1 | /*************************************************************************** |
2 | * _ _ ____ _ |
3 | * Project ___| | | | _ \| | |
4 | * / __| | | | |_) | | |
5 | * | (__| |_| | _ <| |___ |
6 | * \___|\___/|_| \_\_____| |
7 | * |
8 | * Copyright (C) 1998 - 2020, Daniel Stenberg, <daniel@haxx.se>, et al. |
9 | * |
10 | * This software is licensed as described in the file COPYING, which |
11 | * you should have received as part of this distribution. The terms |
12 | * are also available at https://curl.se/docs/copyright.html. |
13 | * |
14 | * You may opt to use, copy, modify, merge, publish, distribute and/or sell |
15 | * copies of the Software, and permit persons to whom the Software is |
16 | * furnished to do so, under the terms of the COPYING file. |
17 | * |
18 | * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY |
19 | * KIND, either express or implied. |
20 | * |
21 | ***************************************************************************/ |
22 | |
23 | #include "curl_setup.h" |
24 | #ifndef CURL_DISABLE_FTP |
25 | #include <curl/curl.h> |
26 | |
27 | #include "curl_fnmatch.h" |
28 | #include "curl_memory.h" |
29 | |
30 | /* The last #include file should be: */ |
31 | #include "memdebug.h" |
32 | |
33 | #ifndef HAVE_FNMATCH |
34 | |
35 | #define CURLFNM_CHARSET_LEN (sizeof(char) * 256) |
36 | #define CURLFNM_CHSET_SIZE (CURLFNM_CHARSET_LEN + 15) |
37 | |
38 | #define CURLFNM_NEGATE CURLFNM_CHARSET_LEN |
39 | |
40 | #define CURLFNM_ALNUM (CURLFNM_CHARSET_LEN + 1) |
41 | #define CURLFNM_DIGIT (CURLFNM_CHARSET_LEN + 2) |
42 | #define CURLFNM_XDIGIT (CURLFNM_CHARSET_LEN + 3) |
43 | #define CURLFNM_ALPHA (CURLFNM_CHARSET_LEN + 4) |
44 | #define CURLFNM_PRINT (CURLFNM_CHARSET_LEN + 5) |
45 | #define CURLFNM_BLANK (CURLFNM_CHARSET_LEN + 6) |
46 | #define CURLFNM_LOWER (CURLFNM_CHARSET_LEN + 7) |
47 | #define CURLFNM_GRAPH (CURLFNM_CHARSET_LEN + 8) |
48 | #define CURLFNM_SPACE (CURLFNM_CHARSET_LEN + 9) |
49 | #define CURLFNM_UPPER (CURLFNM_CHARSET_LEN + 10) |
50 | |
51 | typedef enum { |
52 | CURLFNM_SCHS_DEFAULT = 0, |
53 | CURLFNM_SCHS_RIGHTBR, |
54 | CURLFNM_SCHS_RIGHTBRLEFTBR |
55 | } setcharset_state; |
56 | |
57 | typedef enum { |
58 | CURLFNM_PKW_INIT = 0, |
59 | CURLFNM_PKW_DDOT |
60 | } parsekey_state; |
61 | |
62 | typedef enum { |
63 | CCLASS_OTHER = 0, |
64 | CCLASS_DIGIT, |
65 | CCLASS_UPPER, |
66 | CCLASS_LOWER |
67 | } char_class; |
68 | |
69 | #define SETCHARSET_OK 1 |
70 | #define SETCHARSET_FAIL 0 |
71 | |
72 | static int parsekeyword(unsigned char **pattern, unsigned char *charset) |
73 | { |
74 | parsekey_state state = CURLFNM_PKW_INIT; |
75 | #define KEYLEN 10 |
76 | char keyword[KEYLEN] = { 0 }; |
77 | int found = FALSE; |
78 | int i; |
79 | unsigned char *p = *pattern; |
80 | for(i = 0; !found; i++) { |
81 | char c = *p++; |
82 | if(i >= KEYLEN) |
83 | return SETCHARSET_FAIL; |
84 | switch(state) { |
85 | case CURLFNM_PKW_INIT: |
86 | if(ISLOWER(c)) |
87 | keyword[i] = c; |
88 | else if(c == ':') |
89 | state = CURLFNM_PKW_DDOT; |
90 | else |
91 | return SETCHARSET_FAIL; |
92 | break; |
93 | case CURLFNM_PKW_DDOT: |
94 | if(c == ']') |
95 | found = TRUE; |
96 | else |
97 | return SETCHARSET_FAIL; |
98 | } |
99 | } |
100 | #undef KEYLEN |
101 | |
102 | *pattern = p; /* move caller's pattern pointer */ |
103 | if(strcmp(keyword, "digit" ) == 0) |
104 | charset[CURLFNM_DIGIT] = 1; |
105 | else if(strcmp(keyword, "alnum" ) == 0) |
106 | charset[CURLFNM_ALNUM] = 1; |
107 | else if(strcmp(keyword, "alpha" ) == 0) |
108 | charset[CURLFNM_ALPHA] = 1; |
109 | else if(strcmp(keyword, "xdigit" ) == 0) |
110 | charset[CURLFNM_XDIGIT] = 1; |
111 | else if(strcmp(keyword, "print" ) == 0) |
112 | charset[CURLFNM_PRINT] = 1; |
113 | else if(strcmp(keyword, "graph" ) == 0) |
114 | charset[CURLFNM_GRAPH] = 1; |
115 | else if(strcmp(keyword, "space" ) == 0) |
116 | charset[CURLFNM_SPACE] = 1; |
117 | else if(strcmp(keyword, "blank" ) == 0) |
118 | charset[CURLFNM_BLANK] = 1; |
119 | else if(strcmp(keyword, "upper" ) == 0) |
120 | charset[CURLFNM_UPPER] = 1; |
121 | else if(strcmp(keyword, "lower" ) == 0) |
122 | charset[CURLFNM_LOWER] = 1; |
123 | else |
124 | return SETCHARSET_FAIL; |
125 | return SETCHARSET_OK; |
126 | } |
127 | |
128 | /* Return the character class. */ |
129 | static char_class charclass(unsigned char c) |
130 | { |
131 | if(ISUPPER(c)) |
132 | return CCLASS_UPPER; |
133 | if(ISLOWER(c)) |
134 | return CCLASS_LOWER; |
135 | if(ISDIGIT(c)) |
136 | return CCLASS_DIGIT; |
137 | return CCLASS_OTHER; |
138 | } |
139 | |
140 | /* Include a character or a range in set. */ |
141 | static void setcharorrange(unsigned char **pp, unsigned char *charset) |
142 | { |
143 | unsigned char *p = (*pp)++; |
144 | unsigned char c = *p++; |
145 | |
146 | charset[c] = 1; |
147 | if(ISALNUM(c) && *p++ == '-') { |
148 | char_class cc = charclass(c); |
149 | unsigned char endrange = *p++; |
150 | |
151 | if(endrange == '\\') |
152 | endrange = *p++; |
153 | if(endrange >= c && charclass(endrange) == cc) { |
154 | while(c++ != endrange) |
155 | if(charclass(c) == cc) /* Chars in class may be not consecutive. */ |
156 | charset[c] = 1; |
157 | *pp = p; |
158 | } |
159 | } |
160 | } |
161 | |
162 | /* returns 1 (true) if pattern is OK, 0 if is bad ("p" is pattern pointer) */ |
163 | static int setcharset(unsigned char **p, unsigned char *charset) |
164 | { |
165 | setcharset_state state = CURLFNM_SCHS_DEFAULT; |
166 | bool something_found = FALSE; |
167 | unsigned char c; |
168 | |
169 | memset(charset, 0, CURLFNM_CHSET_SIZE); |
170 | for(;;) { |
171 | c = **p; |
172 | if(!c) |
173 | return SETCHARSET_FAIL; |
174 | |
175 | switch(state) { |
176 | case CURLFNM_SCHS_DEFAULT: |
177 | if(c == ']') { |
178 | if(something_found) |
179 | return SETCHARSET_OK; |
180 | something_found = TRUE; |
181 | state = CURLFNM_SCHS_RIGHTBR; |
182 | charset[c] = 1; |
183 | (*p)++; |
184 | } |
185 | else if(c == '[') { |
186 | unsigned char *pp = *p + 1; |
187 | |
188 | if(*pp++ == ':' && parsekeyword(&pp, charset)) |
189 | *p = pp; |
190 | else { |
191 | charset[c] = 1; |
192 | (*p)++; |
193 | } |
194 | something_found = TRUE; |
195 | } |
196 | else if(c == '^' || c == '!') { |
197 | if(!something_found) { |
198 | if(charset[CURLFNM_NEGATE]) { |
199 | charset[c] = 1; |
200 | something_found = TRUE; |
201 | } |
202 | else |
203 | charset[CURLFNM_NEGATE] = 1; /* negate charset */ |
204 | } |
205 | else |
206 | charset[c] = 1; |
207 | (*p)++; |
208 | } |
209 | else if(c == '\\') { |
210 | c = *(++(*p)); |
211 | if(c) |
212 | setcharorrange(p, charset); |
213 | else |
214 | charset['\\'] = 1; |
215 | something_found = TRUE; |
216 | } |
217 | else { |
218 | setcharorrange(p, charset); |
219 | something_found = TRUE; |
220 | } |
221 | break; |
222 | case CURLFNM_SCHS_RIGHTBR: |
223 | if(c == '[') { |
224 | state = CURLFNM_SCHS_RIGHTBRLEFTBR; |
225 | charset[c] = 1; |
226 | (*p)++; |
227 | } |
228 | else if(c == ']') { |
229 | return SETCHARSET_OK; |
230 | } |
231 | else if(ISPRINT(c)) { |
232 | charset[c] = 1; |
233 | (*p)++; |
234 | state = CURLFNM_SCHS_DEFAULT; |
235 | } |
236 | else |
237 | /* used 'goto fail' instead of 'return SETCHARSET_FAIL' to avoid a |
238 | * nonsense warning 'statement not reached' at end of the fnc when |
239 | * compiling on Solaris */ |
240 | goto fail; |
241 | break; |
242 | case CURLFNM_SCHS_RIGHTBRLEFTBR: |
243 | if(c == ']') |
244 | return SETCHARSET_OK; |
245 | state = CURLFNM_SCHS_DEFAULT; |
246 | charset[c] = 1; |
247 | (*p)++; |
248 | break; |
249 | } |
250 | } |
251 | fail: |
252 | return SETCHARSET_FAIL; |
253 | } |
254 | |
255 | static int loop(const unsigned char *pattern, const unsigned char *string, |
256 | int maxstars) |
257 | { |
258 | unsigned char *p = (unsigned char *)pattern; |
259 | unsigned char *s = (unsigned char *)string; |
260 | unsigned char charset[CURLFNM_CHSET_SIZE] = { 0 }; |
261 | |
262 | for(;;) { |
263 | unsigned char *pp; |
264 | |
265 | switch(*p) { |
266 | case '*': |
267 | if(!maxstars) |
268 | return CURL_FNMATCH_NOMATCH; |
269 | /* Regroup consecutive stars and question marks. This can be done because |
270 | '*?*?*' can be expressed as '??*'. */ |
271 | for(;;) { |
272 | if(*++p == '\0') |
273 | return CURL_FNMATCH_MATCH; |
274 | if(*p == '?') { |
275 | if(!*s++) |
276 | return CURL_FNMATCH_NOMATCH; |
277 | } |
278 | else if(*p != '*') |
279 | break; |
280 | } |
281 | /* Skip string characters until we find a match with pattern suffix. */ |
282 | for(maxstars--; *s; s++) { |
283 | if(loop(p, s, maxstars) == CURL_FNMATCH_MATCH) |
284 | return CURL_FNMATCH_MATCH; |
285 | } |
286 | return CURL_FNMATCH_NOMATCH; |
287 | case '?': |
288 | if(!*s) |
289 | return CURL_FNMATCH_NOMATCH; |
290 | s++; |
291 | p++; |
292 | break; |
293 | case '\0': |
294 | return *s? CURL_FNMATCH_NOMATCH: CURL_FNMATCH_MATCH; |
295 | case '\\': |
296 | if(p[1]) |
297 | p++; |
298 | if(*s++ != *p++) |
299 | return CURL_FNMATCH_NOMATCH; |
300 | break; |
301 | case '[': |
302 | pp = p + 1; /* Copy in case of syntax error in set. */ |
303 | if(setcharset(&pp, charset)) { |
304 | int found = FALSE; |
305 | if(!*s) |
306 | return CURL_FNMATCH_NOMATCH; |
307 | if(charset[(unsigned int)*s]) |
308 | found = TRUE; |
309 | else if(charset[CURLFNM_ALNUM]) |
310 | found = ISALNUM(*s); |
311 | else if(charset[CURLFNM_ALPHA]) |
312 | found = ISALPHA(*s); |
313 | else if(charset[CURLFNM_DIGIT]) |
314 | found = ISDIGIT(*s); |
315 | else if(charset[CURLFNM_XDIGIT]) |
316 | found = ISXDIGIT(*s); |
317 | else if(charset[CURLFNM_PRINT]) |
318 | found = ISPRINT(*s); |
319 | else if(charset[CURLFNM_SPACE]) |
320 | found = ISSPACE(*s); |
321 | else if(charset[CURLFNM_UPPER]) |
322 | found = ISUPPER(*s); |
323 | else if(charset[CURLFNM_LOWER]) |
324 | found = ISLOWER(*s); |
325 | else if(charset[CURLFNM_BLANK]) |
326 | found = ISBLANK(*s); |
327 | else if(charset[CURLFNM_GRAPH]) |
328 | found = ISGRAPH(*s); |
329 | |
330 | if(charset[CURLFNM_NEGATE]) |
331 | found = !found; |
332 | |
333 | if(!found) |
334 | return CURL_FNMATCH_NOMATCH; |
335 | p = pp + 1; |
336 | s++; |
337 | break; |
338 | } |
339 | /* Syntax error in set; mismatch! */ |
340 | return CURL_FNMATCH_NOMATCH; |
341 | |
342 | default: |
343 | if(*p++ != *s++) |
344 | return CURL_FNMATCH_NOMATCH; |
345 | break; |
346 | } |
347 | } |
348 | } |
349 | |
350 | /* |
351 | * @unittest: 1307 |
352 | */ |
353 | int Curl_fnmatch(void *ptr, const char *pattern, const char *string) |
354 | { |
355 | (void)ptr; /* the argument is specified by the curl_fnmatch_callback |
356 | prototype, but not used by Curl_fnmatch() */ |
357 | if(!pattern || !string) { |
358 | return CURL_FNMATCH_FAIL; |
359 | } |
360 | return loop((unsigned char *)pattern, (unsigned char *)string, 2); |
361 | } |
362 | #else |
363 | #include <fnmatch.h> |
364 | /* |
365 | * @unittest: 1307 |
366 | */ |
367 | int Curl_fnmatch(void *ptr, const char *pattern, const char *string) |
368 | { |
369 | int rc; |
370 | (void)ptr; /* the argument is specified by the curl_fnmatch_callback |
371 | prototype, but not used by Curl_fnmatch() */ |
372 | if(!pattern || !string) { |
373 | return CURL_FNMATCH_FAIL; |
374 | } |
375 | rc = fnmatch(pattern, string, 0); |
376 | switch(rc) { |
377 | case 0: |
378 | return CURL_FNMATCH_MATCH; |
379 | case FNM_NOMATCH: |
380 | return CURL_FNMATCH_NOMATCH; |
381 | default: |
382 | return CURL_FNMATCH_FAIL; |
383 | } |
384 | /* not reached */ |
385 | } |
386 | |
387 | #endif |
388 | |
389 | #endif /* if FTP is disabled */ |
390 | |