | continue; |
381 | 0 | } |
382 | | |
383 | | /* check to skip whitespace at start, but also after ( */ |
384 | 0 | if(skipw && i==0 && !com && !quoted && lc != '\\') { |
385 | 0 | if(strchr(skipw, c)) { |
386 | 0 | lc = c; |
387 | 0 | continue; |
388 | 0 | } |
389 | 0 | } |
390 | | |
391 | | /* check if we hit the delim */ |
392 | 0 | for (d = del; *d; d++) { |
393 | | /* we can only exit if no parens or user tracks them */ |
394 | 0 | if (c == *d && lc != '\\' && (p == 0 || par)) { |
395 | 0 | goto tokenread; |
396 | 0 | } |
397 | 0 | } |
398 | | |
399 | 0 | i++; |
400 | 0 | if (limit > 0 && (i+1 >= limit || (size_t)(t-token)+1 >= limit)) { |
401 | 0 | *t = '\0'; |
402 | 0 | return -1; |
403 | 0 | } |
404 | 0 | *t++ = c; |
405 | |
|
406 | 0 | if (c == '\\' && lc == '\\') { |
407 | 0 | lc = 0; |
408 | 0 | } else { |
409 | 0 | lc = c; |
410 | 0 | } |
411 | 0 | } |
412 | 0 | *t = '\0'; |
413 | 0 | if (i == 0) { |
414 | | /* nothing read */ |
415 | 0 | return -1; |
416 | 0 | } |
417 | 0 | if (!par && p != 0) { |
418 | 0 | return -1; |
419 | 0 | } |
420 | 0 | return (ssize_t)i; |
421 | | |
422 | 0 | tokenread: |
423 | 0 | if(*del == '"') |
424 | | /* do not skip over quotes after the string, they are part |
425 | | * of the next string. But skip over whitespace (if needed)*/ |
426 | 0 | sldns_bskipcs(b, del+1); |
427 | 0 | else sldns_bskipcs(b, del); |
428 | 0 | *t = '\0'; |
429 | |
|
430 | 0 | if (!par && p != 0) { |
431 | 0 | return -1; |
432 | 0 | } |
433 | 0 | return (ssize_t)i; |
434 | 0 | } |
435 | | |
436 | | |
437 | | void |
438 | | sldns_bskipcs(sldns_buffer *buffer, const char *s) |
439 | 0 | { |
440 | 0 | int found; |
441 | 0 | char c; |
442 | 0 | const char *d; |
443 | |
|
444 | 0 | while(sldns_buffer_available_at(buffer, buffer->_position, sizeof(char))) { |
445 | 0 | c = (char) sldns_buffer_read_u8_at(buffer, buffer->_position); |
446 | 0 | found = 0; |
447 | 0 | for (d = s; *d; d++) { |
448 | 0 | if (*d == c) { |
449 | 0 | found = 1; |
450 | 0 | } |
451 | 0 | } |
452 | 0 | if (found && buffer->_limit > buffer->_position) { |
453 | 0 | buffer->_position += sizeof(char); |
454 | 0 | } else { |
455 | 0 | return; |
456 | 0 | } |
457 | 0 | } |
458 | 0 | } |
459 | | |
460 | | void |
461 | | sldns_fskipcs(FILE *fp, const char *s) |
462 | 0 | { |
463 | 0 | sldns_fskipcs_l(fp, s, NULL); |
464 | 0 | } |
465 | | |
466 | | void |
467 | | sldns_fskipcs_l(FILE *fp, const char *s, int *line_nr) |
468 | 0 | { |
469 | 0 | int found; |
470 | 0 | int c; |
471 | 0 | const char *d; |
472 | |
|
473 | 0 | while ((c = fgetc(fp)) != EOF) { |
474 | 0 | if (line_nr && c == '\n') { |
475 | 0 | *line_nr = *line_nr + 1; |
476 | 0 | } |
477 | 0 | found = 0; |
478 | 0 | for (d = s; *d; d++) { |
479 | 0 | if (*d == c) { |
480 | 0 | found = 1; |
481 | 0 | } |
482 | 0 | } |
483 | 0 | if (!found) { |
484 | | /* with getc, we've read too far */ |
485 | 0 | ungetc(c, fp); |
486 | 0 | return; |
487 | 0 | } |
488 | 0 | } |
489 | 0 | } |
490 | | |
491 | | ssize_t |
492 | | sldns_bget_keyword_data(sldns_buffer *b, const char *keyword, const char *k_del, char |
493 | | *data, const char *d_del, size_t data_limit) |
494 | 0 | { |
495 | | /* we assume: keyword|sep|data */ |
496 | 0 | char *fkeyword; |
497 | 0 | ssize_t i; |
498 | |
|
499 | 0 | if(strlen(keyword) >= LDNS_MAX_KEYWORDLEN) |
500 | 0 | return -1; |
501 | 0 | fkeyword = (char*)malloc(LDNS_MAX_KEYWORDLEN); |
502 | 0 | if(!fkeyword) |
503 | 0 | return -1; /* out of memory */ |
504 | | |
505 | 0 | i = sldns_bget_token(b, fkeyword, k_del, data_limit); |
506 | 0 | if(i==0 || i==-1) { |
507 | 0 | free(fkeyword); |
508 | 0 | return -1; /* nothing read */ |
509 | 0 | } |
510 | | |
511 | | /* case??? */ |
512 | 0 | if (strncmp(fkeyword, keyword, strlen(keyword)) == 0) { |
513 | 0 | free(fkeyword); |
514 | | /* whee, the match! */ |
515 | | /* retrieve it's data */ |
516 | 0 | i = sldns_bget_token(b, data, d_del, 0); |
517 | 0 | return i; |
518 | 0 | } else { |
519 | 0 | free(fkeyword); |
520 | 0 | return -1; |
521 | 0 | } |
522 | 0 | } |
523 | | |