1 This file contains the PCRE man page that described the pcretest program. Note
2 that not all of the features of PCRE are available in the limited version that
4 -------------------------------------------------------------------------------
6 PCRETEST(1) PCRETEST(1)
10 pcretest - a program for testing Perl-compatible regular expressions.
15 pcretest [options] [source] [destination]
17 pcretest was written as a test program for the PCRE regular expression
18 library itself, but it can also be used for experimenting with regular
19 expressions. This document describes the features of the test program;
20 for details of the regular expressions themselves, see the pcrepattern
21 documentation. For details of the PCRE library function calls and their
22 options, see the pcreapi documentation.
27 -C Output the version number of the PCRE library, and all avail-
28 able information about the optional features that are
29 included, and then exit.
31 -d Behave as if each regex has the /D (debug) modifier; the
32 internal form is output after compilation.
34 -dfa Behave as if each data line contains the \D escape sequence;
35 this causes the alternative matching function,
36 pcre_dfa_exec(), to be used instead of the standard
37 pcre_exec() function (more detail is given below).
39 -i Behave as if each regex has the /I modifier; information
40 about the compiled pattern is given after compilation.
42 -m Output the size of each compiled pattern after it has been
43 compiled. This is equivalent to adding /M to each regular
44 expression. For compatibility with earlier versions of
45 pcretest, -s is a synonym for -m.
47 -o osize Set the number of elements in the output vector that is used
48 when calling pcre_exec() to be osize. The default value is
49 45, which is enough for 14 capturing subexpressions. The vec-
50 tor size can be changed for individual matching calls by
51 including \O in the data line (see below).
53 -p Behave as if each regex has the /P modifier; the POSIX wrap-
54 per API is used to call PCRE. None of the other options has
55 any effect when -p is set.
57 -q Do not output the version number of pcretest at the start of
60 -S size On Unix-like systems, set the size of the runtime stack to
63 -t Run each compile, study, and match many times with a timer,
64 and output resulting time per compile or match (in millisec-
65 onds). Do not set -m with -t, because you will then get the
66 size output a zillion times, and the timing will be dis-
72 If pcretest is given two filename arguments, it reads from the first
73 and writes to the second. If it is given only one filename argument, it
74 reads from that file and writes to stdout. Otherwise, it reads from
75 stdin and writes to stdout, and prompts for each line of input, using
76 "re>" to prompt for regular expressions, and "data>" to prompt for data
79 The program handles any number of sets of input on a single input file.
80 Each set starts with a regular expression, and continues with any num-
81 ber of data lines to be matched against the pattern.
83 Each data line is matched separately and independently. If you want to
84 do multi-line matches, you have to use the \n escape sequence (or \r or
85 \r\n, depending on the newline setting) in a single line of input to
86 encode the newline characters. There is no limit on the length of data
87 lines; the input buffer is automatically extended if it is too small.
89 An empty line signals the end of the data lines, at which point a new
90 regular expression is read. The regular expressions are given enclosed
91 in any non-alphanumeric delimiters other than backslash, for example:
95 White space before the initial delimiter is ignored. A regular expres-
96 sion may be continued over several input lines, in which case the new-
97 line characters are included within it. It is possible to include the
98 delimiter within the pattern by escaping it, for example
102 If you do so, the escape and the delimiter form part of the pattern,
103 but since delimiters are always non-alphanumeric, this does not affect
104 its interpretation. If the terminating delimiter is immediately fol-
105 lowed by a backslash, for example,
109 then a backslash is added to the end of the pattern. This is done to
110 provide a way of testing the error condition that arises if a pattern
111 finishes with a backslash, because
115 is interpreted as the first line of a pattern that starts with "abc/",
116 causing pcretest to read the next line as a continuation of the regular
122 A pattern may be followed by any number of modifiers, which are mostly
123 single characters. Following Perl usage, these are referred to below
124 as, for example, "the /i modifier", even though the delimiter of the
125 pattern need not always be a slash, and no slash is used when writing
126 modifiers. Whitespace may appear between the final pattern delimiter
127 and the first modifier, and between the modifiers themselves.
129 The /i, /m, /s, and /x modifiers set the PCRE_CASELESS, PCRE_MULTILINE,
130 PCRE_DOTALL, or PCRE_EXTENDED options, respectively, when pcre_com-
131 pile() is called. These four modifier letters have the same effect as
132 they do in Perl. For example:
136 The following table shows additional modifiers for setting PCRE options
137 that do not correspond to anything in Perl:
141 /E PCRE_DOLLAR_ENDONLY
144 /N PCRE_NO_AUTO_CAPTURE
147 /<cr> PCRE_NEWLINE_CR
148 /<lf> PCRE_NEWLINE_LF
149 /<crlf> PCRE_NEWLINE_CRLF
151 Those specifying line endings are literal strings as shown. Details of
152 the meanings of these PCRE options are given in the pcreapi documenta-
155 Finding all matches in a string
157 Searching for all possible matches within each subject string can be
158 requested by the /g or /G modifier. After finding a match, PCRE is
159 called again to search the remainder of the subject string. The differ-
160 ence between /g and /G is that the former uses the startoffset argument
161 to pcre_exec() to start searching at a new point within the entire
162 string (which is in effect what Perl does), whereas the latter passes
163 over a shortened substring. This makes a difference to the matching
164 process if the pattern begins with a lookbehind assertion (including \b
167 If any call to pcre_exec() in a /g or /G sequence matches an empty
168 string, the next call is done with the PCRE_NOTEMPTY and PCRE_ANCHORED
169 flags set in order to search for another, non-empty, match at the same
170 point. If this second match fails, the start offset is advanced by
171 one, and the normal match is retried. This imitates the way Perl han-
172 dles such cases when using the /g modifier or the split() function.
176 There are yet more modifiers for controlling the way pcretest operates.
178 The /+ modifier requests that as well as outputting the substring that
179 matched the entire pattern, pcretest should in addition output the
180 remainder of the subject string. This is useful for tests where the
181 subject contains multiple copies of the same substring.
183 The /L modifier must be followed directly by the name of a locale, for
188 For this reason, it must be the last modifier. The given locale is set,
189 pcre_maketables() is called to build a set of character tables for the
190 locale, and this is then passed to pcre_compile() when compiling the
191 regular expression. Without an /L modifier, NULL is passed as the
192 tables pointer; that is, /L applies only to the expression on which it
195 The /I modifier requests that pcretest output information about the
196 compiled pattern (whether it is anchored, has a fixed first character,
197 and so on). It does this by calling pcre_fullinfo() after compiling a
198 pattern. If the pattern is studied, the results of that are also out-
201 The /D modifier is a PCRE debugging feature, which also assumes /I. It
202 causes the internal form of compiled regular expressions to be output
203 after compilation. If the pattern was studied, the information returned
206 The /F modifier causes pcretest to flip the byte order of the fields in
207 the compiled pattern that contain 2-byte and 4-byte numbers. This
208 facility is for testing the feature in PCRE that allows it to execute
209 patterns that were compiled on a host with a different endianness. This
210 feature is not available when the POSIX interface to PCRE is being
211 used, that is, when the /P pattern modifier is specified. See also the
212 section about saving and reloading compiled patterns below.
214 The /S modifier causes pcre_study() to be called after the expression
215 has been compiled, and the results used when the expression is matched.
217 The /M modifier causes the size of memory block used to hold the com-
218 piled pattern to be output.
220 The /P modifier causes pcretest to call PCRE via the POSIX wrapper API
221 rather than its native API. When this is done, all other modifiers
222 except /i, /m, and /+ are ignored. REG_ICASE is set if /i is present,
223 and REG_NEWLINE is set if /m is present. The wrapper functions force
224 PCRE_DOLLAR_ENDONLY always, and PCRE_DOTALL unless REG_NEWLINE is set.
226 The /8 modifier causes pcretest to call PCRE with the PCRE_UTF8 option
227 set. This turns on support for UTF-8 character handling in PCRE, pro-
228 vided that it was compiled with this support enabled. This modifier
229 also causes any non-printing characters in output strings to be printed
230 using the \x{hh...} notation if they are valid UTF-8 sequences.
232 If the /? modifier is used with /8, it causes pcretest to call
233 pcre_compile() with the PCRE_NO_UTF8_CHECK option, to suppress the
234 checking of the string for UTF-8 validity.
239 Before each data line is passed to pcre_exec(), leading and trailing
240 whitespace is removed, and it is then scanned for \ escapes. Some of
241 these are pretty esoteric features, intended for checking out some of
242 the more complicated features of PCRE. If you are just testing "ordi-
243 nary" regular expressions, you probably don't need any of these. The
244 following escapes are recognized:
251 \qdd set the PCRE_MATCH_LIMIT limit to dd
252 (any number of digits)
256 \nnn octal character (up to 3 octal digits)
257 \xhh hexadecimal character (up to 2 hex digits)
258 \x{hh...} hexadecimal character, any number of digits
260 \A pass the PCRE_ANCHORED option to pcre_exec()
262 \B pass the PCRE_NOTBOL option to pcre_exec()
264 \Cdd call pcre_copy_substring() for substring dd
265 after a successful match (number less than 32)
266 \Cname call pcre_copy_named_substring() for substring
267 "name" after a successful match (name termin-
268 ated by next non alphanumeric character)
269 \C+ show the current captured substrings at callout
271 \C- do not supply a callout function
272 \C!n return 1 instead of 0 when callout number n is
274 \C!n!m return 1 instead of 0 when callout number n is
275 reached for the nth time
276 \C*n pass the number n (may be negative) as callout
277 data; this is used as the callout return value
278 \D use the pcre_dfa_exec() match function
279 \F only shortest match for pcre_dfa_exec()
280 \Gdd call pcre_get_substring() for substring dd
281 after a successful match (number less than 32)
282 \Gname call pcre_get_named_substring() for substring
283 "name" after a successful match (name termin-
284 ated by next non-alphanumeric character)
285 \L call pcre_get_substringlist() after a
287 \M discover the minimum MATCH_LIMIT and
288 MATCH_LIMIT_RECURSION settings
289 \N pass the PCRE_NOTEMPTY option to pcre_exec()
291 \Odd set the size of the output vector passed to
292 pcre_exec() to dd (any number of digits)
293 \P pass the PCRE_PARTIAL option to pcre_exec()
295 \Qdd set the PCRE_MATCH_LIMIT_RECURSION limit to dd
296 (any number of digits)
297 \R pass the PCRE_DFA_RESTART option to pcre_dfa_exec()
298 \S output details of memory get/free calls during matching
299 \Z pass the PCRE_NOTEOL option to pcre_exec()
301 \? pass the PCRE_NO_UTF8_CHECK option to
302 pcre_exec() or pcre_dfa_exec()
303 \>dd start the match at offset dd (any number of digits);
304 this sets the startoffset argument for pcre_exec()
306 \<cr> pass the PCRE_NEWLINE_CR option to pcre_exec()
308 \<lf> pass the PCRE_NEWLINE_LF option to pcre_exec()
310 \<crlf> pass the PCRE_NEWLINE_CRLF option to pcre_exec()
313 The escapes that specify line endings are literal strings, exactly as
314 shown. A backslash followed by anything else just escapes the anything
315 else. If the very last character is a backslash, it is ignored. This
316 gives a way of passing an empty line as data, since a real empty line
317 terminates the data input.
319 If \M is present, pcretest calls pcre_exec() several times, with dif-
320 ferent values in the match_limit and match_limit_recursion fields of
321 the pcre_extra data structure, until it finds the minimum numbers for
322 each parameter that allow pcre_exec() to complete. The match_limit num-
323 ber is a measure of the amount of backtracking that takes place, and
324 checking it out can be instructive. For most simple matches, the number
325 is quite small, but for patterns with very large numbers of matching
326 possibilities, it can become large very quickly with increasing length
327 of subject string. The match_limit_recursion number is a measure of how
328 much stack (or, if PCRE is compiled with NO_RECURSE, how much heap)
329 memory is needed to complete the match attempt.
331 When \O is used, the value specified may be higher or lower than the
332 size set by the -O command line option (or defaulted to 45); \O applies
333 only to the call of pcre_exec() for the line in which it appears.
335 If the /P modifier was present on the pattern, causing the POSIX wrap-
336 per API to be used, the only option-setting sequences that have any
337 effect are \B and \Z, causing REG_NOTBOL and REG_NOTEOL, respectively,
338 to be passed to regexec().
340 The use of \x{hh...} to represent UTF-8 characters is not dependent on
341 the use of the /8 modifier on the pattern. It is recognized always.
342 There may be any number of hexadecimal digits inside the braces. The
343 result is from one to six bytes, encoded according to the UTF-8 rules.
346 THE ALTERNATIVE MATCHING FUNCTION
348 By default, pcretest uses the standard PCRE matching function,
349 pcre_exec() to match each data line. From release 6.0, PCRE supports an
350 alternative matching function, pcre_dfa_test(), which operates in a
351 different way, and has some restrictions. The differences between the
352 two functions are described in the pcrematching documentation.
354 If a data line contains the \D escape sequence, or if the command line
355 contains the -dfa option, the alternative matching function is called.
356 This function finds all possible matches at a given point. If, however,
357 the \F escape sequence is present in the data line, it stops after the
358 first match is found. This is always the shortest possible match.
361 DEFAULT OUTPUT FROM PCRETEST
363 This section describes the output when the normal matching function,
364 pcre_exec(), is being used.
366 When a match succeeds, pcretest outputs the list of captured substrings
367 that pcre_exec() returns, starting with number 0 for the string that
368 matched the whole pattern. Otherwise, it outputs "No match" or "Partial
369 match" when pcre_exec() returns PCRE_ERROR_NOMATCH or PCRE_ERROR_PAR-
370 TIAL, respectively, and otherwise the PCRE negative error number. Here
371 is an example of an interactive pcretest run.
374 PCRE version 5.00 07-Sep-2004
383 If the strings contain any non-printing characters, they are output as
384 \0x escapes, or as \x{...} escapes if the /8 modifier was present on
385 the pattern. If the pattern has the /+ modifier, the output for sub-
386 string 0 is followed by the the rest of the subject string, identified
394 If the pattern has the /g or /G modifier, the results of successive
395 matching attempts are output in sequence, like this:
406 "No match" is output only if the first match attempt fails.
408 If any of the sequences \C, \G, or \L are present in a data line that
409 is successfully matched, the substrings extracted by the convenience
410 functions are output with C, G, or L after the string number instead of
411 a colon. This is in addition to the normal full list. The string length
412 (that is, the return from the extraction function) is given in paren-
413 theses after each string for \C and \G.
415 Note that while patterns can be continued over several lines (a plain
416 ">" prompt is used for continuations), data lines may not. However new-
417 lines can be included in data by means of the \n escape (or \r or \r\n
418 for those newline settings).
421 OUTPUT FROM THE ALTERNATIVE MATCHING FUNCTION
423 When the alternative matching function, pcre_dfa_exec(), is used (by
424 means of the \D escape sequence or the -dfa command line option), the
425 output consists of a list of all the matches that start at the first
426 point in the subject where there is at least one match. For example:
428 re> /(tang|tangerine|tan)/
429 data> yellow tangerine\D
434 (Using the normal matching function on this data finds only "tang".)
435 The longest matching string is always given first (and numbered zero).
437 If /gP is present on the pattern, the search for further matches
438 resumes at the end of the longest match. For example:
440 re> /(tang|tangerine|tan)/g
441 data> yellow tangerine and tangy sultana\D
449 Since the matching function does not support substring capture, the
450 escape sequences that are concerned with captured substrings are not
454 RESTARTING AFTER A PARTIAL MATCH
456 When the alternative matching function has given the PCRE_ERROR_PARTIAL
457 return, indicating that the subject partially matched the pattern, you
458 can restart the match with additional subject data by means of the \R
459 escape sequence. For example:
461 re> /^?(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)$/
467 For further information about partial matching, see the pcrepartial
473 If the pattern contains any callout requests, pcretest's callout func-
474 tion is called during matching. This works with both matching func-
475 tions. By default, the called function displays the callout number, the
476 start and current positions in the text at the callout time, and the
477 next pattern item to be tested. For example, the output
482 indicates that callout number 0 occurred for a match attempt starting
483 at the fourth character of the subject string, when the pointer was at
484 the seventh character of the data, and when the next pattern item was
485 \d. Just one circumflex is output if the start and current positions
488 Callouts numbered 255 are assumed to be automatic callouts, inserted as
489 a result of the /C pattern modifier. In this case, instead of showing
490 the callout number, the offset in the pattern, preceded by a plus, is
502 The callout function in pcretest returns zero (carry on matching) by
503 default, but you can use a \C item in a data line (as described above)
506 Inserting callouts can be helpful when using pcretest to check compli-
507 cated regular expressions. For further information about callouts, see
508 the pcrecallout documentation.
511 SAVING AND RELOADING COMPILED PATTERNS
513 The facilities described in this section are not available when the
514 POSIX inteface to PCRE is being used, that is, when the /P pattern mod-
517 When the POSIX interface is not in use, you can cause pcretest to write
518 a compiled pattern to a file, by following the modifiers with > and a
519 file name. For example:
521 /pattern/im >/some/file
523 See the pcreprecompile documentation for a discussion about saving and
524 re-using compiled patterns.
526 The data that is written is binary. The first eight bytes are the
527 length of the compiled pattern data followed by the length of the
528 optional study data, each written as four bytes in big-endian order
529 (most significant byte first). If there is no study data (either the
530 pattern was not studied, or studying did not return any data), the sec-
531 ond length is zero. The lengths are followed by an exact copy of the
532 compiled pattern. If there is additional study data, this follows imme-
533 diately after the compiled pattern. After writing the file, pcretest
534 expects to read a new pattern.
536 A saved pattern can be reloaded into pcretest by specifing < and a file
537 name instead of a pattern. The name of the file must not contain a <
538 character, as otherwise pcretest will interpret the line as a pattern
539 delimited by < characters. For example:
542 Compiled regex loaded from /some/file
545 When the pattern has been loaded, pcretest proceeds to read data lines
548 You can copy a file written by pcretest to a different host and reload
549 it there, even if the new host has opposite endianness to the one on
550 which the pattern was compiled. For example, you can compile on an i86
551 machine and run on a SPARC machine.
553 File names for saving and reloading can be absolute or relative, but
554 note that the shell facility of expanding a file name that starts with
555 a tilde (~) is not available.
557 The ability to save and reload files in pcretest is intended for test-
558 ing and experimentation. It is not intended for production use because
559 only a single pattern can be written to a file. Furthermore, there is
560 no facility for supplying custom character tables for use with a
561 reloaded pattern. If the original pattern was compiled with custom
562 tables, an attempt to match a subject string using a reloaded pattern
563 is likely to cause pcretest to crash. Finally, if you attempt to load
564 a file that is not in the correct format, the result is undefined.
570 University Computing Service,
571 Cambridge CB2 3QG, England.
573 Last updated: 29 June 2006
574 Copyright (c) 1997-2006 University of Cambridge.