| 1 | /* ============================================================= */ |
|---|
| 2 | /* */ |
|---|
| 3 | /* File : admatch.c */ |
|---|
| 4 | /* Purpose : functions related to string match/replace */ |
|---|
| 5 | /* */ |
|---|
| 6 | /* ReCoded for POSIX ERE by Ralf Westram (coder@reallysoft.de) */ |
|---|
| 7 | /* in April 2009 */ |
|---|
| 8 | /* Institute of Microbiology (Technical University Munich) */ |
|---|
| 9 | /* http://www.arb-home.de/ */ |
|---|
| 10 | /* */ |
|---|
| 11 | /* ============================================================= */ |
|---|
| 12 | |
|---|
| 13 | #include "adlocal.h" |
|---|
| 14 | |
|---|
| 15 | #include <stdlib.h> |
|---|
| 16 | #include <string.h> |
|---|
| 17 | #include <ctype.h> |
|---|
| 18 | #include <regex.h> |
|---|
| 19 | |
|---|
| 20 | /* ------------------------ */ |
|---|
| 21 | /* string matcher */ |
|---|
| 22 | |
|---|
| 23 | struct GBS_string_matcher { |
|---|
| 24 | enum { |
|---|
| 25 | SM_ANY, // matches any string |
|---|
| 26 | SM_WILDCARDED, // match with wildcards (GBS_string_matches) |
|---|
| 27 | SM_REGEXPR, // match using regexpr |
|---|
| 28 | } type; |
|---|
| 29 | |
|---|
| 30 | GB_CASE case_flag; |
|---|
| 31 | char *wildexpr; |
|---|
| 32 | GBS_REGEX *regexpr; |
|---|
| 33 | }; |
|---|
| 34 | |
|---|
| 35 | GBS_MATCHER *GBS_compile_matcher(const char *search_expr, GB_CASE case_flag) { |
|---|
| 36 | /* returns a valid string matcher (to be used with GBS_string_matches_regexp) |
|---|
| 37 | * or NULL (in which case an error was exported) |
|---|
| 38 | */ |
|---|
| 39 | |
|---|
| 40 | GBS_MATCHER *matcher = malloc(sizeof(*matcher)); |
|---|
| 41 | GB_ERROR error = 0; |
|---|
| 42 | |
|---|
| 43 | matcher->type = -1; |
|---|
| 44 | matcher->case_flag = case_flag; |
|---|
| 45 | matcher->wildexpr = NULL; |
|---|
| 46 | matcher->regexpr = NULL; |
|---|
| 47 | |
|---|
| 48 | if (search_expr[0] == '/') { |
|---|
| 49 | const char *end = strchr(search_expr, 0)-1; |
|---|
| 50 | if (end>search_expr && end[0] == '/') { |
|---|
| 51 | GB_CASE expr_attached_case; |
|---|
| 52 | const char *unwrapped_expr = GBS_unwrap_regexpr(search_expr, &expr_attached_case, &error); |
|---|
| 53 | |
|---|
| 54 | if (unwrapped_expr) { |
|---|
| 55 | if (expr_attached_case != GB_MIND_CASE) error = "format '/../i' not allowed here"; |
|---|
| 56 | else { |
|---|
| 57 | matcher->regexpr = GBS_compile_regexpr(unwrapped_expr, case_flag, &error); |
|---|
| 58 | if (matcher->regexpr) { |
|---|
| 59 | matcher->type = SM_REGEXPR;; |
|---|
| 60 | } |
|---|
| 61 | } |
|---|
| 62 | } |
|---|
| 63 | } |
|---|
| 64 | } |
|---|
| 65 | |
|---|
| 66 | if (matcher->regexpr == NULL && !error) { |
|---|
| 67 | if (strcmp(search_expr, "*") == 0) { |
|---|
| 68 | matcher->type = SM_ANY; |
|---|
| 69 | } |
|---|
| 70 | else { |
|---|
| 71 | matcher->type = SM_WILDCARDED; |
|---|
| 72 | matcher->wildexpr = strdup(search_expr); |
|---|
| 73 | } |
|---|
| 74 | } |
|---|
| 75 | |
|---|
| 76 | if (error) { |
|---|
| 77 | GBS_free_matcher(matcher); |
|---|
| 78 | matcher = 0; |
|---|
| 79 | GB_export_error(error); |
|---|
| 80 | } |
|---|
| 81 | return matcher; |
|---|
| 82 | } |
|---|
| 83 | |
|---|
| 84 | void GBS_free_matcher(GBS_MATCHER *matcher) { |
|---|
| 85 | free(matcher->wildexpr); |
|---|
| 86 | if (matcher->regexpr) GBS_free_regexpr(matcher->regexpr); |
|---|
| 87 | free(matcher); |
|---|
| 88 | } |
|---|
| 89 | |
|---|
| 90 | /* --------------------------------------------- */ |
|---|
| 91 | /* Regular Expressions search/replace */ |
|---|
| 92 | |
|---|
| 93 | struct GBS_regex { regex_t compiled; }; // definition exists twice (see ../SL/REGEXPR/RegExpr.cxx) |
|---|
| 94 | |
|---|
| 95 | GBS_REGEX *GBS_compile_regexpr(const char *regexpr, GB_CASE case_flag, GB_ERROR *error) { |
|---|
| 96 | GBS_REGEX *comreg = malloc(sizeof(*comreg)); |
|---|
| 97 | int cflags = REG_EXTENDED|(case_flag == GB_IGNORE_CASE ? REG_ICASE : 0)|REG_NEWLINE; |
|---|
| 98 | int errcode = regcomp(&comreg->compiled, regexpr, cflags); |
|---|
| 99 | |
|---|
| 100 | if (errcode != 0) { // error compiling regexpr |
|---|
| 101 | size_t size = regerror(errcode, &comreg->compiled, NULL, 0); |
|---|
| 102 | GB_BUFFER buf = GB_give_buffer(size); |
|---|
| 103 | |
|---|
| 104 | regerror(errcode, &comreg->compiled, buf, size); |
|---|
| 105 | *error = buf; |
|---|
| 106 | |
|---|
| 107 | free(comreg); |
|---|
| 108 | comreg = NULL; |
|---|
| 109 | } |
|---|
| 110 | else { |
|---|
| 111 | *error = NULL; |
|---|
| 112 | } |
|---|
| 113 | |
|---|
| 114 | return comreg; |
|---|
| 115 | } |
|---|
| 116 | |
|---|
| 117 | void GBS_free_regexpr(GBS_REGEX *toFree) { |
|---|
| 118 | if (toFree) { |
|---|
| 119 | regfree(&toFree->compiled); |
|---|
| 120 | free(toFree); |
|---|
| 121 | } |
|---|
| 122 | } |
|---|
| 123 | |
|---|
| 124 | const char *GBS_unwrap_regexpr(const char *regexpr_in_slashes, GB_CASE *case_flag, GB_ERROR *error) { |
|---|
| 125 | /* unwraps 'expr' from '/expr/[i]' |
|---|
| 126 | * if slashes are not present, 'error' is set |
|---|
| 127 | * 'case_flag' is set to GB_MIND_CASE if format is '/expr/' or |
|---|
| 128 | * to GB_IGNORE_CASE if format is '/expr/i' |
|---|
| 129 | * |
|---|
| 130 | * returns a pointer to a static buffer (containing the unwrapped expression) |
|---|
| 131 | * (Note: The content is invalidated by the next call to GBS_unwrap_regexpr) |
|---|
| 132 | */ |
|---|
| 133 | |
|---|
| 134 | const char *result = NULL; |
|---|
| 135 | const char *end = strchr(regexpr_in_slashes, 0); |
|---|
| 136 | |
|---|
| 137 | if (end >= (regexpr_in_slashes+3)) { |
|---|
| 138 | *case_flag = GB_MIND_CASE; |
|---|
| 139 | if (end[-1] == 'i') { |
|---|
| 140 | *case_flag = GB_IGNORE_CASE; |
|---|
| 141 | end--; |
|---|
| 142 | } |
|---|
| 143 | if (regexpr_in_slashes[0] == '/' && end[-1] == '/') { |
|---|
| 144 | ad_assert(*error == NULL); |
|---|
| 145 | |
|---|
| 146 | static char *result_buffer = 0; |
|---|
| 147 | static size_t max_len = 0; |
|---|
| 148 | |
|---|
| 149 | size_t len = end-regexpr_in_slashes-2; |
|---|
| 150 | ad_assert(len>0); // don't accept empty expression |
|---|
| 151 | |
|---|
| 152 | if (len>max_len) { |
|---|
| 153 | max_len = len*3/2; |
|---|
| 154 | freeset(result_buffer, malloc(max_len+1)); |
|---|
| 155 | } |
|---|
| 156 | |
|---|
| 157 | memcpy(result_buffer, regexpr_in_slashes+1, len); |
|---|
| 158 | result_buffer[len] = 0; |
|---|
| 159 | |
|---|
| 160 | result = result_buffer; |
|---|
| 161 | } |
|---|
| 162 | } |
|---|
| 163 | |
|---|
| 164 | if (!result) { |
|---|
| 165 | *error = GBS_global_string("Regular expression format is '/expr/' or '/expr/i', not '%s'", |
|---|
| 166 | regexpr_in_slashes); |
|---|
| 167 | } |
|---|
| 168 | return result; |
|---|
| 169 | } |
|---|
| 170 | |
|---|
| 171 | const char *GBS_regmatch_compiled(const char *str, GBS_REGEX *comreg, size_t *matchlen) { |
|---|
| 172 | /* like GBS_regmatch, |
|---|
| 173 | * - but uses a precompiled regular expression |
|---|
| 174 | * - no errors can occur here (beside out of memory, which is not handled) |
|---|
| 175 | */ |
|---|
| 176 | |
|---|
| 177 | regmatch_t match; |
|---|
| 178 | int res = regexec(&comreg->compiled, str, 1, &match, 0); |
|---|
| 179 | const char *matchpos = NULL; |
|---|
| 180 | |
|---|
| 181 | if (res == 0) { // matched |
|---|
| 182 | matchpos = str+match.rm_so; |
|---|
| 183 | if (matchlen) *matchlen = match.rm_eo-match.rm_so; |
|---|
| 184 | } |
|---|
| 185 | |
|---|
| 186 | return matchpos; |
|---|
| 187 | } |
|---|
| 188 | |
|---|
| 189 | const char *GBS_regmatch(const char *str, const char *regExpr, size_t *matchlen, GB_ERROR *error) { |
|---|
| 190 | /* searches 'str' for first occurrence of 'regExpr' |
|---|
| 191 | * 'regExpr' has to be in format "/expr/[i]", where 'expr' is a POSIX extended regular expression |
|---|
| 192 | * |
|---|
| 193 | * returns |
|---|
| 194 | * - pointer to start of first match in 'str' and |
|---|
| 195 | * length of match in 'matchlen' ('matchlen' may be NULL, then no len is reported) |
|---|
| 196 | * or |
|---|
| 197 | * - NULL if nothing matched (in this case 'matchlen' is undefined) |
|---|
| 198 | * |
|---|
| 199 | * 'error' will be set if sth is wrong |
|---|
| 200 | * |
|---|
| 201 | * Note: Only use this function if you do exactly ONE match. |
|---|
| 202 | * Use GBS_regmatch_compiled if you use the regexpr twice or more! |
|---|
| 203 | */ |
|---|
| 204 | const char *firstMatch = NULL; |
|---|
| 205 | GB_CASE case_flag; |
|---|
| 206 | const char *unwrapped_expr = GBS_unwrap_regexpr(regExpr, &case_flag, error); |
|---|
| 207 | |
|---|
| 208 | if (unwrapped_expr) { |
|---|
| 209 | GBS_REGEX *comreg = GBS_compile_regexpr(unwrapped_expr, case_flag, error); |
|---|
| 210 | if (comreg) { |
|---|
| 211 | firstMatch = GBS_regmatch_compiled(str, comreg, matchlen); |
|---|
| 212 | GBS_free_regexpr(comreg); |
|---|
| 213 | } |
|---|
| 214 | } |
|---|
| 215 | |
|---|
| 216 | return firstMatch; |
|---|
| 217 | } |
|---|
| 218 | |
|---|
| 219 | char *GBS_regreplace(const char *str, const char *regReplExpr, GB_ERROR *error) { |
|---|
| 220 | /* search and replace all matches in 'str' using POSIX extended regular expression |
|---|
| 221 | * 'regReplExpr' has to be in format '/regexpr/replace/[i]' |
|---|
| 222 | * |
|---|
| 223 | * returns |
|---|
| 224 | * - a heap copy of the modified string or |
|---|
| 225 | * - NULL if something went wrong (in this case 'error' contains the reason) |
|---|
| 226 | * |
|---|
| 227 | * 'replace' may contain several special substrings: |
|---|
| 228 | * |
|---|
| 229 | * "\n" gets replaced by '\n' |
|---|
| 230 | * "\t" -------''------- '\t' |
|---|
| 231 | * "\\" -------''------- '\\' |
|---|
| 232 | * "\0" -------''------- the complete match to regexpr |
|---|
| 233 | * "\1" -------''------- the match to the first subexpression |
|---|
| 234 | * "\2" -------''------- the match to the second subexpression |
|---|
| 235 | * ... |
|---|
| 236 | * "\9" -------''------- the match to the ninth subexpression |
|---|
| 237 | */ |
|---|
| 238 | |
|---|
| 239 | GB_CASE case_flag; |
|---|
| 240 | const char *unwrapped_expr = GBS_unwrap_regexpr(regReplExpr, &case_flag, error); |
|---|
| 241 | char *result = NULL; |
|---|
| 242 | |
|---|
| 243 | if (unwrapped_expr) { |
|---|
| 244 | const char *sep = unwrapped_expr; |
|---|
| 245 | while (sep) { |
|---|
| 246 | sep = strchr(sep, '/'); |
|---|
| 247 | if (!sep) break; |
|---|
| 248 | if (sep>unwrapped_expr && sep[-1] != '\\') break; |
|---|
| 249 | } |
|---|
| 250 | |
|---|
| 251 | if (!sep) { |
|---|
| 252 | // Warning: GB_command_interpreter() tests for this error message - don't change |
|---|
| 253 | *error = "Missing '/' between search and replace string"; |
|---|
| 254 | } |
|---|
| 255 | else { |
|---|
| 256 | char *regexpr = GB_strpartdup(unwrapped_expr, sep-1); |
|---|
| 257 | char *replexpr = GB_strpartdup(sep+1, NULL); |
|---|
| 258 | GBS_REGEX *comreg = GBS_compile_regexpr(regexpr, case_flag, error); |
|---|
| 259 | |
|---|
| 260 | if (comreg) { |
|---|
| 261 | struct GBS_strstruct *out = GBS_stropen(1000); |
|---|
| 262 | int eflags = 0; |
|---|
| 263 | |
|---|
| 264 | while (str) { |
|---|
| 265 | regmatch_t match[10]; |
|---|
| 266 | int res = regexec(&comreg->compiled, str, 10, match, eflags); |
|---|
| 267 | |
|---|
| 268 | if (res == REG_NOMATCH) { |
|---|
| 269 | GBS_strcat(out, str); |
|---|
| 270 | str = 0; |
|---|
| 271 | } |
|---|
| 272 | else { // found match |
|---|
| 273 | size_t p; |
|---|
| 274 | char c; |
|---|
| 275 | |
|---|
| 276 | GBS_strncat(out, str, match[0].rm_so); |
|---|
| 277 | |
|---|
| 278 | for (p = 0; (c = replexpr[p]); ++p) { |
|---|
| 279 | if (c == '\\') { |
|---|
| 280 | c = replexpr[++p]; |
|---|
| 281 | if (!c) break; |
|---|
| 282 | if (c >= '0' && c <= '9') { |
|---|
| 283 | regoff_t start = match[c-'0'].rm_so; |
|---|
| 284 | GBS_strncat(out, str+start, match[c-'0'].rm_eo-start); |
|---|
| 285 | } |
|---|
| 286 | else { |
|---|
| 287 | switch (c) { |
|---|
| 288 | case 'n': c = '\n'; break; |
|---|
| 289 | case 't': c = '\t'; break; |
|---|
| 290 | default: break; |
|---|
| 291 | } |
|---|
| 292 | GBS_chrcat(out, c); |
|---|
| 293 | } |
|---|
| 294 | } |
|---|
| 295 | else { |
|---|
| 296 | GBS_chrcat(out, c); |
|---|
| 297 | } |
|---|
| 298 | } |
|---|
| 299 | |
|---|
| 300 | str = str+match[0].rm_eo; // continue behind match |
|---|
| 301 | eflags = REG_NOTBOL; // for futher matches, do not regard 'str' as "beginning of line" |
|---|
| 302 | } |
|---|
| 303 | } |
|---|
| 304 | |
|---|
| 305 | GBS_free_regexpr(comreg); |
|---|
| 306 | result = GBS_strclose(out); |
|---|
| 307 | } |
|---|
| 308 | free(replexpr); |
|---|
| 309 | free(regexpr); |
|---|
| 310 | } |
|---|
| 311 | } |
|---|
| 312 | |
|---|
| 313 | return result; |
|---|
| 314 | } |
|---|
| 315 | |
|---|
| 316 | /* ------------------------- */ |
|---|
| 317 | /* wildcard search */ |
|---|
| 318 | |
|---|
| 319 | GB_CSTR GBS_find_string(GB_CSTR str, GB_CSTR substr, int match_mode) { |
|---|
| 320 | /* search a substring in another string |
|---|
| 321 | match_mode == 0 -> exact match |
|---|
| 322 | match_mode == 1 -> a==A |
|---|
| 323 | match_mode == 2 -> a==a && a==? |
|---|
| 324 | match_mode == else -> a==A && a==? |
|---|
| 325 | */ |
|---|
| 326 | const char *p1, *p2; |
|---|
| 327 | char b; |
|---|
| 328 | |
|---|
| 329 | switch (match_mode) { |
|---|
| 330 | |
|---|
| 331 | case 0: /* exact match */ |
|---|
| 332 | for (p1 = str, p2 = substr; *p1;) { |
|---|
| 333 | if (!(b = *p2)) { |
|---|
| 334 | return (char *)str; |
|---|
| 335 | } else { |
|---|
| 336 | if (b == *p1) { |
|---|
| 337 | p1++; |
|---|
| 338 | p2++; |
|---|
| 339 | } else { |
|---|
| 340 | p2 = substr; |
|---|
| 341 | p1 = (++str); |
|---|
| 342 | } |
|---|
| 343 | } |
|---|
| 344 | } |
|---|
| 345 | if (!*p2) return (char *)str; |
|---|
| 346 | break; |
|---|
| 347 | |
|---|
| 348 | case 1: /* a==A */ |
|---|
| 349 | for (p1 = str, p2 = substr; *p1;) { |
|---|
| 350 | if (!(b = *p2)) { |
|---|
| 351 | return (char *)str; |
|---|
| 352 | } else { |
|---|
| 353 | if (toupper(*p1) == toupper(b)) { |
|---|
| 354 | p1++; |
|---|
| 355 | p2++; |
|---|
| 356 | } else { |
|---|
| 357 | p2 = substr; |
|---|
| 358 | p1 = (++str); |
|---|
| 359 | } |
|---|
| 360 | } |
|---|
| 361 | } |
|---|
| 362 | if (!*p2) return (char *)str; |
|---|
| 363 | break; |
|---|
| 364 | case 2: /* a==a && a==? */ |
|---|
| 365 | for (p1 = str, p2 = substr; *p1;) { |
|---|
| 366 | if (!(b = *p2)) { |
|---|
| 367 | return (char *)str; |
|---|
| 368 | } else { |
|---|
| 369 | if (b == *p1 || (b=='?')) { |
|---|
| 370 | p1++; |
|---|
| 371 | p2++; |
|---|
| 372 | } else { |
|---|
| 373 | p2 = substr; |
|---|
| 374 | p1 = (++str); |
|---|
| 375 | } |
|---|
| 376 | } |
|---|
| 377 | } |
|---|
| 378 | if (!*p2) return (char *)str; |
|---|
| 379 | break; |
|---|
| 380 | |
|---|
| 381 | default: /* a==A && a==? */ |
|---|
| 382 | for (p1 = str, p2 = substr; *p1;) { |
|---|
| 383 | if (!(b = *p2)) { |
|---|
| 384 | return (char *)str; |
|---|
| 385 | } else { |
|---|
| 386 | if (toupper(*p1) == toupper(b) || (b=='?') ) { |
|---|
| 387 | p1++; |
|---|
| 388 | p2++; |
|---|
| 389 | } else { |
|---|
| 390 | p2 = substr; |
|---|
| 391 | p1 = (++str); |
|---|
| 392 | } |
|---|
| 393 | } |
|---|
| 394 | } |
|---|
| 395 | if (!*p2) return (char *)str; |
|---|
| 396 | break; |
|---|
| 397 | } |
|---|
| 398 | return 0; |
|---|
| 399 | } |
|---|
| 400 | |
|---|
| 401 | GB_BOOL GBS_string_matches(const char *str, const char *search, GB_CASE case_sens) |
|---|
| 402 | /* Wildcards in 'search' string: |
|---|
| 403 | * ? one character |
|---|
| 404 | * * serveral characters |
|---|
| 405 | * |
|---|
| 406 | * if 'case_sens' == GB_IGNORE_CASE -> change all letters to uppercase |
|---|
| 407 | * |
|---|
| 408 | * returns GB_TRUE if strings are equal, GB_FALSE otherwise |
|---|
| 409 | */ |
|---|
| 410 | { |
|---|
| 411 | const char *p1,*p2; |
|---|
| 412 | char a,b,*d; |
|---|
| 413 | long i; |
|---|
| 414 | char fsbuf[256]; |
|---|
| 415 | |
|---|
| 416 | p1 = str; |
|---|
| 417 | p2 = search; |
|---|
| 418 | while (1) { |
|---|
| 419 | a = *p1; |
|---|
| 420 | b = *p2; |
|---|
| 421 | if (b == '*') { |
|---|
| 422 | if (!p2[1]) break; /* '*' also matches nothing */ |
|---|
| 423 | i = 0; |
|---|
| 424 | d = fsbuf; |
|---|
| 425 | for (p2++; (b=*p2)&&(b!='*'); ) { |
|---|
| 426 | *(d++) = b; |
|---|
| 427 | p2++; |
|---|
| 428 | i++; |
|---|
| 429 | if (i > 250) break; |
|---|
| 430 | } |
|---|
| 431 | if (*p2 != '*' ) { |
|---|
| 432 | p1 += strlen(p1)-i; /* check the end of the string */ |
|---|
| 433 | if (p1 < str) return GB_FALSE; |
|---|
| 434 | p2 -= i; |
|---|
| 435 | } |
|---|
| 436 | else { |
|---|
| 437 | *d = 0; |
|---|
| 438 | p1 = GBS_find_string(p1,fsbuf,2+(case_sens == GB_IGNORE_CASE)); // match with '?' wildcard |
|---|
| 439 | if (!p1) return GB_FALSE; |
|---|
| 440 | p1 += i; |
|---|
| 441 | } |
|---|
| 442 | continue; |
|---|
| 443 | } |
|---|
| 444 | |
|---|
| 445 | if (!a) return b ? GB_FALSE : GB_TRUE; |
|---|
| 446 | if (a != b) { |
|---|
| 447 | if (b != '?') { |
|---|
| 448 | if (!b) return a ? GB_FALSE : GB_TRUE; |
|---|
| 449 | if (case_sens == GB_IGNORE_CASE) { |
|---|
| 450 | a = toupper(a); |
|---|
| 451 | b = toupper(b); |
|---|
| 452 | if (a != b) return GB_FALSE; |
|---|
| 453 | } |
|---|
| 454 | else { |
|---|
| 455 | return GB_FALSE; |
|---|
| 456 | } |
|---|
| 457 | } |
|---|
| 458 | } |
|---|
| 459 | p1++; |
|---|
| 460 | p2++; |
|---|
| 461 | } |
|---|
| 462 | return GB_TRUE; |
|---|
| 463 | } |
|---|
| 464 | |
|---|
| 465 | GB_BOOL GBS_string_matches_regexp(const char *str, const GBS_MATCHER *expr) { |
|---|
| 466 | /* Wildcard or regular expression match |
|---|
| 467 | * Returns GB_TRUE if match |
|---|
| 468 | * |
|---|
| 469 | * Use GBS_compile_matcher() and GBS_free_matcher() to maintain 'expr' |
|---|
| 470 | */ |
|---|
| 471 | GB_BOOL matches = GB_FALSE; |
|---|
| 472 | |
|---|
| 473 | switch (expr->type) { |
|---|
| 474 | case SM_ANY: { |
|---|
| 475 | matches = GB_TRUE; |
|---|
| 476 | break; |
|---|
| 477 | } |
|---|
| 478 | case SM_WILDCARDED: { |
|---|
| 479 | matches = GBS_string_matches(str, expr->wildexpr, expr->case_flag); |
|---|
| 480 | break; |
|---|
| 481 | } |
|---|
| 482 | case SM_REGEXPR: { |
|---|
| 483 | matches = GBS_regmatch_compiled(str, expr->regexpr, NULL) != NULL; |
|---|
| 484 | break; |
|---|
| 485 | } |
|---|
| 486 | default: { |
|---|
| 487 | ad_assert(0); |
|---|
| 488 | break; |
|---|
| 489 | } |
|---|
| 490 | } |
|---|
| 491 | |
|---|
| 492 | return matches; |
|---|
| 493 | } |
|---|
| 494 | |
|---|
| 495 | /* ----------------------------------- */ |
|---|
| 496 | /* Search replace tool (SRT) */ |
|---|
| 497 | |
|---|
| 498 | #define GBS_SET ((char)1) |
|---|
| 499 | #define GBS_SEP ((char)2) |
|---|
| 500 | #define GBS_MWILD ((char)3) |
|---|
| 501 | #define GBS_WILD ((char)4) |
|---|
| 502 | |
|---|
| 503 | int GBS_reference_not_found; |
|---|
| 504 | |
|---|
| 505 | ATTRIBUTED(__ATTR__USERESULT, |
|---|
| 506 | static GB_ERROR gbs_build_replace_string(void *strstruct, |
|---|
| 507 | char *bar,char *wildcards, long max_wildcard, |
|---|
| 508 | char **mwildcards, long max_mwildcard, GBDATA *gb_container)) |
|---|
| 509 | { |
|---|
| 510 | char *p,c,d; |
|---|
| 511 | int wildcardcnt = 0; |
|---|
| 512 | int mwildcardcnt = 0; |
|---|
| 513 | int wildcard_num; |
|---|
| 514 | char *entry; |
|---|
| 515 | |
|---|
| 516 | p = bar; |
|---|
| 517 | |
|---|
| 518 | wildcardcnt = 0; |
|---|
| 519 | mwildcardcnt = 0; |
|---|
| 520 | |
|---|
| 521 | p = bar; |
|---|
| 522 | while ( (c=*(p++)) ) { |
|---|
| 523 | switch (c){ |
|---|
| 524 | case GBS_MWILD: case GBS_WILD: |
|---|
| 525 | d = *(p++); |
|---|
| 526 | if (gb_container && (d=='(')) { /* if a gbcont then replace till ')' */ |
|---|
| 527 | GBDATA *gb_entry; |
|---|
| 528 | char *klz; |
|---|
| 529 | char *psym; |
|---|
| 530 | klz = gbs_search_second_bracket(p); |
|---|
| 531 | if (klz) { /* reference found: $(gbd) */ |
|---|
| 532 | int seperator = 0; |
|---|
| 533 | *klz = 0; |
|---|
| 534 | psym = strpbrk(p,"#|:"); |
|---|
| 535 | if (psym) { |
|---|
| 536 | seperator = *psym; |
|---|
| 537 | *psym =0; |
|---|
| 538 | } |
|---|
| 539 | if (*p){ |
|---|
| 540 | gb_entry = GB_search(gb_container,p,GB_FIND); |
|---|
| 541 | }else{ |
|---|
| 542 | gb_entry = gb_container; |
|---|
| 543 | } |
|---|
| 544 | if (psym) *psym = seperator; |
|---|
| 545 | |
|---|
| 546 | if (!gb_entry || gb_entry == gb_container) { |
|---|
| 547 | GBS_reference_not_found = 1; |
|---|
| 548 | entry = strdup(""); |
|---|
| 549 | }else{ |
|---|
| 550 | entry = GB_read_as_string(gb_entry); |
|---|
| 551 | } |
|---|
| 552 | if (entry) { |
|---|
| 553 | char *h; |
|---|
| 554 | switch(seperator) { |
|---|
| 555 | case ':': |
|---|
| 556 | h = GBS_string_eval(entry,psym+1,gb_container); |
|---|
| 557 | if (!h) return GB_await_error(); |
|---|
| 558 | |
|---|
| 559 | GBS_strcat(strstruct,h); |
|---|
| 560 | free(h); |
|---|
| 561 | break; |
|---|
| 562 | |
|---|
| 563 | case '|': |
|---|
| 564 | h = GB_command_interpreter(GB_get_root(gb_container), entry,psym+1,gb_container, 0); |
|---|
| 565 | if (!h) return GB_await_error(); |
|---|
| 566 | |
|---|
| 567 | GBS_strcat(strstruct,h); |
|---|
| 568 | free(h); |
|---|
| 569 | break; |
|---|
| 570 | |
|---|
| 571 | case '#': |
|---|
| 572 | if (!gb_entry){ |
|---|
| 573 | GBS_strcat(strstruct,psym+1); |
|---|
| 574 | break; |
|---|
| 575 | } |
|---|
| 576 | // fall-through |
|---|
| 577 | default: |
|---|
| 578 | GBS_strcat(strstruct,entry); |
|---|
| 579 | break; |
|---|
| 580 | } |
|---|
| 581 | free(entry); |
|---|
| 582 | } |
|---|
| 583 | *klz = ')'; |
|---|
| 584 | p = klz+1; |
|---|
| 585 | break; |
|---|
| 586 | } |
|---|
| 587 | c = '*'; |
|---|
| 588 | GBS_chrcat(strstruct,c); |
|---|
| 589 | GBS_chrcat(strstruct,d); |
|---|
| 590 | } |
|---|
| 591 | else { |
|---|
| 592 | wildcard_num = d - '1'; |
|---|
| 593 | if (c == GBS_WILD) { |
|---|
| 594 | c = '?'; |
|---|
| 595 | if ( (wildcard_num<0)||(wildcard_num>9) ) { |
|---|
| 596 | p--; /* use this character */ |
|---|
| 597 | wildcard_num = wildcardcnt++; |
|---|
| 598 | } |
|---|
| 599 | if (wildcard_num>=max_wildcard) { |
|---|
| 600 | GBS_chrcat(strstruct,c); |
|---|
| 601 | } |
|---|
| 602 | else { |
|---|
| 603 | GBS_chrcat(strstruct,wildcards[wildcard_num]); |
|---|
| 604 | } |
|---|
| 605 | } |
|---|
| 606 | else { |
|---|
| 607 | c = '*'; |
|---|
| 608 | if ( (wildcard_num<0)||(wildcard_num>9) ) { |
|---|
| 609 | p--; /* use this character */ |
|---|
| 610 | wildcard_num = mwildcardcnt++; |
|---|
| 611 | } |
|---|
| 612 | if (wildcard_num>=max_mwildcard) { |
|---|
| 613 | GBS_chrcat(strstruct,c); |
|---|
| 614 | }else{ |
|---|
| 615 | GBS_strcat(strstruct,mwildcards[wildcard_num]); |
|---|
| 616 | } |
|---|
| 617 | } |
|---|
| 618 | } |
|---|
| 619 | break; |
|---|
| 620 | default: |
|---|
| 621 | GBS_chrcat(strstruct,c); |
|---|
| 622 | break; |
|---|
| 623 | } /* switch c */ |
|---|
| 624 | } |
|---|
| 625 | return 0; |
|---|
| 626 | } |
|---|
| 627 | |
|---|
| 628 | static char *gbs_compress_command(const char *com) { |
|---|
| 629 | /* Prepare SRT. |
|---|
| 630 | * |
|---|
| 631 | * Replaces all |
|---|
| 632 | * '=' by GBS_SET |
|---|
| 633 | * ':' by GBS_SEP |
|---|
| 634 | * '?' by GBS_WILD if followed by a number or '?' |
|---|
| 635 | * '*' by GBS_MWILD or '(' |
|---|
| 636 | * \ is the escape charakter |
|---|
| 637 | */ |
|---|
| 638 | char *result,*s,*d; |
|---|
| 639 | int ch; |
|---|
| 640 | |
|---|
| 641 | s = d = result = strdup(com); |
|---|
| 642 | while ( (ch = *(s++)) ){ |
|---|
| 643 | switch (ch) { |
|---|
| 644 | case '=': *(d++) = GBS_SET;break; |
|---|
| 645 | case ':': *(d++) = GBS_SEP;break; |
|---|
| 646 | case '?': |
|---|
| 647 | ch = *s; |
|---|
| 648 | /*if ( (ch>='0' && ch <='9') || ch=='?'){ *(d++) = GBS_WILD;break;}*/ |
|---|
| 649 | *(d++) = GBS_WILD; break; |
|---|
| 650 | case '*': |
|---|
| 651 | ch = *s; |
|---|
| 652 | /* if ( (ch>='0' && ch <='9') || ch=='('){ *(d++) = GBS_MWILD;break;}*/ |
|---|
| 653 | *(d++) = GBS_MWILD; break; |
|---|
| 654 | case '\\': |
|---|
| 655 | ch = *(s++); if (!ch) { s--; break; }; |
|---|
| 656 | switch (ch) { |
|---|
| 657 | case 'n': *(d++) = '\n';break; |
|---|
| 658 | case 't': *(d++) = '\t';break; |
|---|
| 659 | case '0': *(d++) = '\0';break; |
|---|
| 660 | default: *(d++) = ch;break; |
|---|
| 661 | } |
|---|
| 662 | break; |
|---|
| 663 | default: |
|---|
| 664 | *(d++) = ch; |
|---|
| 665 | } |
|---|
| 666 | } |
|---|
| 667 | *d = 0; |
|---|
| 668 | return result; |
|---|
| 669 | } |
|---|
| 670 | |
|---|
| 671 | |
|---|
| 672 | char *GBS_string_eval(const char *insource, const char *icommand, GBDATA *gb_container) |
|---|
| 673 | /* GBS_string_eval replaces substrings in source |
|---|
| 674 | * Syntax: command = "oliver=olli:peter=peti" |
|---|
| 675 | * |
|---|
| 676 | * Returns a heapcopy of result of replacement. |
|---|
| 677 | * |
|---|
| 678 | * * is a wildcard for any number of character |
|---|
| 679 | * ? is a wildcard for exactly one character |
|---|
| 680 | * |
|---|
| 681 | * To reference to the wildcards on the left side of the '=' |
|---|
| 682 | * use ? and *, to reference in a different order use: |
|---|
| 683 | * *0 to reference to the first occurrence of * |
|---|
| 684 | * *1 second |
|---|
| 685 | * ... |
|---|
| 686 | * *9 |
|---|
| 687 | * |
|---|
| 688 | * if the last and first characters of the search string are no '*' wildcards then |
|---|
| 689 | * the replace is repeated as many times as possible |
|---|
| 690 | * '\' is the escape character: e.g. \n is newline; '\\' is '\'; '\=' is '='; .... |
|---|
| 691 | * |
|---|
| 692 | * eg: |
|---|
| 693 | * print first three characters of first word and the whole second word: |
|---|
| 694 | * |
|---|
| 695 | * *(arb_key) is the value of the a database entry arb key |
|---|
| 696 | * *(arb_key#string) value of the database entry or 'string' if the entry does not exist |
|---|
| 697 | * *(arb_key\:SRT) runs SRT recursively on the value of the database entry |
|---|
| 698 | * *([arb_key]|ACI) runs the ACI command interpreter on the value of the database entry (or on an empty string) |
|---|
| 699 | * |
|---|
| 700 | * If an error occurs it returns NULL - in this case the error was exported. |
|---|
| 701 | */ |
|---|
| 702 | { |
|---|
| 703 | GB_CSTR source; /* pointer into the current string when parsed */ |
|---|
| 704 | char *search; /* pointer into the current command when parsed */ |
|---|
| 705 | GB_CSTR p; /* short live pointer */ |
|---|
| 706 | char c; |
|---|
| 707 | GB_CSTR already_transferred; /* point into 'in' string to non parsed position */ |
|---|
| 708 | |
|---|
| 709 | char wildcard[40]; |
|---|
| 710 | char *mwildcard[10]; |
|---|
| 711 | GB_ERROR error; |
|---|
| 712 | |
|---|
| 713 | long i; |
|---|
| 714 | long max_wildcard; |
|---|
| 715 | long max_mwildcard; |
|---|
| 716 | |
|---|
| 717 | char *start_of_wildcard; |
|---|
| 718 | char what_wild_card; |
|---|
| 719 | |
|---|
| 720 | GB_CSTR start_match; |
|---|
| 721 | |
|---|
| 722 | char *doppelpunkt; |
|---|
| 723 | |
|---|
| 724 | char *bar; |
|---|
| 725 | char *in; |
|---|
| 726 | char *nextdp; |
|---|
| 727 | void *strstruct; |
|---|
| 728 | char *command; |
|---|
| 729 | |
|---|
| 730 | if (!icommand || !icommand[0]) return strdup(insource); |
|---|
| 731 | |
|---|
| 732 | command = gbs_compress_command(icommand); |
|---|
| 733 | in = strdup(insource); /* copy insource to allow to destroy it */ |
|---|
| 734 | |
|---|
| 735 | for (doppelpunkt = command; doppelpunkt; doppelpunkt = nextdp) { /* loop over command string */ |
|---|
| 736 | /* in is in , strstruct is out */ |
|---|
| 737 | max_wildcard = 0; |
|---|
| 738 | max_mwildcard = 0; |
|---|
| 739 | nextdp = strchr(doppelpunkt, GBS_SEP); |
|---|
| 740 | if (nextdp){ |
|---|
| 741 | *(nextdp++) = 0; |
|---|
| 742 | } |
|---|
| 743 | if (!doppelpunkt[0]) { /* empty command -> search next */ |
|---|
| 744 | continue; |
|---|
| 745 | } |
|---|
| 746 | |
|---|
| 747 | bar = strchr(doppelpunkt+1, GBS_SET); /* Parse the command string !!!! */ |
|---|
| 748 | if (bar) { |
|---|
| 749 | *(bar++) = 0; |
|---|
| 750 | } else { |
|---|
| 751 | GB_export_errorf("SRT ERROR: no '=' found in command '%s' (position > %zi)", icommand, doppelpunkt-command+1); |
|---|
| 752 | free(command); |
|---|
| 753 | free(in); |
|---|
| 754 | return 0; |
|---|
| 755 | } |
|---|
| 756 | |
|---|
| 757 | already_transferred = in; |
|---|
| 758 | strstruct = GBS_stropen(1000); /* create output stream */ |
|---|
| 759 | |
|---|
| 760 | if ( (!*in) && doppelpunkt[0] == GBS_MWILD && doppelpunkt[1] == 0) { /* empty string -> pars myself */ |
|---|
| 761 | /* * matches empty string !!!! */ |
|---|
| 762 | mwildcard[max_mwildcard++] = strdup(""); |
|---|
| 763 | gbs_build_replace_string(strstruct,bar,wildcard, max_wildcard, mwildcard, max_mwildcard,gb_container); |
|---|
| 764 | goto gbs_pars_unsuccessfull; /* successfull search*/ |
|---|
| 765 | } |
|---|
| 766 | |
|---|
| 767 | for (source = in;*source; ) { /* loop over string */ |
|---|
| 768 | search = doppelpunkt; |
|---|
| 769 | |
|---|
| 770 | start_match = 0; /* match string for '*' */ |
|---|
| 771 | while ( (c = *(search++)) ) { /* search matching command */ |
|---|
| 772 | switch (c) { |
|---|
| 773 | case GBS_MWILD: |
|---|
| 774 | if (!start_match) start_match = source; |
|---|
| 775 | |
|---|
| 776 | start_of_wildcard = search; |
|---|
| 777 | if ( !(c = *(search++) ) ) { /* last character is a wildcard -> that was it */ |
|---|
| 778 | mwildcard[max_mwildcard++] = strdup(source); |
|---|
| 779 | source += strlen(source); |
|---|
| 780 | goto gbs_pars_successfull; /* successfull search and end wildcard*/ |
|---|
| 781 | } |
|---|
| 782 | while ( (c=*(search++)) && c!=GBS_MWILD && c!=GBS_WILD ); /* search the next wildcardstring */ |
|---|
| 783 | search--; /* back one character */ |
|---|
| 784 | *search = 0; |
|---|
| 785 | what_wild_card = c; |
|---|
| 786 | p = GBS_find_string(source,start_of_wildcard,0); |
|---|
| 787 | if (!p){ /* string not found -> unsuccessful search */ |
|---|
| 788 | goto gbs_pars_unsuccessfull; |
|---|
| 789 | } |
|---|
| 790 | c = *p; /* set wildcard */ |
|---|
| 791 | mwildcard[max_mwildcard++] = GB_strpartdup(source, p-1); |
|---|
| 792 | source = p + strlen(start_of_wildcard); /* we parsed it */ |
|---|
| 793 | *search = what_wild_card; |
|---|
| 794 | break; |
|---|
| 795 | |
|---|
| 796 | case GBS_WILD: |
|---|
| 797 | if ( !source[0] ) { |
|---|
| 798 | goto gbs_pars_unsuccessfull; |
|---|
| 799 | } |
|---|
| 800 | if (!start_match) start_match = source; |
|---|
| 801 | wildcard[max_wildcard++] = *(source++); |
|---|
| 802 | break; |
|---|
| 803 | default: |
|---|
| 804 | if (start_match) { |
|---|
| 805 | if (c != *(source++)) { |
|---|
| 806 | goto gbs_pars_unsuccessfull; |
|---|
| 807 | } |
|---|
| 808 | break; |
|---|
| 809 | }else{ |
|---|
| 810 | char *buf1; |
|---|
| 811 | buf1 = search-1; |
|---|
| 812 | while ( (c=*(search++)) && c!=GBS_MWILD && c!=GBS_WILD ); /* search the next wildcardstring */ |
|---|
| 813 | search--; /* back one character */ |
|---|
| 814 | *search = 0; |
|---|
| 815 | what_wild_card = c; |
|---|
| 816 | p = GBS_find_string(source,buf1,0); |
|---|
| 817 | if (!p){ /* string not found -> unsuccessful search */ |
|---|
| 818 | goto gbs_pars_unsuccessfull; |
|---|
| 819 | } |
|---|
| 820 | start_match = p; |
|---|
| 821 | source = p + strlen(buf1); /* we parsed it */ |
|---|
| 822 | *search = what_wild_card; |
|---|
| 823 | } |
|---|
| 824 | break; |
|---|
| 825 | } /* switch */ |
|---|
| 826 | } /* while */ /* search matching command */ |
|---|
| 827 | |
|---|
| 828 | gbs_pars_successfull: /* now we got source: pointer to end of match |
|---|
| 829 | start_match: pointer to start of match |
|---|
| 830 | in: pointer to the entire string |
|---|
| 831 | already_transferred: pointer to the start of the unparsed string |
|---|
| 832 | bar: the replace string |
|---|
| 833 | */ |
|---|
| 834 | /* now look for the replace string */ |
|---|
| 835 | GBS_strncat(strstruct,already_transferred,start_match-already_transferred); /* cat old data */ |
|---|
| 836 | error = gbs_build_replace_string(strstruct,bar,wildcard, max_wildcard, /* do the command */ |
|---|
| 837 | mwildcard, max_mwildcard,gb_container); |
|---|
| 838 | already_transferred = source; |
|---|
| 839 | |
|---|
| 840 | for (i = 0; i < max_mwildcard; i++) { |
|---|
| 841 | freeset(mwildcard[i], 0); |
|---|
| 842 | } |
|---|
| 843 | max_wildcard = 0; |
|---|
| 844 | max_mwildcard = 0; |
|---|
| 845 | |
|---|
| 846 | if (error) { |
|---|
| 847 | free(GBS_strclose(strstruct)); |
|---|
| 848 | free(command); |
|---|
| 849 | free(in); |
|---|
| 850 | GB_export_error(error); |
|---|
| 851 | return 0; |
|---|
| 852 | } |
|---|
| 853 | } /* while parsing */ |
|---|
| 854 | gbs_pars_unsuccessfull: |
|---|
| 855 | GBS_strcat(strstruct,already_transferred); /* cat the rest data */ |
|---|
| 856 | |
|---|
| 857 | for (i = 0; i < max_mwildcard; i++){ |
|---|
| 858 | freeset(mwildcard[i], 0); |
|---|
| 859 | } |
|---|
| 860 | max_wildcard = 0; |
|---|
| 861 | max_mwildcard = 0; |
|---|
| 862 | |
|---|
| 863 | freeset(in, GBS_strclose(strstruct)); |
|---|
| 864 | } |
|---|
| 865 | free(command); |
|---|
| 866 | return in; |
|---|
| 867 | } |
|---|
| 868 | |
|---|