Tristan Matthews | 0461646 | 2013-11-14 16:09:34 -0500 | [diff] [blame] | 1 | PCRETEST(1) PCRETEST(1) |
| 2 | |
| 3 | |
| 4 | NAME |
| 5 | pcretest - a program for testing Perl-compatible regular expressions. |
| 6 | |
| 7 | |
| 8 | SYNOPSIS |
| 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 documentation. The input for pcretest is a |
| 18 | sequence of regular expression patterns and strings to be matched, as |
| 19 | described below. The output shows the result of each match. Options on |
| 20 | the command line and the patterns control PCRE options and exactly what |
| 21 | is output. |
| 22 | |
| 23 | |
| 24 | COMMAND LINE OPTIONS |
| 25 | |
| 26 | -b Behave as if each pattern has the /B (show byte code) modi- |
| 27 | fier; the internal form is output after compilation. |
| 28 | |
| 29 | -C Output the version number of the PCRE library, and all avail- |
| 30 | able information about the optional features that are |
| 31 | included, and then exit. |
| 32 | |
| 33 | -d Behave as if each pattern has the /D (debug) modifier; the |
| 34 | internal form and information about the compiled pattern is |
| 35 | output after compilation; -d is equivalent to -b -i. |
| 36 | |
| 37 | -dfa Behave as if each data line contains the \D escape sequence; |
| 38 | this causes the alternative matching function, |
| 39 | pcre_dfa_exec(), to be used instead of the standard |
| 40 | pcre_exec() function (more detail is given below). |
| 41 | |
| 42 | -help Output a brief summary these options and then exit. |
| 43 | |
| 44 | -i Behave as if each pattern has the /I modifier; information |
| 45 | about the compiled pattern is given after compilation. |
| 46 | |
| 47 | -M Behave as if each data line contains the \M escape sequence; |
| 48 | this causes PCRE to discover the minimum MATCH_LIMIT and |
| 49 | MATCH_LIMIT_RECURSION settings by calling pcre_exec() repeat- |
| 50 | edly with different limits. |
| 51 | |
| 52 | -m Output the size of each compiled pattern after it has been |
| 53 | compiled. This is equivalent to adding /M to each regular |
| 54 | expression. |
| 55 | |
| 56 | -o osize Set the number of elements in the output vector that is used |
| 57 | when calling pcre_exec() or pcre_dfa_exec() to be osize. The |
| 58 | default value is 45, which is enough for 14 capturing subex- |
| 59 | pressions for pcre_exec() or 22 different matches for |
| 60 | pcre_dfa_exec(). The vector size can be changed for individ- |
| 61 | ual matching calls by including \O in the data line (see |
| 62 | below). |
| 63 | |
| 64 | -p Behave as if each pattern has the /P modifier; the POSIX |
| 65 | wrapper API is used to call PCRE. None of the other options |
| 66 | has any effect when -p is set. |
| 67 | |
| 68 | -q Do not output the version number of pcretest at the start of |
| 69 | execution. |
| 70 | |
| 71 | -S size On Unix-like systems, set the size of the run-time stack to |
| 72 | size megabytes. |
| 73 | |
| 74 | -s or -s+ Behave as if each pattern has the /S modifier; in other |
| 75 | words, force each pattern to be studied. If -s+ is used, the |
| 76 | PCRE_STUDY_JIT_COMPILE flag is passed to pcre_study(), caus- |
| 77 | ing just-in-time optimization to be set up if it is avail- |
| 78 | able. If the /I or /D option is present on a pattern |
| 79 | (requesting output about the compiled pattern), information |
| 80 | about the result of studying is not included when studying is |
| 81 | caused only by -s and neither -i nor -d is present on the |
| 82 | command line. This behaviour means that the output from tests |
| 83 | that are run with and without -s should be identical, except |
| 84 | when options that output information about the actual running |
| 85 | of a match are set. The -M, -t, and -tm options, which give |
| 86 | information about resources used, are likely to produce dif- |
| 87 | ferent output with and without -s. Output may also differ if |
| 88 | the /C option is present on an individual pattern. This uses |
| 89 | callouts to trace the the matching process, and this may be |
| 90 | different between studied and non-studied patterns. If the |
| 91 | pattern contains (*MARK) items there may also be differences, |
| 92 | for the same reason. The -s command line option can be over- |
| 93 | ridden for specific patterns that should never be studied |
| 94 | (see the /S pattern modifier below). |
| 95 | |
| 96 | -t Run each compile, study, and match many times with a timer, |
| 97 | and output resulting time per compile or match (in millisec- |
| 98 | onds). Do not set -m with -t, because you will then get the |
| 99 | size output a zillion times, and the timing will be dis- |
| 100 | torted. You can control the number of iterations that are |
| 101 | used for timing by following -t with a number (as a separate |
| 102 | item on the command line). For example, "-t 1000" would iter- |
| 103 | ate 1000 times. The default is to iterate 500000 times. |
| 104 | |
| 105 | -tm This is like -t except that it times only the matching phase, |
| 106 | not the compile or study phases. |
| 107 | |
| 108 | |
| 109 | DESCRIPTION |
| 110 | |
| 111 | If pcretest is given two filename arguments, it reads from the first |
| 112 | and writes to the second. If it is given only one filename argument, it |
| 113 | reads from that file and writes to stdout. Otherwise, it reads from |
| 114 | stdin and writes to stdout, and prompts for each line of input, using |
| 115 | "re>" to prompt for regular expressions, and "data>" to prompt for data |
| 116 | lines. |
| 117 | |
| 118 | When pcretest is built, a configuration option can specify that it |
| 119 | should be linked with the libreadline library. When this is done, if |
| 120 | the input is from a terminal, it is read using the readline() function. |
| 121 | This provides line-editing and history facilities. The output from the |
| 122 | -help option states whether or not readline() will be used. |
| 123 | |
| 124 | The program handles any number of sets of input on a single input file. |
| 125 | Each set starts with a regular expression, and continues with any num- |
| 126 | ber of data lines to be matched against the pattern. |
| 127 | |
| 128 | Each data line is matched separately and independently. If you want to |
| 129 | do multi-line matches, you have to use the \n escape sequence (or \r or |
| 130 | \r\n, etc., depending on the newline setting) in a single line of input |
| 131 | to encode the newline sequences. There is no limit on the length of |
| 132 | data lines; the input buffer is automatically extended if it is too |
| 133 | small. |
| 134 | |
| 135 | An empty line signals the end of the data lines, at which point a new |
| 136 | regular expression is read. The regular expressions are given enclosed |
| 137 | in any non-alphanumeric delimiters other than backslash, for example: |
| 138 | |
| 139 | /(a|bc)x+yz/ |
| 140 | |
| 141 | White space before the initial delimiter is ignored. A regular expres- |
| 142 | sion may be continued over several input lines, in which case the new- |
| 143 | line characters are included within it. It is possible to include the |
| 144 | delimiter within the pattern by escaping it, for example |
| 145 | |
| 146 | /abc\/def/ |
| 147 | |
| 148 | If you do so, the escape and the delimiter form part of the pattern, |
| 149 | but since delimiters are always non-alphanumeric, this does not affect |
| 150 | its interpretation. If the terminating delimiter is immediately fol- |
| 151 | lowed by a backslash, for example, |
| 152 | |
| 153 | /abc/\ |
| 154 | |
| 155 | then a backslash is added to the end of the pattern. This is done to |
| 156 | provide a way of testing the error condition that arises if a pattern |
| 157 | finishes with a backslash, because |
| 158 | |
| 159 | /abc\/ |
| 160 | |
| 161 | is interpreted as the first line of a pattern that starts with "abc/", |
| 162 | causing pcretest to read the next line as a continuation of the regular |
| 163 | expression. |
| 164 | |
| 165 | |
| 166 | PATTERN MODIFIERS |
| 167 | |
| 168 | A pattern may be followed by any number of modifiers, which are mostly |
| 169 | single characters. Following Perl usage, these are referred to below |
| 170 | as, for example, "the /i modifier", even though the delimiter of the |
| 171 | pattern need not always be a slash, and no slash is used when writing |
| 172 | modifiers. White space may appear between the final pattern delimiter |
| 173 | and the first modifier, and between the modifiers themselves. |
| 174 | |
| 175 | The /i, /m, /s, and /x modifiers set the PCRE_CASELESS, PCRE_MULTILINE, |
| 176 | PCRE_DOTALL, or PCRE_EXTENDED options, respectively, when pcre_com- |
| 177 | pile() is called. These four modifier letters have the same effect as |
| 178 | they do in Perl. For example: |
| 179 | |
| 180 | /caseless/i |
| 181 | |
| 182 | The following table shows additional modifiers for setting PCRE com- |
| 183 | pile-time options that do not correspond to anything in Perl: |
| 184 | |
| 185 | /8 PCRE_UTF8 |
| 186 | /? PCRE_NO_UTF8_CHECK |
| 187 | /A PCRE_ANCHORED |
| 188 | /C PCRE_AUTO_CALLOUT |
| 189 | /E PCRE_DOLLAR_ENDONLY |
| 190 | /f PCRE_FIRSTLINE |
| 191 | /J PCRE_DUPNAMES |
| 192 | /N PCRE_NO_AUTO_CAPTURE |
| 193 | /U PCRE_UNGREEDY |
| 194 | /W PCRE_UCP |
| 195 | /X PCRE_EXTRA |
| 196 | /Y PCRE_NO_START_OPTIMIZE |
| 197 | /<JS> PCRE_JAVASCRIPT_COMPAT |
| 198 | /<cr> PCRE_NEWLINE_CR |
| 199 | /<lf> PCRE_NEWLINE_LF |
| 200 | /<crlf> PCRE_NEWLINE_CRLF |
| 201 | /<anycrlf> PCRE_NEWLINE_ANYCRLF |
| 202 | /<any> PCRE_NEWLINE_ANY |
| 203 | /<bsr_anycrlf> PCRE_BSR_ANYCRLF |
| 204 | /<bsr_unicode> PCRE_BSR_UNICODE |
| 205 | |
| 206 | The modifiers that are enclosed in angle brackets are literal strings |
| 207 | as shown, including the angle brackets, but the letters within can be |
| 208 | in either case. This example sets multiline matching with CRLF as the |
| 209 | line ending sequence: |
| 210 | |
| 211 | /^abc/m<CRLF> |
| 212 | |
| 213 | As well as turning on the PCRE_UTF8 option, the /8 modifier also causes |
| 214 | any non-printing characters in output strings to be printed using the |
| 215 | \x{hh...} notation if they are valid UTF-8 sequences. Full details of |
| 216 | the PCRE options are given in the pcreapi documentation. |
| 217 | |
| 218 | Finding all matches in a string |
| 219 | |
| 220 | Searching for all possible matches within each subject string can be |
| 221 | requested by the /g or /G modifier. After finding a match, PCRE is |
| 222 | called again to search the remainder of the subject string. The differ- |
| 223 | ence between /g and /G is that the former uses the startoffset argument |
| 224 | to pcre_exec() to start searching at a new point within the entire |
| 225 | string (which is in effect what Perl does), whereas the latter passes |
| 226 | over a shortened substring. This makes a difference to the matching |
| 227 | process if the pattern begins with a lookbehind assertion (including \b |
| 228 | or \B). |
| 229 | |
| 230 | If any call to pcre_exec() in a /g or /G sequence matches an empty |
| 231 | string, the next call is done with the PCRE_NOTEMPTY_ATSTART and |
| 232 | PCRE_ANCHORED flags set in order to search for another, non-empty, |
| 233 | match at the same point. If this second match fails, the start offset |
| 234 | is advanced, and the normal match is retried. This imitates the way |
| 235 | Perl handles such cases when using the /g modifier or the split() func- |
| 236 | tion. Normally, the start offset is advanced by one character, but if |
| 237 | the newline convention recognizes CRLF as a newline, and the current |
| 238 | character is CR followed by LF, an advance of two is used. |
| 239 | |
| 240 | Other modifiers |
| 241 | |
| 242 | There are yet more modifiers for controlling the way pcretest operates. |
| 243 | |
| 244 | The /+ modifier requests that as well as outputting the substring that |
| 245 | matched the entire pattern, pcretest should in addition output the |
| 246 | remainder of the subject string. This is useful for tests where the |
| 247 | subject contains multiple copies of the same substring. If the + modi- |
| 248 | fier appears twice, the same action is taken for captured substrings. |
| 249 | In each case the remainder is output on the following line with a plus |
| 250 | character following the capture number. Note that this modifier must |
| 251 | not immediately follow the /S modifier because /S+ has another meaning. |
| 252 | |
| 253 | The /= modifier requests that the values of all potential captured |
| 254 | parentheses be output after a match by pcre_exec(). By default, only |
| 255 | those up to the highest one actually used in the match are output (cor- |
| 256 | responding to the return code from pcre_exec()). Values in the offsets |
| 257 | vector corresponding to higher numbers should be set to -1, and these |
| 258 | are output as "<unset>". This modifier gives a way of checking that |
| 259 | this is happening. |
| 260 | |
| 261 | The /B modifier is a debugging feature. It requests that pcretest out- |
| 262 | put a representation of the compiled byte code after compilation. Nor- |
| 263 | mally this information contains length and offset values; however, if |
| 264 | /Z is also present, this data is replaced by spaces. This is a special |
| 265 | feature for use in the automatic test scripts; it ensures that the same |
| 266 | output is generated for different internal link sizes. |
| 267 | |
| 268 | The /D modifier is a PCRE debugging feature, and is equivalent to /BI, |
| 269 | that is, both the /B and the /I modifiers. |
| 270 | |
| 271 | The /F modifier causes pcretest to flip the byte order of the fields in |
| 272 | the compiled pattern that contain 2-byte and 4-byte numbers. This |
| 273 | facility is for testing the feature in PCRE that allows it to execute |
| 274 | patterns that were compiled on a host with a different endianness. This |
| 275 | feature is not available when the POSIX interface to PCRE is being |
| 276 | used, that is, when the /P pattern modifier is specified. See also the |
| 277 | section about saving and reloading compiled patterns below. |
| 278 | |
| 279 | The /I modifier requests that pcretest output information about the |
| 280 | compiled pattern (whether it is anchored, has a fixed first character, |
| 281 | and so on). It does this by calling pcre_fullinfo() after compiling a |
| 282 | pattern. If the pattern is studied, the results of that are also out- |
| 283 | put. |
| 284 | |
| 285 | The /K modifier requests pcretest to show names from backtracking con- |
| 286 | trol verbs that are returned from calls to pcre_exec(). It causes |
| 287 | pcretest to create a pcre_extra block if one has not already been cre- |
| 288 | ated by a call to pcre_study(), and to set the PCRE_EXTRA_MARK flag and |
| 289 | the mark field within it, every time that pcre_exec() is called. If the |
| 290 | variable that the mark field points to is non-NULL for a match, non- |
| 291 | match, or partial match, pcretest prints the string to which it points. |
| 292 | For a match, this is shown on a line by itself, tagged with "MK:". For |
| 293 | a non-match it is added to the message. |
| 294 | |
| 295 | The /L modifier must be followed directly by the name of a locale, for |
| 296 | example, |
| 297 | |
| 298 | /pattern/Lfr_FR |
| 299 | |
| 300 | For this reason, it must be the last modifier. The given locale is set, |
| 301 | pcre_maketables() is called to build a set of character tables for the |
| 302 | locale, and this is then passed to pcre_compile() when compiling the |
| 303 | regular expression. Without an /L (or /T) modifier, NULL is passed as |
| 304 | the tables pointer; that is, /L applies only to the expression on which |
| 305 | it appears. |
| 306 | |
| 307 | The /M modifier causes the size of memory block used to hold the com- |
| 308 | piled pattern to be output. This does not include the size of the pcre |
| 309 | block; it is just the actual compiled data. If the pattern is success- |
| 310 | fully studied with the PCRE_STUDY_JIT_COMPILE option, the size of the |
| 311 | JIT compiled code is also output. |
| 312 | |
| 313 | If the /S modifier appears once, it causes pcre_study() to be called |
| 314 | after the expression has been compiled, and the results used when the |
| 315 | expression is matched. If /S appears twice, it suppresses studying, |
| 316 | even if it was requested externally by the -s command line option. This |
| 317 | makes it possible to specify that certain patterns are always studied, |
| 318 | and others are never studied, independently of -s. This feature is used |
| 319 | in the test files in a few cases where the output is different when the |
| 320 | pattern is studied. |
| 321 | |
| 322 | If the /S modifier is immediately followed by a + character, the call |
| 323 | to pcre_study() is made with the PCRE_STUDY_JIT_COMPILE option, |
| 324 | requesting just-in-time optimization support if it is available. Note |
| 325 | that there is also a /+ modifier; it must not be given immediately |
| 326 | after /S because this will be misinterpreted. If JIT studying is suc- |
| 327 | cessful, it will automatically be used when pcre_exec() is run, except |
| 328 | when incompatible run-time options are specified. These include the |
| 329 | partial matching options; a complete list is given in the pcrejit docu- |
| 330 | mentation. See also the \J escape sequence below for a way of setting |
| 331 | the size of the JIT stack. |
| 332 | |
| 333 | The /T modifier must be followed by a single digit. It causes a spe- |
| 334 | cific set of built-in character tables to be passed to pcre_compile(). |
| 335 | It is used in the standard PCRE tests to check behaviour with different |
| 336 | character tables. The digit specifies the tables as follows: |
| 337 | |
| 338 | 0 the default ASCII tables, as distributed in |
| 339 | pcre_chartables.c.dist |
| 340 | 1 a set of tables defining ISO 8859 characters |
| 341 | |
| 342 | In table 1, some characters whose codes are greater than 128 are iden- |
| 343 | tified as letters, digits, spaces, etc. |
| 344 | |
| 345 | Using the POSIX wrapper API |
| 346 | |
| 347 | The /P modifier causes pcretest to call PCRE via the POSIX wrapper API |
| 348 | rather than its native API. When /P is set, the following modifiers set |
| 349 | options for the regcomp() function: |
| 350 | |
| 351 | /i REG_ICASE |
| 352 | /m REG_NEWLINE |
| 353 | /N REG_NOSUB |
| 354 | /s REG_DOTALL ) |
| 355 | /U REG_UNGREEDY ) These options are not part of |
| 356 | /W REG_UCP ) the POSIX standard |
| 357 | /8 REG_UTF8 ) |
| 358 | |
| 359 | The /+ modifier works as described above. All other modifiers are |
| 360 | ignored. |
| 361 | |
| 362 | |
| 363 | DATA LINES |
| 364 | |
| 365 | Before each data line is passed to pcre_exec(), leading and trailing |
| 366 | white space is removed, and it is then scanned for \ escapes. Some of |
| 367 | these are pretty esoteric features, intended for checking out some of |
| 368 | the more complicated features of PCRE. If you are just testing "ordi- |
| 369 | nary" regular expressions, you probably don't need any of these. The |
| 370 | following escapes are recognized: |
| 371 | |
| 372 | \a alarm (BEL, \x07) |
| 373 | \b backspace (\x08) |
| 374 | \e escape (\x27) |
| 375 | \f form feed (\x0c) |
| 376 | \n newline (\x0a) |
| 377 | \qdd set the PCRE_MATCH_LIMIT limit to dd |
| 378 | (any number of digits) |
| 379 | \r carriage return (\x0d) |
| 380 | \t tab (\x09) |
| 381 | \v vertical tab (\x0b) |
| 382 | \nnn octal character (up to 3 octal digits) |
| 383 | always a byte unless > 255 in UTF-8 mode |
| 384 | \xhh hexadecimal byte (up to 2 hex digits) |
| 385 | \x{hh...} hexadecimal character, any number of digits |
| 386 | in UTF-8 mode |
| 387 | \A pass the PCRE_ANCHORED option to pcre_exec() |
| 388 | or pcre_dfa_exec() |
| 389 | \B pass the PCRE_NOTBOL option to pcre_exec() |
| 390 | or pcre_dfa_exec() |
| 391 | \Cdd call pcre_copy_substring() for substring dd |
| 392 | after a successful match (number less than 32) |
| 393 | \Cname call pcre_copy_named_substring() for substring |
| 394 | "name" after a successful match (name termin- |
| 395 | ated by next non alphanumeric character) |
| 396 | \C+ show the current captured substrings at callout |
| 397 | time |
| 398 | \C- do not supply a callout function |
| 399 | \C!n return 1 instead of 0 when callout number n is |
| 400 | reached |
| 401 | \C!n!m return 1 instead of 0 when callout number n is |
| 402 | reached for the nth time |
| 403 | \C*n pass the number n (may be negative) as callout |
| 404 | data; this is used as the callout return value |
| 405 | \D use the pcre_dfa_exec() match function |
| 406 | \F only shortest match for pcre_dfa_exec() |
| 407 | \Gdd call pcre_get_substring() for substring dd |
| 408 | after a successful match (number less than 32) |
| 409 | \Gname call pcre_get_named_substring() for substring |
| 410 | "name" after a successful match (name termin- |
| 411 | ated by next non-alphanumeric character) |
| 412 | \Jdd set up a JIT stack of dd kilobytes maximum (any |
| 413 | number of digits) |
| 414 | \L call pcre_get_substringlist() after a |
| 415 | successful match |
| 416 | \M discover the minimum MATCH_LIMIT and |
| 417 | MATCH_LIMIT_RECURSION settings |
| 418 | \N pass the PCRE_NOTEMPTY option to pcre_exec() |
| 419 | or pcre_dfa_exec(); if used twice, pass the |
| 420 | PCRE_NOTEMPTY_ATSTART option |
| 421 | \Odd set the size of the output vector passed to |
| 422 | pcre_exec() to dd (any number of digits) |
| 423 | \P pass the PCRE_PARTIAL_SOFT option to pcre_exec() |
| 424 | or pcre_dfa_exec(); if used twice, pass the |
| 425 | PCRE_PARTIAL_HARD option |
| 426 | \Qdd set the PCRE_MATCH_LIMIT_RECURSION limit to dd |
| 427 | (any number of digits) |
| 428 | \R pass the PCRE_DFA_RESTART option to pcre_dfa_exec() |
| 429 | \S output details of memory get/free calls during matching |
| 430 | \Y pass the PCRE_NO_START_OPTIMIZE option to pcre_exec() |
| 431 | or pcre_dfa_exec() |
| 432 | \Z pass the PCRE_NOTEOL option to pcre_exec() |
| 433 | or pcre_dfa_exec() |
| 434 | \? pass the PCRE_NO_UTF8_CHECK option to |
| 435 | pcre_exec() or pcre_dfa_exec() |
| 436 | \>dd start the match at offset dd (optional "-"; then |
| 437 | any number of digits); this sets the startoffset |
| 438 | argument for pcre_exec() or pcre_dfa_exec() |
| 439 | \<cr> pass the PCRE_NEWLINE_CR option to pcre_exec() |
| 440 | or pcre_dfa_exec() |
| 441 | \<lf> pass the PCRE_NEWLINE_LF option to pcre_exec() |
| 442 | or pcre_dfa_exec() |
| 443 | \<crlf> pass the PCRE_NEWLINE_CRLF option to pcre_exec() |
| 444 | or pcre_dfa_exec() |
| 445 | \<anycrlf> pass the PCRE_NEWLINE_ANYCRLF option to pcre_exec() |
| 446 | or pcre_dfa_exec() |
| 447 | \<any> pass the PCRE_NEWLINE_ANY option to pcre_exec() |
| 448 | or pcre_dfa_exec() |
| 449 | |
| 450 | Note that \xhh always specifies one byte, even in UTF-8 mode; this |
| 451 | makes it possible to construct invalid UTF-8 sequences for testing pur- |
| 452 | poses. On the other hand, \x{hh} is interpreted as a UTF-8 character in |
| 453 | UTF-8 mode, generating more than one byte if the value is greater than |
| 454 | 127. When not in UTF-8 mode, it generates one byte for values less than |
| 455 | 256, and causes an error for greater values. |
| 456 | |
| 457 | The escapes that specify line ending sequences are literal strings, |
| 458 | exactly as shown. No more than one newline setting should be present in |
| 459 | any data line. |
| 460 | |
| 461 | A backslash followed by anything else just escapes the anything else. |
| 462 | If the very last character is a backslash, it is ignored. This gives a |
| 463 | way of passing an empty line as data, since a real empty line termi- |
| 464 | nates the data input. |
| 465 | |
| 466 | The \J escape provides a way of setting the maximum stack size that is |
| 467 | used by the just-in-time optimization code. It is ignored if JIT opti- |
| 468 | mization is not being used. Providing a stack that is larger than the |
| 469 | default 32K is necessary only for very complicated patterns. |
| 470 | |
| 471 | If \M is present, pcretest calls pcre_exec() several times, with dif- |
| 472 | ferent values in the match_limit and match_limit_recursion fields of |
| 473 | the pcre_extra data structure, until it finds the minimum numbers for |
| 474 | each parameter that allow pcre_exec() to complete without error. |
| 475 | Because this is testing a specific feature of the normal interpretive |
| 476 | pcre_exec() execution, the use of any JIT optimization that might have |
| 477 | been set up by the /S+ qualifier of -s+ option is disabled. |
| 478 | |
| 479 | The match_limit number is a measure of the amount of backtracking that |
| 480 | takes place, and checking it out can be instructive. For most simple |
| 481 | matches, the number is quite small, but for patterns with very large |
| 482 | numbers of matching possibilities, it can become large very quickly |
| 483 | with increasing length of subject string. The match_limit_recursion |
| 484 | number is a measure of how much stack (or, if PCRE is compiled with |
| 485 | NO_RECURSE, how much heap) memory is needed to complete the match |
| 486 | attempt. |
| 487 | |
| 488 | When \O is used, the value specified may be higher or lower than the |
| 489 | size set by the -O command line option (or defaulted to 45); \O applies |
| 490 | only to the call of pcre_exec() for the line in which it appears. |
| 491 | |
| 492 | If the /P modifier was present on the pattern, causing the POSIX wrap- |
| 493 | per API to be used, the only option-setting sequences that have any |
| 494 | effect are \B, \N, and \Z, causing REG_NOTBOL, REG_NOTEMPTY, and |
| 495 | REG_NOTEOL, respectively, to be passed to regexec(). |
| 496 | |
| 497 | The use of \x{hh...} to represent UTF-8 characters is not dependent on |
| 498 | the use of the /8 modifier on the pattern. It is recognized always. |
| 499 | There may be any number of hexadecimal digits inside the braces. The |
| 500 | result is from one to six bytes, encoded according to the original |
| 501 | UTF-8 rules of RFC 2279. This allows for values in the range 0 to |
| 502 | 0x7FFFFFFF. Note that not all of those are valid Unicode code points, |
| 503 | or indeed valid UTF-8 characters according to the later rules in RFC |
| 504 | 3629. |
| 505 | |
| 506 | |
| 507 | THE ALTERNATIVE MATCHING FUNCTION |
| 508 | |
| 509 | By default, pcretest uses the standard PCRE matching function, |
| 510 | pcre_exec() to match each data line. From release 6.0, PCRE supports an |
| 511 | alternative matching function, pcre_dfa_test(), which operates in a |
| 512 | different way, and has some restrictions. The differences between the |
| 513 | two functions are described in the pcrematching documentation. |
| 514 | |
| 515 | If a data line contains the \D escape sequence, or if the command line |
| 516 | contains the -dfa option, the alternative matching function is called. |
| 517 | This function finds all possible matches at a given point. If, however, |
| 518 | the \F escape sequence is present in the data line, it stops after the |
| 519 | first match is found. This is always the shortest possible match. |
| 520 | |
| 521 | |
| 522 | DEFAULT OUTPUT FROM PCRETEST |
| 523 | |
| 524 | This section describes the output when the normal matching function, |
| 525 | pcre_exec(), is being used. |
| 526 | |
| 527 | When a match succeeds, pcretest outputs the list of captured substrings |
| 528 | that pcre_exec() returns, starting with number 0 for the string that |
| 529 | matched the whole pattern. Otherwise, it outputs "No match" when the |
| 530 | return is PCRE_ERROR_NOMATCH, and "Partial match:" followed by the par- |
| 531 | tially matching substring when pcre_exec() returns PCRE_ERROR_PARTIAL. |
| 532 | (Note that this is the entire substring that was inspected during the |
| 533 | partial match; it may include characters before the actual match start |
| 534 | if a lookbehind assertion, \K, \b, or \B was involved.) For any other |
| 535 | return, pcretest outputs the PCRE negative error number and a short |
| 536 | descriptive phrase. If the error is a failed UTF-8 string check, the |
| 537 | byte offset of the start of the failing character and the reason code |
| 538 | are also output, provided that the size of the output vector is at |
| 539 | least two. Here is an example of an interactive pcretest run. |
| 540 | |
| 541 | $ pcretest |
| 542 | PCRE version 8.13 2011-04-30 |
| 543 | |
| 544 | re> /^abc(\d+)/ |
| 545 | data> abc123 |
| 546 | 0: abc123 |
| 547 | 1: 123 |
| 548 | data> xyz |
| 549 | No match |
| 550 | |
| 551 | Unset capturing substrings that are not followed by one that is set are |
| 552 | not returned by pcre_exec(), and are not shown by pcretest. In the fol- |
| 553 | lowing example, there are two capturing substrings, but when the first |
| 554 | data line is matched, the second, unset substring is not shown. An |
| 555 | "internal" unset substring is shown as "<unset>", as for the second |
| 556 | data line. |
| 557 | |
| 558 | re> /(a)|(b)/ |
| 559 | data> a |
| 560 | 0: a |
| 561 | 1: a |
| 562 | data> b |
| 563 | 0: b |
| 564 | 1: <unset> |
| 565 | 2: b |
| 566 | |
| 567 | If the strings contain any non-printing characters, they are output as |
| 568 | \0x escapes, or as \x{...} escapes if the /8 modifier was present on |
| 569 | the pattern. See below for the definition of non-printing characters. |
| 570 | If the pattern has the /+ modifier, the output for substring 0 is fol- |
| 571 | lowed by the the rest of the subject string, identified by "0+" like |
| 572 | this: |
| 573 | |
| 574 | re> /cat/+ |
| 575 | data> cataract |
| 576 | 0: cat |
| 577 | 0+ aract |
| 578 | |
| 579 | If the pattern has the /g or /G modifier, the results of successive |
| 580 | matching attempts are output in sequence, like this: |
| 581 | |
| 582 | re> /\Bi(\w\w)/g |
| 583 | data> Mississippi |
| 584 | 0: iss |
| 585 | 1: ss |
| 586 | 0: iss |
| 587 | 1: ss |
| 588 | 0: ipp |
| 589 | 1: pp |
| 590 | |
| 591 | "No match" is output only if the first match attempt fails. Here is an |
| 592 | example of a failure message (the offset 4 that is specified by \>4 is |
| 593 | past the end of the subject string): |
| 594 | |
| 595 | re> /xyz/ |
| 596 | data> xyz\>4 |
| 597 | Error -24 (bad offset value) |
| 598 | |
| 599 | If any of the sequences \C, \G, or \L are present in a data line that |
| 600 | is successfully matched, the substrings extracted by the convenience |
| 601 | functions are output with C, G, or L after the string number instead of |
| 602 | a colon. This is in addition to the normal full list. The string length |
| 603 | (that is, the return from the extraction function) is given in paren- |
| 604 | theses after each string for \C and \G. |
| 605 | |
| 606 | Note that whereas patterns can be continued over several lines (a plain |
| 607 | ">" prompt is used for continuations), data lines may not. However new- |
| 608 | lines can be included in data by means of the \n escape (or \r, \r\n, |
| 609 | etc., depending on the newline sequence setting). |
| 610 | |
| 611 | |
| 612 | OUTPUT FROM THE ALTERNATIVE MATCHING FUNCTION |
| 613 | |
| 614 | When the alternative matching function, pcre_dfa_exec(), is used (by |
| 615 | means of the \D escape sequence or the -dfa command line option), the |
| 616 | output consists of a list of all the matches that start at the first |
| 617 | point in the subject where there is at least one match. For example: |
| 618 | |
| 619 | re> /(tang|tangerine|tan)/ |
| 620 | data> yellow tangerine\D |
| 621 | 0: tangerine |
| 622 | 1: tang |
| 623 | 2: tan |
| 624 | |
| 625 | (Using the normal matching function on this data finds only "tang".) |
| 626 | The longest matching string is always given first (and numbered zero). |
| 627 | After a PCRE_ERROR_PARTIAL return, the output is "Partial match:", fol- |
| 628 | lowed by the partially matching substring. (Note that this is the |
| 629 | entire substring that was inspected during the partial match; it may |
| 630 | include characters before the actual match start if a lookbehind asser- |
| 631 | tion, \K, \b, or \B was involved.) |
| 632 | |
| 633 | If /g is present on the pattern, the search for further matches resumes |
| 634 | at the end of the longest match. For example: |
| 635 | |
| 636 | re> /(tang|tangerine|tan)/g |
| 637 | data> yellow tangerine and tangy sultana\D |
| 638 | 0: tangerine |
| 639 | 1: tang |
| 640 | 2: tan |
| 641 | 0: tang |
| 642 | 1: tan |
| 643 | 0: tan |
| 644 | |
| 645 | Since the matching function does not support substring capture, the |
| 646 | escape sequences that are concerned with captured substrings are not |
| 647 | relevant. |
| 648 | |
| 649 | |
| 650 | RESTARTING AFTER A PARTIAL MATCH |
| 651 | |
| 652 | When the alternative matching function has given the PCRE_ERROR_PARTIAL |
| 653 | return, indicating that the subject partially matched the pattern, you |
| 654 | can restart the match with additional subject data by means of the \R |
| 655 | escape sequence. For example: |
| 656 | |
| 657 | re> /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/ |
| 658 | data> 23ja\P\D |
| 659 | Partial match: 23ja |
| 660 | data> n05\R\D |
| 661 | 0: n05 |
| 662 | |
| 663 | For further information about partial matching, see the pcrepartial |
| 664 | documentation. |
| 665 | |
| 666 | |
| 667 | CALLOUTS |
| 668 | |
| 669 | If the pattern contains any callout requests, pcretest's callout func- |
| 670 | tion is called during matching. This works with both matching func- |
| 671 | tions. By default, the called function displays the callout number, the |
| 672 | start and current positions in the text at the callout time, and the |
| 673 | next pattern item to be tested. For example, the output |
| 674 | |
| 675 | --->pqrabcdef |
| 676 | 0 ^ ^ \d |
| 677 | |
| 678 | indicates that callout number 0 occurred for a match attempt starting |
| 679 | at the fourth character of the subject string, when the pointer was at |
| 680 | the seventh character of the data, and when the next pattern item was |
| 681 | \d. Just one circumflex is output if the start and current positions |
| 682 | are the same. |
| 683 | |
| 684 | Callouts numbered 255 are assumed to be automatic callouts, inserted as |
| 685 | a result of the /C pattern modifier. In this case, instead of showing |
| 686 | the callout number, the offset in the pattern, preceded by a plus, is |
| 687 | output. For example: |
| 688 | |
| 689 | re> /\d?[A-E]\*/C |
| 690 | data> E* |
| 691 | --->E* |
| 692 | +0 ^ \d? |
| 693 | +3 ^ [A-E] |
| 694 | +8 ^^ \* |
| 695 | +10 ^ ^ |
| 696 | 0: E* |
| 697 | |
| 698 | If a pattern contains (*MARK) items, an additional line is output when- |
| 699 | ever a change of latest mark is passed to the callout function. For |
| 700 | example: |
| 701 | |
| 702 | re> /a(*MARK:X)bc/C |
| 703 | data> abc |
| 704 | --->abc |
| 705 | +0 ^ a |
| 706 | +1 ^^ (*MARK:X) |
| 707 | +10 ^^ b |
| 708 | Latest Mark: X |
| 709 | +11 ^ ^ c |
| 710 | +12 ^ ^ |
| 711 | 0: abc |
| 712 | |
| 713 | The mark changes between matching "a" and "b", but stays the same for |
| 714 | the rest of the match, so nothing more is output. If, as a result of |
| 715 | backtracking, the mark reverts to being unset, the text "<unset>" is |
| 716 | output. |
| 717 | |
| 718 | The callout function in pcretest returns zero (carry on matching) by |
| 719 | default, but you can use a \C item in a data line (as described above) |
| 720 | to change this and other parameters of the callout. |
| 721 | |
| 722 | Inserting callouts can be helpful when using pcretest to check compli- |
| 723 | cated regular expressions. For further information about callouts, see |
| 724 | the pcrecallout documentation. |
| 725 | |
| 726 | |
| 727 | NON-PRINTING CHARACTERS |
| 728 | |
| 729 | When pcretest is outputting text in the compiled version of a pattern, |
| 730 | bytes other than 32-126 are always treated as non-printing characters |
| 731 | are are therefore shown as hex escapes. |
| 732 | |
| 733 | When pcretest is outputting text that is a matched part of a subject |
| 734 | string, it behaves in the same way, unless a different locale has been |
| 735 | set for the pattern (using the /L modifier). In this case, the |
| 736 | isprint() function to distinguish printing and non-printing characters. |
| 737 | |
| 738 | |
| 739 | SAVING AND RELOADING COMPILED PATTERNS |
| 740 | |
| 741 | The facilities described in this section are not available when the |
| 742 | POSIX interface to PCRE is being used, that is, when the /P pattern |
| 743 | modifier is specified. |
| 744 | |
| 745 | When the POSIX interface is not in use, you can cause pcretest to write |
| 746 | a compiled pattern to a file, by following the modifiers with > and a |
| 747 | file name. For example: |
| 748 | |
| 749 | /pattern/im >/some/file |
| 750 | |
| 751 | See the pcreprecompile documentation for a discussion about saving and |
| 752 | re-using compiled patterns. Note that if the pattern was successfully |
| 753 | studied with JIT optimization, the JIT data cannot be saved. |
| 754 | |
| 755 | The data that is written is binary. The first eight bytes are the |
| 756 | length of the compiled pattern data followed by the length of the |
| 757 | optional study data, each written as four bytes in big-endian order |
| 758 | (most significant byte first). If there is no study data (either the |
| 759 | pattern was not studied, or studying did not return any data), the sec- |
| 760 | ond length is zero. The lengths are followed by an exact copy of the |
| 761 | compiled pattern. If there is additional study data, this (excluding |
| 762 | any JIT data) follows immediately after the compiled pattern. After |
| 763 | writing the file, pcretest expects to read a new pattern. |
| 764 | |
| 765 | A saved pattern can be reloaded into pcretest by specifying < and a |
| 766 | file name instead of a pattern. The name of the file must not contain a |
| 767 | < character, as otherwise pcretest will interpret the line as a pattern |
| 768 | delimited by < characters. For example: |
| 769 | |
| 770 | re> </some/file |
| 771 | Compiled pattern loaded from /some/file |
| 772 | No study data |
| 773 | |
| 774 | If the pattern was previously studied with the JIT optimization, the |
| 775 | JIT information cannot be saved and restored, and so is lost. When the |
| 776 | pattern has been loaded, pcretest proceeds to read data lines in the |
| 777 | usual way. |
| 778 | |
| 779 | You can copy a file written by pcretest to a different host and reload |
| 780 | it there, even if the new host has opposite endianness to the one on |
| 781 | which the pattern was compiled. For example, you can compile on an i86 |
| 782 | machine and run on a SPARC machine. |
| 783 | |
| 784 | File names for saving and reloading can be absolute or relative, but |
| 785 | note that the shell facility of expanding a file name that starts with |
| 786 | a tilde (~) is not available. |
| 787 | |
| 788 | The ability to save and reload files in pcretest is intended for test- |
| 789 | ing and experimentation. It is not intended for production use because |
| 790 | only a single pattern can be written to a file. Furthermore, there is |
| 791 | no facility for supplying custom character tables for use with a |
| 792 | reloaded pattern. If the original pattern was compiled with custom |
| 793 | tables, an attempt to match a subject string using a reloaded pattern |
| 794 | is likely to cause pcretest to crash. Finally, if you attempt to load |
| 795 | a file that is not in the correct format, the result is undefined. |
| 796 | |
| 797 | |
| 798 | SEE ALSO |
| 799 | |
| 800 | pcre(3), pcreapi(3), pcrecallout(3), pcrejit, pcrematching(3), pcrepar- |
| 801 | tial(d), pcrepattern(3), pcreprecompile(3). |
| 802 | |
| 803 | |
| 804 | AUTHOR |
| 805 | |
| 806 | Philip Hazel |
| 807 | University Computing Service |
| 808 | Cambridge CB2 3QH, England. |
| 809 | |
| 810 | |
| 811 | REVISION |
| 812 | |
| 813 | Last updated: 02 December 2011 |
| 814 | Copyright (c) 1997-2011 University of Cambridge. |