BadVPN – Blame information for rev 1
?pathlinks?
Rev | Author | Line No. | Line |
---|---|---|---|
1 | office | 1 | /** |
2 | * @file regex_match.c |
||
3 | * @author Ambroz Bizjak <ambrop7@gmail.com> |
||
4 | * |
||
5 | * @section LICENSE |
||
6 | * |
||
7 | * Redistribution and use in source and binary forms, with or without |
||
8 | * modification, are permitted provided that the following conditions are met: |
||
9 | * 1. Redistributions of source code must retain the above copyright |
||
10 | * notice, this list of conditions and the following disclaimer. |
||
11 | * 2. Redistributions in binary form must reproduce the above copyright |
||
12 | * notice, this list of conditions and the following disclaimer in the |
||
13 | * documentation and/or other materials provided with the distribution. |
||
14 | * 3. Neither the name of the author nor the |
||
15 | * names of its contributors may be used to endorse or promote products |
||
16 | * derived from this software without specific prior written permission. |
||
17 | * |
||
18 | * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND |
||
19 | * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED |
||
20 | * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE |
||
21 | * DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY |
||
22 | * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES |
||
23 | * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; |
||
24 | * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND |
||
25 | * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
||
26 | * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS |
||
27 | * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
||
28 | * |
||
29 | * @section DESCRIPTION |
||
30 | * |
||
31 | * Regular expression matching module. |
||
32 | * |
||
33 | * Synopsis: |
||
34 | * regex_match(string input, string regex) |
||
35 | * |
||
36 | * Variables: |
||
37 | * succeeded - "true" or "false", indicating whether input matched regex |
||
38 | * matchN - for N=0,1,2,..., the matching data for the N-th subexpression |
||
39 | * (match0 = whole match) |
||
40 | * |
||
41 | * Description: |
||
42 | * Matches 'input' with the POSIX extended regular expression 'regex'. |
||
43 | * 'regex' must be a string without null bytes, but 'input' can contain null bytes. |
||
44 | * However, it's difficult, if not impossible, to actually match nulls with the regular |
||
45 | * expression. |
||
46 | * The input and regex strings are interpreted according to the POSIX regex functions |
||
47 | * (regcomp(), regexec()); in particular, the current locale setting affects the |
||
48 | * interpretation. |
||
49 | * |
||
50 | * Synopsis: |
||
51 | * regex_replace(string input, list(string) regex, list(string) replace) |
||
52 | * |
||
53 | * Variables: |
||
54 | * string (empty) - transformed input |
||
55 | * |
||
56 | * Description: |
||
57 | * Replaces matching parts of a string. Replacement is performed by repetedly matching |
||
58 | * the remaining part of the string with all regular expressions. On each step, out of |
||
59 | * all regular expressions that match the remainder of the string, the one whose match |
||
60 | * starts at the least position wins, and the matching part is replaced with the |
||
61 | * replacement string corresponding to this regular expression. The process continues |
||
62 | * from the end of the just-replaced portion until no more regular expressions match. |
||
63 | * If multiple regular expressions match at the least position, the one that appears |
||
64 | * first in the 'regex' argument wins. |
||
65 | */ |
||
66 | |||
67 | #include <stdlib.h> |
||
68 | #include <string.h> |
||
69 | #include <limits.h> |
||
70 | #include <regex.h> |
||
71 | |||
72 | #include <misc/string_begins_with.h> |
||
73 | #include <misc/parse_number.h> |
||
74 | #include <misc/expstring.h> |
||
75 | #include <misc/debug.h> |
||
76 | #include <misc/balloc.h> |
||
77 | |||
78 | #include <ncd/module_common.h> |
||
79 | |||
80 | #include <generated/blog_channel_ncd_regex_match.h> |
||
81 | |||
82 | #define MAX_MATCHES 64 |
||
83 | |||
84 | struct instance { |
||
85 | NCDModuleInst *i; |
||
86 | MemRef input; |
||
87 | int succeeded; |
||
88 | int num_matches; |
||
89 | regmatch_t matches[MAX_MATCHES]; |
||
90 | }; |
||
91 | |||
92 | struct replace_instance { |
||
93 | NCDModuleInst *i; |
||
94 | MemRef output; |
||
95 | }; |
||
96 | |||
97 | static void func_new (void *vo, NCDModuleInst *i, const struct NCDModuleInst_new_params *params) |
||
98 | { |
||
99 | struct instance *o = vo; |
||
100 | o->i = i; |
||
101 | |||
102 | // read arguments |
||
103 | NCDValRef input_arg; |
||
104 | NCDValRef regex_arg; |
||
105 | if (!NCDVal_ListRead(params->args, 2, &input_arg, ®ex_arg)) { |
||
106 | ModuleLog(o->i, BLOG_ERROR, "wrong arity"); |
||
107 | goto fail0; |
||
108 | } |
||
109 | if (!NCDVal_IsString(input_arg) || !NCDVal_IsStringNoNulls(regex_arg)) { |
||
110 | ModuleLog(o->i, BLOG_ERROR, "wrong type"); |
||
111 | goto fail0; |
||
112 | } |
||
113 | o->input = NCDVal_StringMemRef(input_arg); |
||
114 | |||
115 | // make sure we don't overflow regoff_t |
||
116 | if (o->input.len > INT_MAX) { |
||
117 | ModuleLog(o->i, BLOG_ERROR, "input string too long"); |
||
118 | goto fail0; |
||
119 | } |
||
120 | |||
121 | // null terminate regex |
||
122 | NCDValNullTermString regex_nts; |
||
123 | if (!NCDVal_StringNullTerminate(regex_arg, ®ex_nts)) { |
||
124 | ModuleLog(i, BLOG_ERROR, "NCDVal_StringNullTerminate failed"); |
||
125 | goto fail0; |
||
126 | } |
||
127 | |||
128 | // compile regex |
||
129 | regex_t preg; |
||
130 | int ret = regcomp(&preg, regex_nts.data, REG_EXTENDED); |
||
131 | NCDValNullTermString_Free(®ex_nts); |
||
132 | if (ret != 0) { |
||
133 | ModuleLog(o->i, BLOG_ERROR, "regcomp failed (error=%d)", ret); |
||
134 | goto fail0; |
||
135 | } |
||
136 | |||
137 | // execute match |
||
138 | o->matches[0].rm_so = 0; |
||
139 | o->matches[0].rm_eo = o->input.len; |
||
140 | o->succeeded = (regexec(&preg, o->input.ptr, MAX_MATCHES, o->matches, REG_STARTEND) == 0); |
||
141 | |||
142 | // free regex |
||
143 | regfree(&preg); |
||
144 | |||
145 | // signal up |
||
146 | NCDModuleInst_Backend_Up(o->i); |
||
147 | return; |
||
148 | |||
149 | fail0: |
||
150 | NCDModuleInst_Backend_DeadError(i); |
||
151 | } |
||
152 | |||
153 | static int func_getvar (void *vo, const char *name, NCDValMem *mem, NCDValRef *out) |
||
154 | { |
||
155 | struct instance *o = vo; |
||
156 | |||
157 | if (!strcmp(name, "succeeded")) { |
||
158 | *out = ncd_make_boolean(mem, o->succeeded); |
||
159 | return 1; |
||
160 | } |
||
161 | |||
162 | size_t pos; |
||
163 | uintmax_t n; |
||
164 | if ((pos = string_begins_with(name, "match")) && parse_unsigned_integer(MemRef_MakeCstr(name + pos), &n)) { |
||
165 | if (o->succeeded && n < MAX_MATCHES && o->matches[n].rm_so >= 0) { |
||
166 | regmatch_t *m = &o->matches[n]; |
||
167 | |||
168 | ASSERT(m->rm_so <= o->input.len) |
||
169 | ASSERT(m->rm_eo >= m->rm_so) |
||
170 | ASSERT(m->rm_eo <= o->input.len) |
||
171 | |||
172 | size_t len = m->rm_eo - m->rm_so; |
||
173 | |||
174 | *out = NCDVal_NewStringBinMr(mem, MemRef_Sub(o->input, m->rm_so, len)); |
||
175 | return 1; |
||
176 | } |
||
177 | } |
||
178 | |||
179 | return 0; |
||
180 | } |
||
181 | |||
182 | static void replace_func_new (void *vo, NCDModuleInst *i, const struct NCDModuleInst_new_params *params) |
||
183 | { |
||
184 | struct replace_instance *o = vo; |
||
185 | o->i = i; |
||
186 | |||
187 | // read arguments |
||
188 | NCDValRef input_arg; |
||
189 | NCDValRef regex_arg; |
||
190 | NCDValRef replace_arg; |
||
191 | if (!NCDVal_ListRead(params->args, 3, &input_arg, ®ex_arg, &replace_arg)) { |
||
192 | ModuleLog(i, BLOG_ERROR, "wrong arity"); |
||
193 | goto fail1; |
||
194 | } |
||
195 | if (!NCDVal_IsString(input_arg) || !NCDVal_IsList(regex_arg) || !NCDVal_IsList(replace_arg)) { |
||
196 | ModuleLog(i, BLOG_ERROR, "wrong type"); |
||
197 | goto fail1; |
||
198 | } |
||
199 | |||
200 | // check number of regex/replace |
||
201 | if (NCDVal_ListCount(regex_arg) != NCDVal_ListCount(replace_arg)) { |
||
202 | ModuleLog(i, BLOG_ERROR, "number of regex's is not the same as number of replacements"); |
||
203 | goto fail1; |
||
204 | } |
||
205 | size_t num_regex = NCDVal_ListCount(regex_arg); |
||
206 | |||
207 | // allocate array for compiled regex's |
||
208 | regex_t *regs = BAllocArray(num_regex, sizeof(regs[0])); |
||
209 | if (!regs) { |
||
210 | ModuleLog(i, BLOG_ERROR, "BAllocArray failed"); |
||
211 | goto fail1; |
||
212 | } |
||
213 | size_t num_done_regex = 0; |
||
214 | |||
215 | // compile regex's, check arguments |
||
216 | while (num_done_regex < num_regex) { |
||
217 | NCDValRef regex = NCDVal_ListGet(regex_arg, num_done_regex); |
||
218 | NCDValRef replace = NCDVal_ListGet(replace_arg, num_done_regex); |
||
219 | |||
220 | if (!NCDVal_IsStringNoNulls(regex) || !NCDVal_IsString(replace)) { |
||
221 | ModuleLog(i, BLOG_ERROR, "wrong regex/replace type for pair %zu", num_done_regex); |
||
222 | goto fail2; |
||
223 | } |
||
224 | |||
225 | // null terminate regex |
||
226 | NCDValNullTermString regex_nts; |
||
227 | if (!NCDVal_StringNullTerminate(regex, ®ex_nts)) { |
||
228 | ModuleLog(i, BLOG_ERROR, "NCDVal_StringNullTerminate failed"); |
||
229 | goto fail2; |
||
230 | } |
||
231 | |||
232 | int res = regcomp(®s[num_done_regex], regex_nts.data, REG_EXTENDED); |
||
233 | NCDValNullTermString_Free(®ex_nts); |
||
234 | if (res != 0) { |
||
235 | ModuleLog(i, BLOG_ERROR, "regcomp failed for pair %zu (error=%d)", num_done_regex, res); |
||
236 | goto fail2; |
||
237 | } |
||
238 | |||
239 | num_done_regex++; |
||
240 | } |
||
241 | |||
242 | // init output string |
||
243 | ExpString out; |
||
244 | if (!ExpString_Init(&out)) { |
||
245 | ModuleLog(i, BLOG_ERROR, "ExpString_Init failed"); |
||
246 | goto fail2; |
||
247 | } |
||
248 | |||
249 | // input state |
||
250 | MemRef in = NCDVal_StringMemRef(input_arg); |
||
251 | size_t in_pos = 0; |
||
252 | |||
253 | // process input |
||
254 | while (in_pos < in.len) { |
||
255 | // find first match |
||
256 | int have_match = 0; |
||
257 | size_t match_regex = 0; // to remove warning |
||
258 | regmatch_t match = {0, 0}; // to remove warning |
||
259 | for (size_t j = 0; j < num_regex; j++) { |
||
260 | regmatch_t this_match; |
||
261 | this_match.rm_so = 0; |
||
262 | this_match.rm_eo = in.len - in_pos; |
||
263 | if (regexec(®s[j], in.ptr + in_pos, 1, &this_match, REG_STARTEND) == 0 && (!have_match || this_match.rm_so < match.rm_so)) { |
||
264 | have_match = 1; |
||
265 | match_regex = j; |
||
266 | match = this_match; |
||
267 | } |
||
268 | } |
||
269 | |||
270 | // if no match, append remaining data and finish |
||
271 | if (!have_match) { |
||
272 | if (!ExpString_AppendBinaryMr(&out, MemRef_SubFrom(in, in_pos))) { |
||
273 | ModuleLog(i, BLOG_ERROR, "ExpString_AppendBinaryMr failed"); |
||
274 | goto fail3; |
||
275 | } |
||
276 | break; |
||
277 | } |
||
278 | |||
279 | // append data before match |
||
280 | if (!ExpString_AppendBinaryMr(&out, MemRef_Sub(in, in_pos, match.rm_so))) { |
||
281 | ModuleLog(i, BLOG_ERROR, "ExpString_AppendBinaryMr failed"); |
||
282 | goto fail3; |
||
283 | } |
||
284 | |||
285 | // append replacement data |
||
286 | NCDValRef replace = NCDVal_ListGet(replace_arg, match_regex); |
||
287 | if (!ExpString_AppendBinaryMr(&out, NCDVal_StringMemRef(replace))) { |
||
288 | ModuleLog(i, BLOG_ERROR, "ExpString_AppendBinaryMr failed"); |
||
289 | goto fail3; |
||
290 | } |
||
291 | |||
292 | in_pos += match.rm_eo; |
||
293 | } |
||
294 | |||
295 | // set output |
||
296 | o->output = ExpString_GetMr(&out); |
||
297 | |||
298 | // free compiled regex's |
||
299 | while (num_done_regex-- > 0) { |
||
300 | regfree(®s[num_done_regex]); |
||
301 | } |
||
302 | |||
303 | // free array |
||
304 | BFree(regs); |
||
305 | |||
306 | // signal up |
||
307 | NCDModuleInst_Backend_Up(i); |
||
308 | return; |
||
309 | |||
310 | fail3: |
||
311 | ExpString_Free(&out); |
||
312 | fail2: |
||
313 | while (num_done_regex-- > 0) { |
||
314 | regfree(®s[num_done_regex]); |
||
315 | } |
||
316 | BFree(regs); |
||
317 | fail1: |
||
318 | NCDModuleInst_Backend_DeadError(i); |
||
319 | } |
||
320 | |||
321 | static void replace_func_die (void *vo) |
||
322 | { |
||
323 | struct replace_instance *o = vo; |
||
324 | |||
325 | // free output |
||
326 | BFree((char *)o->output.ptr); |
||
327 | |||
328 | NCDModuleInst_Backend_Dead(o->i); |
||
329 | } |
||
330 | |||
331 | static int replace_func_getvar (void *vo, const char *name, NCDValMem *mem, NCDValRef *out) |
||
332 | { |
||
333 | struct replace_instance *o = vo; |
||
334 | |||
335 | if (!strcmp(name, "")) { |
||
336 | *out = NCDVal_NewStringBinMr(mem, o->output); |
||
337 | return 1; |
||
338 | } |
||
339 | |||
340 | return 0; |
||
341 | } |
||
342 | |||
343 | static struct NCDModule modules[] = { |
||
344 | { |
||
345 | .type = "regex_match", |
||
346 | .func_new2 = func_new, |
||
347 | .func_getvar = func_getvar, |
||
348 | .alloc_size = sizeof(struct instance) |
||
349 | }, { |
||
350 | .type = "regex_replace", |
||
351 | .func_new2 = replace_func_new, |
||
352 | .func_die = replace_func_die, |
||
353 | .func_getvar = replace_func_getvar, |
||
354 | .alloc_size = sizeof(struct replace_instance) |
||
355 | }, { |
||
356 | .type = NULL |
||
357 | } |
||
358 | }; |
||
359 | |||
360 | const struct NCDModuleGroup ncdmodule_regex_match = { |
||
361 | .modules = modules |
||
362 | }; |