1 // Copyright (c) 2005, Google Inc. 2 // All rights reserved. 3 // 4 // Redistribution and use in source and binary forms, with or without 5 // modification, are permitted provided that the following conditions are 6 // met: 7 // 8 // * Redistributions of source code must retain the above copyright 9 // notice, this list of conditions and the following disclaimer. 10 // * Redistributions in binary form must reproduce the above 11 // copyright notice, this list of conditions and the following disclaimer 12 // in the documentation and/or other materials provided with the 13 // distribution. 14 // * Neither the name of Google Inc. nor the names of its 15 // contributors may be used to endorse or promote products derived from 16 // this software without specific prior written permission. 17 // 18 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 19 // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 20 // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 21 // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT 22 // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 23 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT 24 // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 25 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 26 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 27 // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE 28 // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 29 // 30 // Author: Sanjay Ghemawat 31 // Support for PCRE_XXX modifiers added by Giuseppe Maxia, July 2005 32 33 #ifndef _PCRECPP_H 34 #define _PCRECPP_H 35 36 // C++ interface to the pcre regular-expression library. RE supports 37 // Perl-style regular expressions (with extensions like \d, \w, \s, 38 // ...). 39 // 40 // ----------------------------------------------------------------------- 41 // REGEXP SYNTAX: 42 // 43 // This module is part of the pcre library and hence supports its syntax 44 // for regular expressions. 45 // 46 // The syntax is pretty similar to Perl's. For those not familiar 47 // with Perl's regular expressions, here are some examples of the most 48 // commonly used extensions: 49 // 50 // "hello (\\w+) world" -- \w matches a "word" character 51 // "version (\\d+)" -- \d matches a digit 52 // "hello\\s+world" -- \s matches any whitespace character 53 // "\\b(\\w+)\\b" -- \b matches empty string at a word boundary 54 // "(?i)hello" -- (?i) turns on case-insensitive matching 55 // "/\\*(.*?)\\*/" -- .*? matches . minimum no. of times possible 56 // 57 // ----------------------------------------------------------------------- 58 // MATCHING INTERFACE: 59 // 60 // The "FullMatch" operation checks that supplied text matches a 61 // supplied pattern exactly. 62 // 63 // Example: successful match 64 // pcrecpp::RE re("h.*o"); 65 // re.FullMatch("hello"); 66 // 67 // Example: unsuccessful match (requires full match): 68 // pcrecpp::RE re("e"); 69 // !re.FullMatch("hello"); 70 // 71 // Example: creating a temporary RE object: 72 // pcrecpp::RE("h.*o").FullMatch("hello"); 73 // 74 // You can pass in a "const char*" or a "string" for "text". The 75 // examples below tend to use a const char*. 76 // 77 // You can, as in the different examples above, store the RE object 78 // explicitly in a variable or use a temporary RE object. The 79 // examples below use one mode or the other arbitrarily. Either 80 // could correctly be used for any of these examples. 81 // 82 // ----------------------------------------------------------------------- 83 // MATCHING WITH SUB-STRING EXTRACTION: 84 // 85 // You can supply extra pointer arguments to extract matched subpieces. 86 // 87 // Example: extracts "ruby" into "s" and 1234 into "i" 88 // int i; 89 // string s; 90 // pcrecpp::RE re("(\\w+):(\\d+)"); 91 // re.FullMatch("ruby:1234", &s, &i); 92 // 93 // Example: does not try to extract any extra sub-patterns 94 // re.FullMatch("ruby:1234", &s); 95 // 96 // Example: does not try to extract into NULL 97 // re.FullMatch("ruby:1234", NULL, &i); 98 // 99 // Example: integer overflow causes failure 100 // !re.FullMatch("ruby:1234567891234", NULL, &i); 101 // 102 // Example: fails because there aren't enough sub-patterns: 103 // !pcrecpp::RE("\\w+:\\d+").FullMatch("ruby:1234", &s); 104 // 105 // Example: fails because string cannot be stored in integer 106 // !pcrecpp::RE("(.*)").FullMatch("ruby", &i); 107 // 108 // The provided pointer arguments can be pointers to any scalar numeric 109 // type, or one of 110 // string (matched piece is copied to string) 111 // StringPiece (StringPiece is mutated to point to matched piece) 112 // T (where "bool T::ParseFrom(const char*, int)" exists) 113 // NULL (the corresponding matched sub-pattern is not copied) 114 // 115 // CAVEAT: An optional sub-pattern that does not exist in the matched 116 // string is assigned the empty string. Therefore, the following will 117 // return false (because the empty string is not a valid number): 118 // int number; 119 // pcrecpp::RE::FullMatch("abc", "[a-z]+(\\d+)?", &number); 120 // 121 // ----------------------------------------------------------------------- 122 // DO_MATCH 123 // 124 // The matching interface supports at most 16 arguments per call. 125 // If you need more, consider using the more general interface 126 // pcrecpp::RE::DoMatch(). See pcrecpp.h for the signature for DoMatch. 127 // 128 // ----------------------------------------------------------------------- 129 // PARTIAL MATCHES 130 // 131 // You can use the "PartialMatch" operation when you want the pattern 132 // to match any substring of the text. 133 // 134 // Example: simple search for a string: 135 // pcrecpp::RE("ell").PartialMatch("hello"); 136 // 137 // Example: find first number in a string: 138 // int number; 139 // pcrecpp::RE re("(\\d+)"); 140 // re.PartialMatch("x*100 + 20", &number); 141 // assert(number == 100); 142 // 143 // ----------------------------------------------------------------------- 144 // UTF-8 AND THE MATCHING INTERFACE: 145 // 146 // By default, pattern and text are plain text, one byte per character. 147 // The UTF8 flag, passed to the constructor, causes both pattern 148 // and string to be treated as UTF-8 text, still a byte stream but 149 // potentially multiple bytes per character. In practice, the text 150 // is likelier to be UTF-8 than the pattern, but the match returned 151 // may depend on the UTF8 flag, so always use it when matching 152 // UTF8 text. E.g., "." will match one byte normally but with UTF8 153 // set may match up to three bytes of a multi-byte character. 154 // 155 // Example: 156 // pcrecpp::RE_Options options; 157 // options.set_utf8(); 158 // pcrecpp::RE re(utf8_pattern, options); 159 // re.FullMatch(utf8_string); 160 // 161 // Example: using the convenience function UTF8(): 162 // pcrecpp::RE re(utf8_pattern, pcrecpp::UTF8()); 163 // re.FullMatch(utf8_string); 164 // 165 // NOTE: The UTF8 option is ignored if pcre was not configured with the 166 // --enable-utf8 flag. 167 // 168 // ----------------------------------------------------------------------- 169 // PASSING MODIFIERS TO THE REGULAR EXPRESSION ENGINE 170 // 171 // PCRE defines some modifiers to change the behavior of the regular 172 // expression engine. 173 // The C++ wrapper defines an auxiliary class, RE_Options, as a vehicle 174 // to pass such modifiers to a RE class. 175 // 176 // Currently, the following modifiers are supported 177 // 178 // modifier description Perl corresponding 179 // 180 // PCRE_CASELESS case insensitive match /i 181 // PCRE_MULTILINE multiple lines match /m 182 // PCRE_DOTALL dot matches newlines /s 183 // PCRE_DOLLAR_ENDONLY $ matches only at end N/A 184 // PCRE_EXTRA strict escape parsing N/A 185 // PCRE_EXTENDED ignore whitespaces /x 186 // PCRE_UTF8 handles UTF8 chars built-in 187 // PCRE_UNGREEDY reverses * and *? N/A 188 // PCRE_NO_AUTO_CAPTURE disables matching parens N/A (*) 189 // 190 // (For a full account on how each modifier works, please check the 191 // PCRE API reference manual). 192 // 193 // (*) Both Perl and PCRE allow non matching parentheses by means of the 194 // "?:" modifier within the pattern itself. e.g. (?:ab|cd) does not 195 // capture, while (ab|cd) does. 196 // 197 // For each modifier, there are two member functions whose name is made 198 // out of the modifier in lowercase, without the "PCRE_" prefix. For 199 // instance, PCRE_CASELESS is handled by 200 // bool caseless(), 201 // which returns true if the modifier is set, and 202 // RE_Options & set_caseless(bool), 203 // which sets or unsets the modifier. 204 // 205 // Moreover, PCRE_EXTRA_MATCH_LIMIT can be accessed through the 206 // set_match_limit() and match_limit() member functions. 207 // Setting match_limit to a non-zero value will limit the executation of 208 // pcre to keep it from doing bad things like blowing the stack or taking 209 // an eternity to return a result. A value of 5000 is good enough to stop 210 // stack blowup in a 2MB thread stack. Setting match_limit to zero will 211 // disable match limiting. Alternately, you can set match_limit_recursion() 212 // which uses PCRE_EXTRA_MATCH_LIMIT_RECURSION to limit how much pcre 213 // recurses. match_limit() caps the number of matches pcre does; 214 // match_limit_recrusion() caps the depth of recursion. 215 // 216 // Normally, to pass one or more modifiers to a RE class, you declare 217 // a RE_Options object, set the appropriate options, and pass this 218 // object to a RE constructor. Example: 219 // 220 // RE_options opt; 221 // opt.set_caseless(true); 222 // 223 // if (RE("HELLO", opt).PartialMatch("hello world")) ... 224 // 225 // RE_options has two constructors. The default constructor takes no 226 // arguments and creates a set of flags that are off by default. 227 // 228 // The optional parameter 'option_flags' is to facilitate transfer 229 // of legacy code from C programs. This lets you do 230 // RE(pattern, RE_Options(PCRE_CASELESS|PCRE_MULTILINE)).PartialMatch(str); 231 // 232 // But new code is better off doing 233 // RE(pattern, 234 // RE_Options().set_caseless(true).set_multiline(true)).PartialMatch(str); 235 // (See below) 236 // 237 // If you are going to pass one of the most used modifiers, there are some 238 // convenience functions that return a RE_Options class with the 239 // appropriate modifier already set: 240 // CASELESS(), UTF8(), MULTILINE(), DOTALL(), EXTENDED() 241 // 242 // If you need to set several options at once, and you don't want to go 243 // through the pains of declaring a RE_Options object and setting several 244 // options, there is a parallel method that give you such ability on the 245 // fly. You can concatenate several set_xxxxx member functions, since each 246 // of them returns a reference to its class object. e.g.: to pass 247 // PCRE_CASELESS, PCRE_EXTENDED, and PCRE_MULTILINE to a RE with one 248 // statement, you may write 249 // 250 // RE(" ^ xyz \\s+ .* blah$", RE_Options() 251 // .set_caseless(true) 252 // .set_extended(true) 253 // .set_multiline(true)).PartialMatch(sometext); 254 // 255 // ----------------------------------------------------------------------- 256 // SCANNING TEXT INCREMENTALLY 257 // 258 // The "Consume" operation may be useful if you want to repeatedly 259 // match regular expressions at the front of a string and skip over 260 // them as they match. This requires use of the "StringPiece" type, 261 // which represents a sub-range of a real string. Like RE, StringPiece 262 // is defined in the pcrecpp namespace. 263 // 264 // Example: read lines of the form "var = value" from a string. 265 // string contents = ...; // Fill string somehow 266 // pcrecpp::StringPiece input(contents); // Wrap in a StringPiece 267 // 268 // string var; 269 // int value; 270 // pcrecpp::RE re("(\\w+) = (\\d+)\n"); 271 // while (re.Consume(&input, &var, &value)) { 272 // ...; 273 // } 274 // 275 // Each successful call to "Consume" will set "var/value", and also 276 // advance "input" so it points past the matched text. 277 // 278 // The "FindAndConsume" operation is similar to "Consume" but does not 279 // anchor your match at the beginning of the string. For example, you 280 // could extract all words from a string by repeatedly calling 281 // pcrecpp::RE("(\\w+)").FindAndConsume(&input, &word) 282 // 283 // ----------------------------------------------------------------------- 284 // PARSING HEX/OCTAL/C-RADIX NUMBERS 285 // 286 // By default, if you pass a pointer to a numeric value, the 287 // corresponding text is interpreted as a base-10 number. You can 288 // instead wrap the pointer with a call to one of the operators Hex(), 289 // Octal(), or CRadix() to interpret the text in another base. The 290 // CRadix operator interprets C-style "0" (base-8) and "0x" (base-16) 291 // prefixes, but defaults to base-10. 292 // 293 // Example: 294 // int a, b, c, d; 295 // pcrecpp::RE re("(.*) (.*) (.*) (.*)"); 296 // re.FullMatch("100 40 0100 0x40", 297 // pcrecpp::Octal(&a), pcrecpp::Hex(&b), 298 // pcrecpp::CRadix(&c), pcrecpp::CRadix(&d)); 299 // will leave 64 in a, b, c, and d. 300 // 301 // ----------------------------------------------------------------------- 302 // REPLACING PARTS OF STRINGS 303 // 304 // You can replace the first match of "pattern" in "str" with 305 // "rewrite". Within "rewrite", backslash-escaped digits (\1 to \9) 306 // can be used to insert text matching corresponding parenthesized 307 // group from the pattern. \0 in "rewrite" refers to the entire 308 // matching text. E.g., 309 // 310 // string s = "yabba dabba doo"; 311 // pcrecpp::RE("b+").Replace("d", &s); 312 // 313 // will leave "s" containing "yada dabba doo". The result is true if 314 // the pattern matches and a replacement occurs, or false otherwise. 315 // 316 // GlobalReplace() is like Replace(), except that it replaces all 317 // occurrences of the pattern in the string with the rewrite. 318 // Replacements are not subject to re-matching. E.g., 319 // 320 // string s = "yabba dabba doo"; 321 // pcrecpp::RE("b+").GlobalReplace("d", &s); 322 // 323 // will leave "s" containing "yada dada doo". It returns the number 324 // of replacements made. 325 // 326 // Extract() is like Replace(), except that if the pattern matches, 327 // "rewrite" is copied into "out" (an additional argument) with 328 // substitutions. The non-matching portions of "text" are ignored. 329 // Returns true iff a match occurred and the extraction happened 330 // successfully. If no match occurs, the string is left unaffected. 331 332 333 #include <string> 334 #include <pcre2.h> 335 #include <pcrecpparg.h> // defines the Arg class 336 // This isn't technically needed here, but we include it 337 // anyway so folks who include pcrecpp.h don't have to. 338 #include <pcre_stringpiece.h> 339 #include <memory> 340 341 namespace pcrecpp { 342 343 #define PCRE_SET_OR_CLEAR(b, o) \ 344 if (b) all_options_ |= (o); else all_options_ &= ~(o); \ 345 return *this 346 347 #define PCRE_IS_SET(o) \ 348 (all_options_ & o) == o 349 350 typedef std::shared_ptr<pcre2_match_data> pcre2_match_data_ptr; 351 352 /***** Compiling regular expressions: the RE class *****/ 353 354 // RE_Options allow you to set options to be passed along to pcre, 355 // along with other options we put on top of pcre. 356 // Only 9 modifiers, plus match_limit and match_limit_recursion, 357 // are supported now. 358 class RE_Options { 359 public: 360 // constructor 361 RE_Options() 362 : newline_mode_(0), 363 match_limit_(0), 364 match_limit_recursion_(0), 365 all_options_(0) { 366 } 367 368 // alternative constructor. 369 // To facilitate transfer of legacy code from C programs 370 // 371 // This lets you do 372 // RE(pattern, RE_Options(PCRE_CASELESS|PCRE_MULTILINE)).PartialMatch(str); 373 // But new code is better off doing 374 // RE(pattern, 375 // RE_Options().set_caseless(true).set_multiline(true)).PartialMatch(str); 376 RE_Options(int option_flags) 377 : newline_mode_(0), 378 match_limit_(0), 379 match_limit_recursion_(0), 380 all_options_(option_flags) { 381 } 382 // we're fine with the default destructor, copy constructor, etc. 383 384 // accessors and mutators 385 int match_limit() const { return match_limit_; }; 386 RE_Options &set_match_limit(int limit) { 387 match_limit_ = limit; 388 return *this; 389 } 390 391 int match_limit_recursion() const { return match_limit_recursion_; }; 392 RE_Options &set_match_limit_recursion(int limit) { 393 match_limit_recursion_ = limit; 394 return *this; 395 } 396 397 bool caseless() const { 398 return PCRE_IS_SET(PCRE2_CASELESS); 399 } 400 RE_Options &set_caseless(bool x) { 401 PCRE_SET_OR_CLEAR(x, PCRE2_CASELESS); 402 } 403 404 bool multiline() const { 405 return PCRE_IS_SET(PCRE2_MULTILINE); 406 } 407 RE_Options &set_multiline(bool x) { 408 PCRE_SET_OR_CLEAR(x, PCRE2_MULTILINE); 409 } 410 411 int newline_mode() const { 412 if(newline_mode_) 413 return newline_mode_; 414 else { 415 // if newline_mode_ is 0 return the global configuration default 416 int value; 417 pcre2_config_8(PCRE2_CONFIG_NEWLINE, &value); 418 return value; 419 } 420 } 421 RE_Options & set_newline_mode(int newline_mode) { 422 newline_mode_ = newline_mode; 423 return *this; 424 } 425 426 bool dotall() const { 427 return PCRE_IS_SET(PCRE2_DOTALL); 428 } 429 RE_Options &set_dotall(bool x) { 430 PCRE_SET_OR_CLEAR(x, PCRE2_DOTALL); 431 } 432 433 bool extended() const { 434 return PCRE_IS_SET(PCRE2_EXTENDED); 435 } 436 RE_Options &set_extended(bool x) { 437 PCRE_SET_OR_CLEAR(x, PCRE2_EXTENDED); 438 } 439 440 bool dollar_endonly() const { 441 return PCRE_IS_SET(PCRE2_DOLLAR_ENDONLY); 442 } 443 RE_Options &set_dollar_endonly(bool x) { 444 PCRE_SET_OR_CLEAR(x, PCRE2_DOLLAR_ENDONLY); 445 } 446 447 bool ungreedy() const { 448 return PCRE_IS_SET(PCRE2_UNGREEDY); 449 } 450 RE_Options &set_ungreedy(bool x) { 451 PCRE_SET_OR_CLEAR(x, PCRE2_UNGREEDY); 452 } 453 454 bool utf() const { 455 return PCRE_IS_SET(PCRE2_UTF); 456 } 457 RE_Options &set_utf(bool x) { 458 PCRE_SET_OR_CLEAR(x, PCRE2_UTF); 459 } 460 461 bool no_auto_capture() const { 462 return PCRE_IS_SET(PCRE2_NO_AUTO_CAPTURE); 463 } 464 RE_Options &set_no_auto_capture(bool x) { 465 PCRE_SET_OR_CLEAR(x, PCRE2_NO_AUTO_CAPTURE); 466 } 467 468 RE_Options &set_all_options(int opt) { 469 all_options_ = opt; 470 return *this; 471 } 472 int all_options() const { 473 return all_options_ ; 474 } 475 476 // TODO: add other pcre flags 477 478 private: 479 int newline_mode_; 480 int match_limit_; 481 int match_limit_recursion_; 482 int all_options_; 483 }; 484 485 // These functions return some common RE_Options 486 static inline RE_Options UTF() { 487 return RE_Options().set_utf(true); 488 } 489 490 static inline RE_Options CASELESS() { 491 return RE_Options().set_caseless(true); 492 } 493 static inline RE_Options MULTILINE() { 494 return RE_Options().set_multiline(true); 495 } 496 497 static inline RE_Options DOTALL() { 498 return RE_Options().set_dotall(true); 499 } 500 501 static inline RE_Options EXTENDED() { 502 return RE_Options().set_extended(true); 503 } 504 505 // Interface for regular expression matching. Also corresponds to a 506 // pre-compiled regular expression. An "RE" object is safe for 507 // concurrent use by multiple threads. 508 class RE { 509 public: 510 // We provide implicit conversions from strings so that users can 511 // pass in a string or a "const char*" wherever an "RE" is expected. 512 RE(const string& pat) { Init(pat, NULL); } 513 RE(const string& pat, const RE_Options& option) { Init(pat, &option); } 514 RE(const char* pat) { Init(pat, NULL); } 515 RE(const char* pat, const RE_Options& option) { Init(pat, &option); } 516 RE(const unsigned char* pat) { 517 Init(reinterpret_cast<const char*>(pat), NULL); 518 } 519 RE(const unsigned char* pat, const RE_Options& option) { 520 Init(reinterpret_cast<const char*>(pat), &option); 521 } 522 523 // Copy constructor & assignment - note that these are expensive 524 // because they recompile the expression. 525 RE(const RE& re) { Init(re.pattern_, &re.options_); } 526 const RE& operator=(const RE& re) { 527 if (this != &re) { 528 Cleanup(); 529 530 // This is the code that originally came from Google 531 // Init(re.pattern_.c_str(), &re.options_); 532 533 // This is the replacement from Ari Pollak 534 Init(re.pattern_, &re.options_); 535 } 536 return *this; 537 } 538 539 540 ~RE(); 541 542 // The string specification for this RE. E.g. 543 // RE re("ab*c?d+"); 544 // re.pattern(); // "ab*c?d+" 545 const string& pattern() const { return pattern_; } 546 547 // If RE could not be created properly, returns an error string. 548 // Else returns the empty string. 549 const string& error() const { return error_; } 550 551 /***** The useful part: the matching interface *****/ 552 553 // This is provided so one can do pattern.ReplaceAll() just as 554 // easily as ReplaceAll(pattern-text, ....) 555 556 template<typename ... ARGS> 557 bool FullMatch(const StringPiece & text, ARGS && ...a) const { 558 // create an array with the size of the number of arguments given 559 Arg args[Args<ARGS...>::count()]; 560 // initialize the array with the arguments given 561 Args<ARGS...>::arrayify(args, a...); 562 563 return DoMatchImpl(text, ANCHOR_BOTH, NULL, args, Args<ARGS...>::count()); 564 } 565 566 template<typename ... ARGS> 567 bool PartialMatch(const StringPiece& text, ARGS && ...a) const { 568 // create an array with the size of the number of arguments given 569 Arg args[Args<ARGS...>::count()]; 570 // initialize the array with the arguments given 571 Args<ARGS...>::arrayify(args, a...); 572 573 return DoMatchImpl(text, UNANCHORED, NULL, args, Args<ARGS...>::count()); 574 } 575 576 template<typename ... ARGS> 577 bool Consume(StringPiece* input, ARGS && ...a) const { 578 // create an array with the size of the number of arguments given 579 Arg args[Args<ARGS...>::count()]; 580 // initialize the array with the arguments given 581 Args<ARGS...>::arrayify(args, a...); 582 583 int consumed; 584 if (DoMatchImpl(*input, ANCHOR_START, &consumed, args, 585 Args<ARGS...>::count())) { 586 input->remove_prefix(consumed); 587 return true; 588 } else { 589 return false; 590 } 591 } 592 593 template<typename ... ARGS> 594 bool FindAndConsume(StringPiece* input, ARGS && ...a) const { 595 Arg args[Args<ARGS...>::count()]; 596 Args<ARGS...>::arrayify(args, a...); 597 int consumed; 598 if (DoMatchImpl(*input, UNANCHORED, &consumed, args, 599 Args<ARGS...>::count())) { 600 input->remove_prefix(consumed); 601 return true; 602 } else { 603 return false; 604 } 605 } 606 607 bool Replace(const StringPiece& rewrite, 608 string *str) const; 609 610 int GlobalReplace(const StringPiece& rewrite, 611 string *str) const; 612 613 bool Extract(const StringPiece &rewrite, 614 const StringPiece &text, 615 string *out) const; 616 617 // Escapes all potentially meaningful regexp characters in 618 // 'unquoted'. The returned string, used as a regular expression, 619 // will exactly match the original string. For example, 620 // 1.5-2.0? 621 // may become: 622 // 1\.5\-2\.0\? 623 // Note QuoteMeta behaves the same as perl's QuoteMeta function, 624 // *except* that it escapes the NUL character (\0) as backslash + 0, 625 // rather than backslash + NUL. 626 static string QuoteMeta(const StringPiece& unquoted); 627 628 629 /***** Generic matching interface *****/ 630 631 // Type of match (TODO: Should be restructured as part of RE_Options) 632 enum Anchor { 633 UNANCHORED, // No anchoring 634 ANCHOR_START, // Anchor at start only 635 ANCHOR_BOTH // Anchor at start and end 636 }; 637 638 // General matching routine. Stores the length of the match in 639 // "*consumed" if successful. 640 bool DoMatch(const StringPiece& text, 641 Anchor anchor, 642 int* consumed, 643 Arg const argsp[], int n) const; 644 645 // Return the number of capturing subpatterns, or -1 if the 646 // regexp wasn't valid on construction. 647 int NumberOfCapturingGroups() const; 648 649 private: 650 651 void Init(const string& pattern, const RE_Options* options); 652 void Cleanup(); 653 654 // Match against "text", filling in "vec" (up to "vecsize" * 2/3) with 655 // pairs of integers for the beginning and end positions of matched 656 // text. The first pair corresponds to the entire matched text; 657 // subsequent pairs correspond, in order, to parentheses-captured 658 // matches. Returns the number of pairs (one more than the number of 659 // the last subpattern with a match) if matching was successful 660 // and zero if the match failed. 661 // I.e. for RE("(foo)|(bar)|(baz)") it will return 2, 3, and 4 when matching 662 // against "foo", "bar", and "baz" respectively. 663 // When matching RE("(foo)|hello") against "hello", it will return 1. 664 // But the values for all subpattern are filled in into "vec". 665 int TryMatch(const StringPiece& text, 666 int startpos, 667 Anchor anchor, 668 bool empty_ok, 669 pcre2_match_data_ptr & match_data) const; 670 671 // Append the "rewrite" string, with backslash subsitutions from "text" 672 // and "vec", to string "out". 673 bool Rewrite(string *out, 674 const StringPiece& rewrite, 675 const StringPiece& text, 676 pcre2_match_data_ptr const & match_data) const; 677 678 // internal implementation for DoMatch 679 bool DoMatchImpl(const StringPiece& text, 680 Anchor anchor, 681 int* consumed, 682 const Arg args[], 683 int n) const; 684 685 // Compile the regexp for the specified anchoring mode 686 pcre2_code * Compile(Anchor anchor); 687 688 string pattern_; 689 RE_Options options_; 690 pcre2_code* re_full_; // For full matches 691 pcre2_code* re_partial_; // For partial matches 692 string error_; // Error indicator 693 }; 694 695 } // namespace pcrecpp 696 697 #endif /* _PCRECPP_H */ 698