1PCRETEST(1) General Commands Manual PCRETEST(1) 2 3 4 5NAME 6 pcretest - a program for testing Perl-compatible regular expressions. 7 8SYNOPSIS 9 10 pcretest [options] [input file [output file]] 11 12 pcretest was written as a test program for the PCRE regular expression 13 library itself, but it can also be used for experimenting with regular 14 expressions. This document describes the features of the test program; 15 for details of the regular expressions themselves, see the pcrepattern 16 documentation. For details of the PCRE library function calls and their 17 options, see the pcreapi , pcre16 and pcre32 documentation. 18 19 The input for pcretest is a sequence of regular expression patterns and 20 strings to be matched, as described below. The output shows the result 21 of each match. Options on the command line and the patterns control 22 PCRE options and exactly what is output. 23 24 As PCRE has evolved, it has acquired many different features, and as a 25 result, pcretest now has rather a lot of obscure options for testing 26 every possible feature. Some of these options are specifically designed 27 for use in conjunction with the test script and data files that are 28 distributed as part of PCRE, and are unlikely to be of use otherwise. 29 They are all documented here, but without much justification. 30 31 32INPUT DATA FORMAT 33 34 Input to pcretest is processed line by line, either by calling the C 35 library's fgets() function, or via the libreadline library (see below). 36 In Unix-like environments, fgets() treats any bytes other than newline 37 as data characters. However, in some Windows environments character 26 38 (hex 1A) causes an immediate end of file, and no further data is read. 39 For maximum portability, therefore, it is safest to use only ASCII 40 characters in pcretest input files. 41 42 43PCRE's 8-BIT, 16-BIT AND 32-BIT LIBRARIES 44 45 From release 8.30, two separate PCRE libraries can be built. The origi- 46 nal one supports 8-bit character strings, whereas the newer 16-bit 47 library supports character strings encoded in 16-bit units. From 48 release 8.32, a third library can be built, supporting character 49 strings encoded in 32-bit units. The pcretest program can be used to 50 test all three libraries. However, it is itself still an 8-bit program, 51 reading 8-bit input and writing 8-bit output. When testing the 16-bit 52 or 32-bit library, the patterns and data strings are converted to 16- 53 or 32-bit format before being passed to the PCRE library functions. 54 Results are converted to 8-bit for output. 55 56 References to functions and structures of the form pcre[16|32]_xx below 57 mean "pcre_xx when using the 8-bit library, pcre16_xx when using the 58 16-bit library, or pcre32_xx when using the 32-bit library". 59 60 61COMMAND LINE OPTIONS 62 63 -8 If both the 8-bit library has been built, this option causes 64 the 8-bit library to be used (which is the default); if the 65 8-bit library has not been built, this option causes an 66 error. 67 68 -16 If both the 8-bit or the 32-bit, and the 16-bit libraries 69 have been built, this option causes the 16-bit library to be 70 used. If only the 16-bit library has been built, this is the 71 default (so has no effect). If only the 8-bit or the 32-bit 72 library has been built, this option causes an error. 73 74 -32 If both the 8-bit or the 16-bit, and the 32-bit libraries 75 have been built, this option causes the 32-bit library to be 76 used. If only the 32-bit library has been built, this is the 77 default (so has no effect). If only the 8-bit or the 16-bit 78 library has been built, this option causes an error. 79 80 -b Behave as if each pattern has the /B (show byte code) modi- 81 fier; the internal form is output after compilation. 82 83 -C Output the version number of the PCRE library, and all avail- 84 able information about the optional features that are 85 included, and then exit with zero exit code. All other 86 options are ignored. 87 88 -C option Output information about a specific build-time option, then 89 exit. This functionality is intended for use in scripts such 90 as RunTest. The following options output the value and set 91 the exit code as indicated: 92 93 ebcdic-nl the code for LF (= NL) in an EBCDIC environment: 94 0x15 or 0x25 95 0 if used in an ASCII environment 96 exit code is always 0 97 linksize the configured internal link size (2, 3, or 4) 98 exit code is set to the link size 99 newline the default newline setting: 100 CR, LF, CRLF, ANYCRLF, or ANY 101 exit code is always 0 102 bsr the default setting for what \R matches: 103 ANYCRLF or ANY 104 exit code is always 0 105 106 The following options output 1 for true or 0 for false, and 107 set the exit code to the same value: 108 109 ebcdic compiled for an EBCDIC environment 110 jit just-in-time support is available 111 pcre16 the 16-bit library was built 112 pcre32 the 32-bit library was built 113 pcre8 the 8-bit library was built 114 ucp Unicode property support is available 115 utf UTF-8 and/or UTF-16 and/or UTF-32 support 116 is available 117 118 If an unknown option is given, an error message is output; 119 the exit code is 0. 120 121 -d Behave as if each pattern has the /D (debug) modifier; the 122 internal form and information about the compiled pattern is 123 output after compilation; -d is equivalent to -b -i. 124 125 -dfa Behave as if each data line contains the \D escape sequence; 126 this causes the alternative matching function, 127 pcre[16|32]_dfa_exec(), to be used instead of the standard 128 pcre[16|32]_exec() function (more detail is given below). 129 130 -help Output a brief summary these options and then exit. 131 132 -i Behave as if each pattern has the /I modifier; information 133 about the compiled pattern is given after compilation. 134 135 -M Behave as if each data line contains the \M escape sequence; 136 this causes PCRE to discover the minimum MATCH_LIMIT and 137 MATCH_LIMIT_RECURSION settings by calling pcre[16|32]_exec() 138 repeatedly with different limits. 139 140 -m Output the size of each compiled pattern after it has been 141 compiled. This is equivalent to adding /M to each regular 142 expression. The size is given in bytes for both libraries. 143 144 -O Behave as if each pattern has the /O modifier, that is dis- 145 able auto-possessification for all patterns. 146 147 -o osize Set the number of elements in the output vector that is used 148 when calling pcre[16|32]_exec() or pcre[16|32]_dfa_exec() to 149 be osize. The default value is 45, which is enough for 14 150 capturing subexpressions for pcre[16|32]_exec() or 22 differ- 151 ent matches for pcre[16|32]_dfa_exec(). The vector size can 152 be changed for individual matching calls by including \O in 153 the data line (see below). 154 155 -p Behave as if each pattern has the /P modifier; the POSIX 156 wrapper API is used to call PCRE. None of the other options 157 has any effect when -p is set. This option can be used only 158 with the 8-bit library. 159 160 -q Do not output the version number of pcretest at the start of 161 execution. 162 163 -S size On Unix-like systems, set the size of the run-time stack to 164 size megabytes. 165 166 -s or -s+ Behave as if each pattern has the /S modifier; in other 167 words, force each pattern to be studied. If -s+ is used, all 168 the JIT compile options are passed to pcre[16|32]_study(), 169 causing just-in-time optimization to be set up if it is 170 available, for both full and partial matching. Specific JIT 171 compile options can be selected by following -s+ with a digit 172 in the range 1 to 7, which selects the JIT compile modes as 173 follows: 174 175 1 normal match only 176 2 soft partial match only 177 3 normal match and soft partial match 178 4 hard partial match only 179 6 soft and hard partial match 180 7 all three modes (default) 181 182 If -s++ is used instead of -s+ (with or without a following 183 digit), the text "(JIT)" is added to the first output line 184 after a match or no match when JIT-compiled code was actually 185 used. 186 187 Note that there are pattern options that can override -s, 188 either specifying no studying at all, or suppressing JIT com- 189 pilation. 190 191 If the /I or /D option is present on a pattern (requesting 192 output about the compiled pattern), information about the 193 result of studying is not included when studying is caused 194 only by -s and neither -i nor -d is present on the command 195 line. This behaviour means that the output from tests that 196 are run with and without -s should be identical, except when 197 options that output information about the actual running of a 198 match are set. 199 200 The -M, -t, and -tm options, which give information about 201 resources used, are likely to produce different output with 202 and without -s. Output may also differ if the /C option is 203 present on an individual pattern. This uses callouts to trace 204 the the matching process, and this may be different between 205 studied and non-studied patterns. If the pattern contains 206 (*MARK) items there may also be differences, for the same 207 reason. The -s command line option can be overridden for spe- 208 cific patterns that should never be studied (see the /S pat- 209 tern modifier below). 210 211 -t Run each compile, study, and match many times with a timer, 212 and output the resulting times per compile, study, or match 213 (in milliseconds). Do not set -m with -t, because you will 214 then get the size output a zillion times, and the timing will 215 be distorted. You can control the number of iterations that 216 are used for timing by following -t with a number (as a sepa- 217 rate item on the command line). For example, "-t 1000" iter- 218 ates 1000 times. The default is to iterate 500000 times. 219 220 -tm This is like -t except that it times only the matching phase, 221 not the compile or study phases. 222 223 -T -TM These behave like -t and -tm, but in addition, at the end of 224 a run, the total times for all compiles, studies, and matches 225 are output. 226 227 228DESCRIPTION 229 230 If pcretest is given two filename arguments, it reads from the first 231 and writes to the second. If it is given only one filename argument, it 232 reads from that file and writes to stdout. Otherwise, it reads from 233 stdin and writes to stdout, and prompts for each line of input, using 234 "re>" to prompt for regular expressions, and "data>" to prompt for data 235 lines. 236 237 When pcretest is built, a configuration option can specify that it 238 should be linked with the libreadline library. When this is done, if 239 the input is from a terminal, it is read using the readline() function. 240 This provides line-editing and history facilities. The output from the 241 -help option states whether or not readline() will be used. 242 243 The program handles any number of sets of input on a single input file. 244 Each set starts with a regular expression, and continues with any num- 245 ber of data lines to be matched against that pattern. 246 247 Each data line is matched separately and independently. If you want to 248 do multi-line matches, you have to use the \n escape sequence (or \r or 249 \r\n, etc., depending on the newline setting) in a single line of input 250 to encode the newline sequences. There is no limit on the length of 251 data lines; the input buffer is automatically extended if it is too 252 small. 253 254 An empty line signals the end of the data lines, at which point a new 255 regular expression is read. The regular expressions are given enclosed 256 in any non-alphanumeric delimiters other than backslash, for example: 257 258 /(a|bc)x+yz/ 259 260 White space before the initial delimiter is ignored. A regular expres- 261 sion may be continued over several input lines, in which case the new- 262 line characters are included within it. It is possible to include the 263 delimiter within the pattern by escaping it, for example 264 265 /abc\/def/ 266 267 If you do so, the escape and the delimiter form part of the pattern, 268 but since delimiters are always non-alphanumeric, this does not affect 269 its interpretation. If the terminating delimiter is immediately fol- 270 lowed by a backslash, for example, 271 272 /abc/\ 273 274 then a backslash is added to the end of the pattern. This is done to 275 provide a way of testing the error condition that arises if a pattern 276 finishes with a backslash, because 277 278 /abc\/ 279 280 is interpreted as the first line of a pattern that starts with "abc/", 281 causing pcretest to read the next line as a continuation of the regular 282 expression. 283 284 285PATTERN MODIFIERS 286 287 A pattern may be followed by any number of modifiers, which are mostly 288 single characters, though some of these can be qualified by further 289 characters. Following Perl usage, these are referred to below as, for 290 example, "the /i modifier", even though the delimiter of the pattern 291 need not always be a slash, and no slash is used when writing modi- 292 fiers. White space may appear between the final pattern delimiter and 293 the first modifier, and between the modifiers themselves. For refer- 294 ence, here is a complete list of modifiers. They fall into several 295 groups that are described in detail in the following sections. 296 297 /8 set UTF mode 298 /9 set PCRE_NEVER_UTF (locks out UTF mode) 299 /? disable UTF validity check 300 /+ show remainder of subject after match 301 /= show all captures (not just those that are set) 302 303 /A set PCRE_ANCHORED 304 /B show compiled code 305 /C set PCRE_AUTO_CALLOUT 306 /D same as /B plus /I 307 /E set PCRE_DOLLAR_ENDONLY 308 /F flip byte order in compiled pattern 309 /f set PCRE_FIRSTLINE 310 /G find all matches (shorten string) 311 /g find all matches (use startoffset) 312 /I show information about pattern 313 /i set PCRE_CASELESS 314 /J set PCRE_DUPNAMES 315 /K show backtracking control names 316 /L set locale 317 /M show compiled memory size 318 /m set PCRE_MULTILINE 319 /N set PCRE_NO_AUTO_CAPTURE 320 /O set PCRE_NO_AUTO_POSSESS 321 /P use the POSIX wrapper 322 /Q test external stack check function 323 /S study the pattern after compilation 324 /s set PCRE_DOTALL 325 /T select character tables 326 /U set PCRE_UNGREEDY 327 /W set PCRE_UCP 328 /X set PCRE_EXTRA 329 /x set PCRE_EXTENDED 330 /Y set PCRE_NO_START_OPTIMIZE 331 /Z don't show lengths in /B output 332 333 /<any> set PCRE_NEWLINE_ANY 334 /<anycrlf> set PCRE_NEWLINE_ANYCRLF 335 /<cr> set PCRE_NEWLINE_CR 336 /<crlf> set PCRE_NEWLINE_CRLF 337 /<lf> set PCRE_NEWLINE_LF 338 /<bsr_anycrlf> set PCRE_BSR_ANYCRLF 339 /<bsr_unicode> set PCRE_BSR_UNICODE 340 /<JS> set PCRE_JAVASCRIPT_COMPAT 341 342 343 Perl-compatible modifiers 344 345 The /i, /m, /s, and /x modifiers set the PCRE_CASELESS, PCRE_MULTILINE, 346 PCRE_DOTALL, or PCRE_EXTENDED options, respectively, when 347 pcre[16|32]_compile() is called. These four modifier letters have the 348 same effect as they do in Perl. For example: 349 350 /caseless/i 351 352 353 Modifiers for other PCRE options 354 355 The following table shows additional modifiers for setting PCRE com- 356 pile-time options that do not correspond to anything in Perl: 357 358 /8 PCRE_UTF8 ) when using the 8-bit 359 /? PCRE_NO_UTF8_CHECK ) library 360 361 /8 PCRE_UTF16 ) when using the 16-bit 362 /? PCRE_NO_UTF16_CHECK ) library 363 364 /8 PCRE_UTF32 ) when using the 32-bit 365 /? PCRE_NO_UTF32_CHECK ) library 366 367 /9 PCRE_NEVER_UTF 368 /A PCRE_ANCHORED 369 /C PCRE_AUTO_CALLOUT 370 /E PCRE_DOLLAR_ENDONLY 371 /f PCRE_FIRSTLINE 372 /J PCRE_DUPNAMES 373 /N PCRE_NO_AUTO_CAPTURE 374 /O PCRE_NO_AUTO_POSSESS 375 /U PCRE_UNGREEDY 376 /W PCRE_UCP 377 /X PCRE_EXTRA 378 /Y PCRE_NO_START_OPTIMIZE 379 /<any> PCRE_NEWLINE_ANY 380 /<anycrlf> PCRE_NEWLINE_ANYCRLF 381 /<cr> PCRE_NEWLINE_CR 382 /<crlf> PCRE_NEWLINE_CRLF 383 /<lf> PCRE_NEWLINE_LF 384 /<bsr_anycrlf> PCRE_BSR_ANYCRLF 385 /<bsr_unicode> PCRE_BSR_UNICODE 386 /<JS> PCRE_JAVASCRIPT_COMPAT 387 388 The modifiers that are enclosed in angle brackets are literal strings 389 as shown, including the angle brackets, but the letters within can be 390 in either case. This example sets multiline matching with CRLF as the 391 line ending sequence: 392 393 /^abc/m<CRLF> 394 395 As well as turning on the PCRE_UTF8/16/32 option, the /8 modifier 396 causes all non-printing characters in output strings to be printed 397 using the \x{hh...} notation. Otherwise, those less than 0x100 are out- 398 put in hex without the curly brackets. 399 400 Full details of the PCRE options are given in the pcreapi documenta- 401 tion. 402 403 Finding all matches in a string 404 405 Searching for all possible matches within each subject string can be 406 requested by the /g or /G modifier. After finding a match, PCRE is 407 called again to search the remainder of the subject string. The differ- 408 ence between /g and /G is that the former uses the startoffset argument 409 to pcre[16|32]_exec() to start searching at a new point within the 410 entire string (which is in effect what Perl does), whereas the latter 411 passes over a shortened substring. This makes a difference to the 412 matching process if the pattern begins with a lookbehind assertion 413 (including \b or \B). 414 415 If any call to pcre[16|32]_exec() in a /g or /G sequence matches an 416 empty string, the next call is done with the PCRE_NOTEMPTY_ATSTART and 417 PCRE_ANCHORED flags set in order to search for another, non-empty, 418 match at the same point. If this second match fails, the start offset 419 is advanced, and the normal match is retried. This imitates the way 420 Perl handles such cases when using the /g modifier or the split() func- 421 tion. Normally, the start offset is advanced by one character, but if 422 the newline convention recognizes CRLF as a newline, and the current 423 character is CR followed by LF, an advance of two is used. 424 425 Other modifiers 426 427 There are yet more modifiers for controlling the way pcretest operates. 428 429 The /+ modifier requests that as well as outputting the substring that 430 matched the entire pattern, pcretest should in addition output the 431 remainder of the subject string. This is useful for tests where the 432 subject contains multiple copies of the same substring. If the + modi- 433 fier appears twice, the same action is taken for captured substrings. 434 In each case the remainder is output on the following line with a plus 435 character following the capture number. Note that this modifier must 436 not immediately follow the /S modifier because /S+ and /S++ have other 437 meanings. 438 439 The /= modifier requests that the values of all potential captured 440 parentheses be output after a match. By default, only those up to the 441 highest one actually used in the match are output (corresponding to the 442 return code from pcre[16|32]_exec()). Values in the offsets vector cor- 443 responding to higher numbers should be set to -1, and these are output 444 as "<unset>". This modifier gives a way of checking that this is hap- 445 pening. 446 447 The /B modifier is a debugging feature. It requests that pcretest out- 448 put a representation of the compiled code after compilation. Normally 449 this information contains length and offset values; however, if /Z is 450 also present, this data is replaced by spaces. This is a special fea- 451 ture for use in the automatic test scripts; it ensures that the same 452 output is generated for different internal link sizes. 453 454 The /D modifier is a PCRE debugging feature, and is equivalent to /BI, 455 that is, both the /B and the /I modifiers. 456 457 The /F modifier causes pcretest to flip the byte order of the 2-byte 458 and 4-byte fields in the compiled pattern. This facility is for testing 459 the feature in PCRE that allows it to execute patterns that were com- 460 piled on a host with a different endianness. This feature is not avail- 461 able when the POSIX interface to PCRE is being used, that is, when the 462 /P pattern modifier is specified. See also the section about saving and 463 reloading compiled patterns below. 464 465 The /I modifier requests that pcretest output information about the 466 compiled pattern (whether it is anchored, has a fixed first character, 467 and so on). It does this by calling pcre[16|32]_fullinfo() after com- 468 piling a pattern. If the pattern is studied, the results of that are 469 also output. In this output, the word "char" means a non-UTF character, 470 that is, the value of a single data item (8-bit, 16-bit, or 32-bit, 471 depending on the library that is being tested). 472 473 The /K modifier requests pcretest to show names from backtracking con- 474 trol verbs that are returned from calls to pcre[16|32]_exec(). It 475 causes pcretest to create a pcre[16|32]_extra block if one has not 476 already been created by a call to pcre[16|32]_study(), and to set the 477 PCRE_EXTRA_MARK flag and the mark field within it, every time that 478 pcre[16|32]_exec() is called. If the variable that the mark field 479 points to is non-NULL for a match, non-match, or partial match, 480 pcretest prints the string to which it points. For a match, this is 481 shown on a line by itself, tagged with "MK:". For a non-match it is 482 added to the message. 483 484 The /L modifier must be followed directly by the name of a locale, for 485 example, 486 487 /pattern/Lfr_FR 488 489 For this reason, it must be the last modifier. The given locale is set, 490 pcre[16|32]_maketables() is called to build a set of character tables 491 for the locale, and this is then passed to pcre[16|32]_compile() when 492 compiling the regular expression. Without an /L (or /T) modifier, NULL 493 is passed as the tables pointer; that is, /L applies only to the 494 expression on which it appears. 495 496 The /M modifier causes the size in bytes of the memory block used to 497 hold the compiled pattern to be output. This does not include the size 498 of the pcre[16|32] block; it is just the actual compiled data. If the 499 pattern is successfully studied with the PCRE_STUDY_JIT_COMPILE option, 500 the size of the JIT compiled code is also output. 501 502 The /Q modifier is used to test the use of pcre_stack_guard. It must be 503 followed by '0' or '1', specifying the return code to be given from an 504 external function that is passed to PCRE and used for stack checking 505 during compilation (see the pcreapi documentation for details). 506 507 The /S modifier causes pcre[16|32]_study() to be called after the 508 expression has been compiled, and the results used when the expression 509 is matched. There are a number of qualifying characters that may follow 510 /S. They may appear in any order. 511 512 If /S is followed by an exclamation mark, pcre[16|32]_study() is called 513 with the PCRE_STUDY_EXTRA_NEEDED option, causing it always to return a 514 pcre_extra block, even when studying discovers no useful information. 515 516 If /S is followed by a second S character, it suppresses studying, even 517 if it was requested externally by the -s command line option. This 518 makes it possible to specify that certain patterns are always studied, 519 and others are never studied, independently of -s. This feature is used 520 in the test files in a few cases where the output is different when the 521 pattern is studied. 522 523 If the /S modifier is followed by a + character, the call to 524 pcre[16|32]_study() is made with all the JIT study options, requesting 525 just-in-time optimization support if it is available, for both normal 526 and partial matching. If you want to restrict the JIT compiling modes, 527 you can follow /S+ with a digit in the range 1 to 7: 528 529 1 normal match only 530 2 soft partial match only 531 3 normal match and soft partial match 532 4 hard partial match only 533 6 soft and hard partial match 534 7 all three modes (default) 535 536 If /S++ is used instead of /S+ (with or without a following digit), the 537 text "(JIT)" is added to the first output line after a match or no 538 match when JIT-compiled code was actually used. 539 540 Note that there is also an independent /+ modifier; it must not be 541 given immediately after /S or /S+ because this will be misinterpreted. 542 543 If JIT studying is successful, the compiled JIT code will automatically 544 be used when pcre[16|32]_exec() is run, except when incompatible run- 545 time options are specified. For more details, see the pcrejit documen- 546 tation. See also the \J escape sequence below for a way of setting the 547 size of the JIT stack. 548 549 Finally, if /S is followed by a minus character, JIT compilation is 550 suppressed, even if it was requested externally by the -s command line 551 option. This makes it possible to specify that JIT is never to be used 552 for certain patterns. 553 554 The /T modifier must be followed by a single digit. It causes a spe- 555 cific set of built-in character tables to be passed to pcre[16|32]_com- 556 pile(). It is used in the standard PCRE tests to check behaviour with 557 different character tables. The digit specifies the tables as follows: 558 559 0 the default ASCII tables, as distributed in 560 pcre_chartables.c.dist 561 1 a set of tables defining ISO 8859 characters 562 563 In table 1, some characters whose codes are greater than 128 are iden- 564 tified as letters, digits, spaces, etc. 565 566 Using the POSIX wrapper API 567 568 The /P modifier causes pcretest to call PCRE via the POSIX wrapper API 569 rather than its native API. This supports only the 8-bit library. When 570 /P is set, the following modifiers set options for the regcomp() func- 571 tion: 572 573 /i REG_ICASE 574 /m REG_NEWLINE 575 /N REG_NOSUB 576 /s REG_DOTALL ) 577 /U REG_UNGREEDY ) These options are not part of 578 /W REG_UCP ) the POSIX standard 579 /8 REG_UTF8 ) 580 581 The /+ modifier works as described above. All other modifiers are 582 ignored. 583 584 Locking out certain modifiers 585 586 PCRE can be compiled with or without support for certain features such 587 as UTF-8/16/32 or Unicode properties. Accordingly, the standard tests 588 are split up into a number of different files that are selected for 589 running depending on which features are available. When updating the 590 tests, it is all too easy to put a new test into the wrong file by mis- 591 take; for example, to put a test that requires UTF support into a file 592 that is used when it is not available. To help detect such mistakes as 593 early as possible, there is a facility for locking out specific modi- 594 fiers. If an input line for pcretest starts with the string "< forbid " 595 the following sequence of characters is taken as a list of forbidden 596 modifiers. For example, in the test files that must not use UTF or Uni- 597 code property support, this line appears: 598 599 < forbid 8W 600 601 This locks out the /8 and /W modifiers. An immediate error is given if 602 they are subsequently encountered. If the character string contains < 603 but not >, all the multi-character modifiers that begin with < are 604 locked out. Otherwise, such modifiers must be explicitly listed, for 605 example: 606 607 < forbid <JS><cr> 608 609 There must be a single space between < and "forbid" for this feature to 610 be recognised. If there is not, the line is interpreted either as a 611 request to re-load a pre-compiled pattern (see "SAVING AND RELOADING 612 COMPILED PATTERNS" below) or, if there is a another < character, as a 613 pattern that uses < as its delimiter. 614 615 616DATA LINES 617 618 Before each data line is passed to pcre[16|32]_exec(), leading and 619 trailing white space is removed, and it is then scanned for \ escapes. 620 Some of these are pretty esoteric features, intended for checking out 621 some of the more complicated features of PCRE. If you are just testing 622 "ordinary" regular expressions, you probably don't need any of these. 623 The following escapes are recognized: 624 625 \a alarm (BEL, \x07) 626 \b backspace (\x08) 627 \e escape (\x27) 628 \f form feed (\x0c) 629 \n newline (\x0a) 630 \qdd set the PCRE_MATCH_LIMIT limit to dd 631 (any number of digits) 632 \r carriage return (\x0d) 633 \t tab (\x09) 634 \v vertical tab (\x0b) 635 \nnn octal character (up to 3 octal digits); always 636 a byte unless > 255 in UTF-8 or 16-bit or 32-bit mode 637 \o{dd...} octal character (any number of octal digits} 638 \xhh hexadecimal byte (up to 2 hex digits) 639 \x{hh...} hexadecimal character (any number of hex digits) 640 \A pass the PCRE_ANCHORED option to pcre[16|32]_exec() 641 or pcre[16|32]_dfa_exec() 642 \B pass the PCRE_NOTBOL option to pcre[16|32]_exec() 643 or pcre[16|32]_dfa_exec() 644 \Cdd call pcre[16|32]_copy_substring() for substring dd 645 after a successful match (number less than 32) 646 \Cname call pcre[16|32]_copy_named_substring() for substring 647 "name" after a successful match (name termin- 648 ated by next non alphanumeric character) 649 \C+ show the current captured substrings at callout 650 time 651 \C- do not supply a callout function 652 \C!n return 1 instead of 0 when callout number n is 653 reached 654 \C!n!m return 1 instead of 0 when callout number n is 655 reached for the nth time 656 \C*n pass the number n (may be negative) as callout 657 data; this is used as the callout return value 658 \D use the pcre[16|32]_dfa_exec() match function 659 \F only shortest match for pcre[16|32]_dfa_exec() 660 \Gdd call pcre[16|32]_get_substring() for substring dd 661 after a successful match (number less than 32) 662 \Gname call pcre[16|32]_get_named_substring() for substring 663 "name" after a successful match (name termin- 664 ated by next non-alphanumeric character) 665 \Jdd set up a JIT stack of dd kilobytes maximum (any 666 number of digits) 667 \L call pcre[16|32]_get_substringlist() after a 668 successful match 669 \M discover the minimum MATCH_LIMIT and 670 MATCH_LIMIT_RECURSION settings 671 \N pass the PCRE_NOTEMPTY option to pcre[16|32]_exec() 672 or pcre[16|32]_dfa_exec(); if used twice, pass the 673 PCRE_NOTEMPTY_ATSTART option 674 \Odd set the size of the output vector passed to 675 pcre[16|32]_exec() to dd (any number of digits) 676 \P pass the PCRE_PARTIAL_SOFT option to pcre[16|32]_exec() 677 or pcre[16|32]_dfa_exec(); if used twice, pass the 678 PCRE_PARTIAL_HARD option 679 \Qdd set the PCRE_MATCH_LIMIT_RECURSION limit to dd 680 (any number of digits) 681 \R pass the PCRE_DFA_RESTART option to pcre[16|32]_dfa_exec() 682 \S output details of memory get/free calls during matching 683 \Y pass the PCRE_NO_START_OPTIMIZE option to 684 pcre[16|32]_exec() 685 or pcre[16|32]_dfa_exec() 686 \Z pass the PCRE_NOTEOL option to pcre[16|32]_exec() 687 or pcre[16|32]_dfa_exec() 688 \? pass the PCRE_NO_UTF[8|16|32]_CHECK option to 689 pcre[16|32]_exec() or pcre[16|32]_dfa_exec() 690 \>dd start the match at offset dd (optional "-"; then 691 any number of digits); this sets the startoffset 692 argument for pcre[16|32]_exec() or 693 pcre[16|32]_dfa_exec() 694 \<cr> pass the PCRE_NEWLINE_CR option to pcre[16|32]_exec() 695 or pcre[16|32]_dfa_exec() 696 \<lf> pass the PCRE_NEWLINE_LF option to pcre[16|32]_exec() 697 or pcre[16|32]_dfa_exec() 698 \<crlf> pass the PCRE_NEWLINE_CRLF option to pcre[16|32]_exec() 699 or pcre[16|32]_dfa_exec() 700 \<anycrlf> pass the PCRE_NEWLINE_ANYCRLF option to pcre[16|32]_exec() 701 or pcre[16|32]_dfa_exec() 702 \<any> pass the PCRE_NEWLINE_ANY option to pcre[16|32]_exec() 703 or pcre[16|32]_dfa_exec() 704 705 The use of \x{hh...} is not dependent on the use of the /8 modifier on 706 the pattern. It is recognized always. There may be any number of hexa- 707 decimal digits inside the braces; invalid values provoke error mes- 708 sages. 709 710 Note that \xhh specifies one byte rather than one character in UTF-8 711 mode; this makes it possible to construct invalid UTF-8 sequences for 712 testing purposes. On the other hand, \x{hh} is interpreted as a UTF-8 713 character in UTF-8 mode, generating more than one byte if the value is 714 greater than 127. When testing the 8-bit library not in UTF-8 mode, 715 \x{hh} generates one byte for values less than 256, and causes an error 716 for greater values. 717 718 In UTF-16 mode, all 4-digit \x{hhhh} values are accepted. This makes it 719 possible to construct invalid UTF-16 sequences for testing purposes. 720 721 In UTF-32 mode, all 4- to 8-digit \x{...} values are accepted. This 722 makes it possible to construct invalid UTF-32 sequences for testing 723 purposes. 724 725 The escapes that specify line ending sequences are literal strings, 726 exactly as shown. No more than one newline setting should be present in 727 any data line. 728 729 A backslash followed by anything else just escapes the anything else. 730 If the very last character is a backslash, it is ignored. This gives a 731 way of passing an empty line as data, since a real empty line termi- 732 nates the data input. 733 734 The \J escape provides a way of setting the maximum stack size that is 735 used by the just-in-time optimization code. It is ignored if JIT opti- 736 mization is not being used. Providing a stack that is larger than the 737 default 32K is necessary only for very complicated patterns. 738 739 If \M is present, pcretest calls pcre[16|32]_exec() several times, with 740 different values in the match_limit and match_limit_recursion fields of 741 the pcre[16|32]_extra data structure, until it finds the minimum num- 742 bers for each parameter that allow pcre[16|32]_exec() to complete with- 743 out error. Because this is testing a specific feature of the normal 744 interpretive pcre[16|32]_exec() execution, the use of any JIT optimiza- 745 tion that might have been set up by the /S+ qualifier of -s+ option is 746 disabled. 747 748 The match_limit number is a measure of the amount of backtracking that 749 takes place, and checking it out can be instructive. For most simple 750 matches, the number is quite small, but for patterns with very large 751 numbers of matching possibilities, it can become large very quickly 752 with increasing length of subject string. The match_limit_recursion 753 number is a measure of how much stack (or, if PCRE is compiled with 754 NO_RECURSE, how much heap) memory is needed to complete the match 755 attempt. 756 757 When \O is used, the value specified may be higher or lower than the 758 size set by the -O command line option (or defaulted to 45); \O applies 759 only to the call of pcre[16|32]_exec() for the line in which it 760 appears. 761 762 If the /P modifier was present on the pattern, causing the POSIX wrap- 763 per API to be used, the only option-setting sequences that have any 764 effect are \B, \N, and \Z, causing REG_NOTBOL, REG_NOTEMPTY, and 765 REG_NOTEOL, respectively, to be passed to regexec(). 766 767 768THE ALTERNATIVE MATCHING FUNCTION 769 770 By default, pcretest uses the standard PCRE matching function, 771 pcre[16|32]_exec() to match each data line. PCRE also supports an 772 alternative matching function, pcre[16|32]_dfa_test(), which operates 773 in a different way, and has some restrictions. The differences between 774 the two functions are described in the pcrematching documentation. 775 776 If a data line contains the \D escape sequence, or if the command line 777 contains the -dfa option, the alternative matching function is used. 778 This function finds all possible matches at a given point. If, however, 779 the \F escape sequence is present in the data line, it stops after the 780 first match is found. This is always the shortest possible match. 781 782 783DEFAULT OUTPUT FROM PCRETEST 784 785 This section describes the output when the normal matching function, 786 pcre[16|32]_exec(), is being used. 787 788 When a match succeeds, pcretest outputs the list of captured substrings 789 that pcre[16|32]_exec() returns, starting with number 0 for the string 790 that matched the whole pattern. Otherwise, it outputs "No match" when 791 the return is PCRE_ERROR_NOMATCH, and "Partial match:" followed by the 792 partially matching substring when pcre[16|32]_exec() returns 793 PCRE_ERROR_PARTIAL. (Note that this is the entire substring that was 794 inspected during the partial match; it may include characters before 795 the actual match start if a lookbehind assertion, \K, \b, or \B was 796 involved.) For any other return, pcretest outputs the PCRE negative 797 error number and a short descriptive phrase. If the error is a failed 798 UTF string check, the offset of the start of the failing character and 799 the reason code are also output, provided that the size of the output 800 vector is at least two. Here is an example of an interactive pcretest 801 run. 802 803 $ pcretest 804 PCRE version 8.13 2011-04-30 805 806 re> /^abc(\d+)/ 807 data> abc123 808 0: abc123 809 1: 123 810 data> xyz 811 No match 812 813 Unset capturing substrings that are not followed by one that is set are 814 not returned by pcre[16|32]_exec(), and are not shown by pcretest. In 815 the following example, there are two capturing substrings, but when the 816 first data line is matched, the second, unset substring is not shown. 817 An "internal" unset substring is shown as "<unset>", as for the second 818 data line. 819 820 re> /(a)|(b)/ 821 data> a 822 0: a 823 1: a 824 data> b 825 0: b 826 1: <unset> 827 2: b 828 829 If the strings contain any non-printing characters, they are output as 830 \xhh escapes if the value is less than 256 and UTF mode is not set. 831 Otherwise they are output as \x{hh...} escapes. See below for the defi- 832 nition of non-printing characters. If the pattern has the /+ modifier, 833 the output for substring 0 is followed by the the rest of the subject 834 string, identified by "0+" like this: 835 836 re> /cat/+ 837 data> cataract 838 0: cat 839 0+ aract 840 841 If the pattern has the /g or /G modifier, the results of successive 842 matching attempts are output in sequence, like this: 843 844 re> /\Bi(\w\w)/g 845 data> Mississippi 846 0: iss 847 1: ss 848 0: iss 849 1: ss 850 0: ipp 851 1: pp 852 853 "No match" is output only if the first match attempt fails. Here is an 854 example of a failure message (the offset 4 that is specified by \>4 is 855 past the end of the subject string): 856 857 re> /xyz/ 858 data> xyz\>4 859 Error -24 (bad offset value) 860 861 If any of the sequences \C, \G, or \L are present in a data line that 862 is successfully matched, the substrings extracted by the convenience 863 functions are output with C, G, or L after the string number instead of 864 a colon. This is in addition to the normal full list. The string length 865 (that is, the return from the extraction function) is given in paren- 866 theses after each string for \C and \G. 867 868 Note that whereas patterns can be continued over several lines (a plain 869 ">" prompt is used for continuations), data lines may not. However new- 870 lines can be included in data by means of the \n escape (or \r, \r\n, 871 etc., depending on the newline sequence setting). 872 873 874OUTPUT FROM THE ALTERNATIVE MATCHING FUNCTION 875 876 When the alternative matching function, pcre[16|32]_dfa_exec(), is used 877 (by means of the \D escape sequence or the -dfa command line option), 878 the output consists of a list of all the matches that start at the 879 first point in the subject where there is at least one match. For exam- 880 ple: 881 882 re> /(tang|tangerine|tan)/ 883 data> yellow tangerine\D 884 0: tangerine 885 1: tang 886 2: tan 887 888 (Using the normal matching function on this data finds only "tang".) 889 The longest matching string is always given first (and numbered zero). 890 After a PCRE_ERROR_PARTIAL return, the output is "Partial match:", fol- 891 lowed by the partially matching substring. (Note that this is the 892 entire substring that was inspected during the partial match; it may 893 include characters before the actual match start if a lookbehind asser- 894 tion, \K, \b, or \B was involved.) 895 896 If /g is present on the pattern, the search for further matches resumes 897 at the end of the longest match. For example: 898 899 re> /(tang|tangerine|tan)/g 900 data> yellow tangerine and tangy sultana\D 901 0: tangerine 902 1: tang 903 2: tan 904 0: tang 905 1: tan 906 0: tan 907 908 Since the matching function does not support substring capture, the 909 escape sequences that are concerned with captured substrings are not 910 relevant. 911 912 913RESTARTING AFTER A PARTIAL MATCH 914 915 When the alternative matching function has given the PCRE_ERROR_PARTIAL 916 return, indicating that the subject partially matched the pattern, you 917 can restart the match with additional subject data by means of the \R 918 escape sequence. For example: 919 920 re> /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/ 921 data> 23ja\P\D 922 Partial match: 23ja 923 data> n05\R\D 924 0: n05 925 926 For further information about partial matching, see the pcrepartial 927 documentation. 928 929 930CALLOUTS 931 932 If the pattern contains any callout requests, pcretest's callout func- 933 tion is called during matching. This works with both matching func- 934 tions. By default, the called function displays the callout number, the 935 start and current positions in the text at the callout time, and the 936 next pattern item to be tested. For example: 937 938 --->pqrabcdef 939 0 ^ ^ \d 940 941 This output indicates that callout number 0 occurred for a match 942 attempt starting at the fourth character of the subject string, when 943 the pointer was at the seventh character of the data, and when the next 944 pattern item was \d. Just one circumflex is output if the start and 945 current positions are the same. 946 947 Callouts numbered 255 are assumed to be automatic callouts, inserted as 948 a result of the /C pattern modifier. In this case, instead of showing 949 the callout number, the offset in the pattern, preceded by a plus, is 950 output. For example: 951 952 re> /\d?[A-E]\*/C 953 data> E* 954 --->E* 955 +0 ^ \d? 956 +3 ^ [A-E] 957 +8 ^^ \* 958 +10 ^ ^ 959 0: E* 960 961 If a pattern contains (*MARK) items, an additional line is output when- 962 ever a change of latest mark is passed to the callout function. For 963 example: 964 965 re> /a(*MARK:X)bc/C 966 data> abc 967 --->abc 968 +0 ^ a 969 +1 ^^ (*MARK:X) 970 +10 ^^ b 971 Latest Mark: X 972 +11 ^ ^ c 973 +12 ^ ^ 974 0: abc 975 976 The mark changes between matching "a" and "b", but stays the same for 977 the rest of the match, so nothing more is output. If, as a result of 978 backtracking, the mark reverts to being unset, the text "<unset>" is 979 output. 980 981 The callout function in pcretest returns zero (carry on matching) by 982 default, but you can use a \C item in a data line (as described above) 983 to change this and other parameters of the callout. 984 985 Inserting callouts can be helpful when using pcretest to check compli- 986 cated regular expressions. For further information about callouts, see 987 the pcrecallout documentation. 988 989 990NON-PRINTING CHARACTERS 991 992 When pcretest is outputting text in the compiled version of a pattern, 993 bytes other than 32-126 are always treated as non-printing characters 994 are are therefore shown as hex escapes. 995 996 When pcretest is outputting text that is a matched part of a subject 997 string, it behaves in the same way, unless a different locale has been 998 set for the pattern (using the /L modifier). In this case, the 999 isprint() function to distinguish printing and non-printing characters. 1000 1001 1002SAVING AND RELOADING COMPILED PATTERNS 1003 1004 The facilities described in this section are not available when the 1005 POSIX interface to PCRE is being used, that is, when the /P pattern 1006 modifier is specified. 1007 1008 When the POSIX interface is not in use, you can cause pcretest to write 1009 a compiled pattern to a file, by following the modifiers with > and a 1010 file name. For example: 1011 1012 /pattern/im >/some/file 1013 1014 See the pcreprecompile documentation for a discussion about saving and 1015 re-using compiled patterns. Note that if the pattern was successfully 1016 studied with JIT optimization, the JIT data cannot be saved. 1017 1018 The data that is written is binary. The first eight bytes are the 1019 length of the compiled pattern data followed by the length of the 1020 optional study data, each written as four bytes in big-endian order 1021 (most significant byte first). If there is no study data (either the 1022 pattern was not studied, or studying did not return any data), the sec- 1023 ond length is zero. The lengths are followed by an exact copy of the 1024 compiled pattern. If there is additional study data, this (excluding 1025 any JIT data) follows immediately after the compiled pattern. After 1026 writing the file, pcretest expects to read a new pattern. 1027 1028 A saved pattern can be reloaded into pcretest by specifying < and a 1029 file name instead of a pattern. There must be no space between < and 1030 the file name, which must not contain a < character, as otherwise 1031 pcretest will interpret the line as a pattern delimited by < charac- 1032 ters. For example: 1033 1034 re> </some/file 1035 Compiled pattern loaded from /some/file 1036 No study data 1037 1038 If the pattern was previously studied with the JIT optimization, the 1039 JIT information cannot be saved and restored, and so is lost. When the 1040 pattern has been loaded, pcretest proceeds to read data lines in the 1041 usual way. 1042 1043 You can copy a file written by pcretest to a different host and reload 1044 it there, even if the new host has opposite endianness to the one on 1045 which the pattern was compiled. For example, you can compile on an i86 1046 machine and run on a SPARC machine. When a pattern is reloaded on a 1047 host with different endianness, the confirmation message is changed to: 1048 1049 Compiled pattern (byte-inverted) loaded from /some/file 1050 1051 The test suite contains some saved pre-compiled patterns with different 1052 endianness. These are reloaded using "<!" instead of just "<". This 1053 suppresses the "(byte-inverted)" text so that the output is the same on 1054 all hosts. It also forces debugging output once the pattern has been 1055 reloaded. 1056 1057 File names for saving and reloading can be absolute or relative, but 1058 note that the shell facility of expanding a file name that starts with 1059 a tilde (~) is not available. 1060 1061 The ability to save and reload files in pcretest is intended for test- 1062 ing and experimentation. It is not intended for production use because 1063 only a single pattern can be written to a file. Furthermore, there is 1064 no facility for supplying custom character tables for use with a 1065 reloaded pattern. If the original pattern was compiled with custom 1066 tables, an attempt to match a subject string using a reloaded pattern 1067 is likely to cause pcretest to crash. Finally, if you attempt to load 1068 a file that is not in the correct format, the result is undefined. 1069 1070 1071SEE ALSO 1072 1073 pcre(3), pcre16(3), pcre32(3), pcreapi(3), pcrecallout(3), pcrejit, 1074 pcrematching(3), pcrepartial(d), pcrepattern(3), pcreprecompile(3). 1075 1076 1077AUTHOR 1078 1079 Philip Hazel 1080 University Computing Service 1081 Cambridge CB2 3QH, England. 1082 1083 1084REVISION 1085 1086 Last updated: 09 February 2014 1087 Copyright (c) 1997-2014 University of Cambridge. 1088