/[pcre]/code/trunk/doc/pcre.txt
ViewVC logotype

Diff of /code/trunk/doc/pcre.txt

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 47 by nigel, Sat Feb 24 21:39:29 2007 UTC revision 77 by nigel, Sat Feb 24 21:40:45 2007 UTC
# Line 1  Line 1 
1    -----------------------------------------------------------------------------
2    This file contains a concatenation of the PCRE man pages, converted to plain
3    text format for ease of searching with a text editor, or for use on systems
4    that do not have a man page processor. The small individual files that give
5    synopses of each function in the library have not been included. There are
6    separate text files for the pcregrep and pcretest commands.
7    -----------------------------------------------------------------------------
8    
9    
10    
11  NAME  NAME
12       pcre - Perl-compatible regular expressions.         PCRE - Perl-compatible regular expressions
13    
14    
15    INTRODUCTION
16    
17           The  PCRE  library is a set of functions that implement regular expres-
18           sion pattern matching using the same syntax and semantics as Perl, with
19           just  a  few  differences.  The current implementation of PCRE (release
20           6.x) corresponds approximately with Perl  5.8,  including  support  for
21           UTF-8 encoded strings and Unicode general category properties. However,
22           this support has to be explicitly enabled; it is not the default.
23    
24           In addition to the Perl-compatible matching function,  PCRE  also  con-
25           tains  an  alternative matching function that matches the same compiled
26           patterns in a different way. In certain circumstances, the  alternative
27           function  has  some  advantages.  For  a discussion of the two matching
28           algorithms, see the pcrematching page.
29    
30           PCRE is written in C and released as a C library. A  number  of  people
31           have  written  wrappers and interfaces of various kinds. In particular,
32           Google Inc.  have provided a comprehensive C++  wrapper.  This  is  now
33           included as part of the PCRE distribution. The pcrecpp page has details
34           of this interface. Other people's contributions can  be  found  in  the
35           Contrib directory at the primary FTP site, which is:
36    
37           ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre
38    
39           Details  of  exactly which Perl regular expression features are and are
40           not supported by PCRE are given in separate documents. See the pcrepat-
41           tern and pcrecompat pages.
42    
43           Some  features  of  PCRE can be included, excluded, or changed when the
44           library is built. The pcre_config() function makes it  possible  for  a
45           client  to  discover  which  features are available. The features them-
46           selves are described in the pcrebuild page. Documentation about  build-
47           ing  PCRE for various operating systems can be found in the README file
48           in the source distribution.
49    
50           The library contains a number of undocumented  internal  functions  and
51           data  tables  that  are  used by more than one of the exported external
52           functions, but which are not intended  for  use  by  external  callers.
53           Their  names  all begin with "_pcre_", which hopefully will not provoke
54           any name clashes.
55    
56    
57    USER DOCUMENTATION
58    
59           The user documentation for PCRE comprises a number  of  different  sec-
60           tions.  In the "man" format, each of these is a separate "man page". In
61           the HTML format, each is a separate page, linked from the  index  page.
62           In  the  plain text format, all the sections are concatenated, for ease
63           of searching. The sections are as follows:
64    
65             pcre              this document
66             pcreapi           details of PCRE's native C API
67             pcrebuild         options for building PCRE
68             pcrecallout       details of the callout feature
69             pcrecompat        discussion of Perl compatibility
70             pcrecpp           details of the C++ wrapper
71             pcregrep          description of the pcregrep command
72             pcrematching      discussion of the two matching algorithms
73             pcrepartial       details of the partial matching facility
74             pcrepattern       syntax and semantics of supported
75                                 regular expressions
76             pcreperform       discussion of performance issues
77             pcreposix         the POSIX-compatible C API
78             pcreprecompile    details of saving and re-using precompiled patterns
79             pcresample        discussion of the sample program
80             pcretest          description of the pcretest testing command
81    
82  SYNOPSIS         In  addition,  in the "man" and HTML formats, there is a short page for
83       #include <pcre.h>         each C library function, listing its arguments and results.
84    
      pcre *pcre_compile(const char *pattern, int options,  
           const char **errptr, int *erroffset,  
           const unsigned char *tableptr);  
85    
86       pcre_extra *pcre_study(const pcre *code, int options,  LIMITATIONS
           const char **errptr);  
87    
88       int pcre_exec(const pcre *code, const pcre_extra *extra,         There are some size limitations in PCRE but it is hoped that they  will
89            const char *subject, int length, int startoffset,         never in practice be relevant.
           int options, int *ovector, int ovecsize);  
90    
91       int pcre_copy_substring(const char *subject, int *ovector,         The  maximum  length of a compiled pattern is 65539 (sic) bytes if PCRE
92            int stringcount, int stringnumber, char *buffer,         is compiled with the default internal linkage size of 2. If you want to
93            int buffersize);         process  regular  expressions  that are truly enormous, you can compile
94           PCRE with an internal linkage size of 3 or 4 (see the  README  file  in
95           the  source  distribution and the pcrebuild documentation for details).
96           In these cases the limit is substantially larger.  However,  the  speed
97           of execution will be slower.
98    
99           All values in repeating quantifiers must be less than 65536.  The maxi-
100           mum number of capturing subpatterns is 65535.
101    
102           There is no limit to the number of non-capturing subpatterns,  but  the
103           maximum  depth  of  nesting  of  all kinds of parenthesized subpattern,
104           including capturing subpatterns, assertions, and other types of subpat-
105           tern, is 200.
106    
107           The  maximum  length of a subject string is the largest positive number
108           that an integer variable can hold. However, when using the  traditional
109           matching function, PCRE uses recursion to handle subpatterns and indef-
110           inite repetition.  This means that the available stack space may  limit
111           the size of a subject string that can be processed by certain patterns.
112    
113    
114    UTF-8 AND UNICODE PROPERTY SUPPORT
115    
116           From release 3.3, PCRE has  had  some  support  for  character  strings
117           encoded  in the UTF-8 format. For release 4.0 this was greatly extended
118           to cover most common requirements, and in release 5.0  additional  sup-
119           port for Unicode general category properties was added.
120    
121           In  order  process  UTF-8 strings, you must build PCRE to include UTF-8
122           support in the code, and, in addition,  you  must  call  pcre_compile()
123           with  the PCRE_UTF8 option flag. When you do this, both the pattern and
124           any subject strings that are matched against it are  treated  as  UTF-8
125           strings instead of just strings of bytes.
126    
127           If  you compile PCRE with UTF-8 support, but do not use it at run time,
128           the library will be a bit bigger, but the additional run time  overhead
129           is  limited  to testing the PCRE_UTF8 flag in several places, so should
130           not be very large.
131    
132           If PCRE is built with Unicode character property support (which implies
133           UTF-8  support),  the  escape sequences \p{..}, \P{..}, and \X are sup-
134           ported.  The available properties that can be tested are limited to the
135           general  category  properties such as Lu for an upper case letter or Nd
136           for a decimal number. A full list is given in the pcrepattern  documen-
137           tation. The PCRE library is increased in size by about 90K when Unicode
138           property support is included.
139    
140           The following comments apply when PCRE is running in UTF-8 mode:
141    
142           1. When you set the PCRE_UTF8 flag, the strings passed as patterns  and
143           subjects  are  checked for validity on entry to the relevant functions.
144           If an invalid UTF-8 string is passed, an error return is given. In some
145           situations,  you  may  already  know  that  your strings are valid, and
146           therefore want to skip these checks in order to improve performance. If
147           you  set  the  PCRE_NO_UTF8_CHECK  flag at compile time or at run time,
148           PCRE assumes that the pattern or subject  it  is  given  (respectively)
149           contains  only valid UTF-8 codes. In this case, it does not diagnose an
150           invalid UTF-8 string. If you pass an invalid UTF-8 string to PCRE  when
151           PCRE_NO_UTF8_CHECK  is set, the results are undefined. Your program may
152           crash.
153    
154           2. In a pattern, the escape sequence \x{...}, where the contents of the
155           braces  is  a  string  of hexadecimal digits, is interpreted as a UTF-8
156           character whose code number is the given hexadecimal number, for  exam-
157           ple:  \x{1234}.  If a non-hexadecimal digit appears between the braces,
158           the item is not recognized.  This escape sequence can be used either as
159           a literal, or within a character class.
160    
161           3.  The  original hexadecimal escape sequence, \xhh, matches a two-byte
162           UTF-8 character if the value is greater than 127.
163    
164           4. Repeat quantifiers apply to complete UTF-8 characters, not to  indi-
165           vidual bytes, for example: \x{100}{3}.
166    
167           5.  The dot metacharacter matches one UTF-8 character instead of a sin-
168           gle byte.
169    
170           6. The escape sequence \C can be used to match a single byte  in  UTF-8
171           mode,  but  its  use can lead to some strange effects. This facility is
172           not available in the alternative matching function, pcre_dfa_exec().
173    
174           7. The character escapes \b, \B, \d, \D, \s, \S, \w, and  \W  correctly
175           test  characters of any code value, but the characters that PCRE recog-
176           nizes as digits, spaces, or word characters  remain  the  same  set  as
177           before, all with values less than 256. This remains true even when PCRE
178           includes Unicode property support, because to do otherwise  would  slow
179           down  PCRE in many common cases. If you really want to test for a wider
180           sense of, say, "digit", you must use Unicode  property  tests  such  as
181           \p{Nd}.
182    
183           8.  Similarly,  characters that match the POSIX named character classes
184           are all low-valued characters.
185    
186           9. Case-insensitive matching applies only to  characters  whose  values
187           are  less than 128, unless PCRE is built with Unicode property support.
188           Even when Unicode property support is available, PCRE  still  uses  its
189           own  character  tables when checking the case of low-valued characters,
190           so as not to degrade performance.  The Unicode property information  is
191           used only for characters with higher values.
192    
      int pcre_get_substring(const char *subject, int *ovector,  
           int stringcount, int stringnumber,  
           const char **stringptr);  
193    
194       int pcre_get_substring_list(const char *subject,  AUTHOR
           int *ovector, int stringcount, const char ***listptr);  
195    
196       const unsigned char *pcre_maketables(void);         Philip Hazel
197           University Computing Service,
198           Cambridge CB2 3QG, England.
199    
200           Putting  an actual email address here seems to have been a spam magnet,
201           so I've taken it away. If you want to email me, use my initial and sur-
202           name, separated by a dot, at the domain ucs.cam.ac.uk.
203    
204    Last updated: 07 March 2005
205    Copyright (c) 1997-2005 University of Cambridge.
206    -----------------------------------------------------------------------------
207    
      int pcre_fullinfo(const pcre *code, const pcre_extra *extra,  
           int what, void *where);  
208    
      int pcre_info(const pcre *code, int *optptr, *firstcharptr);  
209    
210       char *pcre_version(void);  NAME
211           PCRE - Perl-compatible regular expressions
212    
      void *(*pcre_malloc)(size_t);  
213    
214       void (*pcre_free)(void *);  PCRE BUILD-TIME OPTIONS
215    
216           This  document  describes  the  optional  features  of PCRE that can be
217           selected when the library is compiled. They are all selected, or  dese-
218           lected, by providing options to the configure script that is run before
219           the make command. The complete list of  options  for  configure  (which
220           includes  the  standard  ones such as the selection of the installation
221           directory) can be obtained by running
222    
223             ./configure --help
224    
225           The following sections describe certain options whose names begin  with
226           --enable  or  --disable. These settings specify changes to the defaults
227           for the configure command. Because of the  way  that  configure  works,
228           --enable  and  --disable  always  come  in  pairs, so the complementary
229           option always exists as well, but as it specifies the  default,  it  is
230           not described.
231    
 DESCRIPTION  
      The PCRE library is a set of functions that implement  regu-  
      lar  expression  pattern  matching using the same syntax and  
      semantics as Perl  5,  with  just  a  few  differences  (see  
      below).  The  current  implementation  corresponds  to  Perl  
      5.005, with some additional features from the Perl  develop-  
      ment release.  
   
      PCRE has its own native API,  which  is  described  in  this  
      document.  There  is  also  a  set of wrapper functions that  
      correspond to the POSIX regular expression API.   These  are  
      described in the pcreposix documentation.  
   
      The native API function prototypes are defined in the header  
      file  pcre.h,  and  on  Unix  systems  the library itself is  
      called libpcre.a, so can be accessed by adding -lpcre to the  
      command  for  linking  an  application  which  calls it. The  
      header file defines the macros PCRE_MAJOR and PCRE_MINOR  to  
      contain the major and minor release numbers for the library.  
      Applications can use these to include support for  different  
      releases.  
   
      The functions pcre_compile(), pcre_study(), and  pcre_exec()  
      are  used  for  compiling  and matching regular expressions,  
      while   pcre_copy_substring(),   pcre_get_substring(),   and  
      pcre_get_substring_list()   are  convenience  functions  for  
      extracting  captured  substrings  from  a  matched   subject  
      string.  The function pcre_maketables() is used (optionally)  
      to build a set of character tables in the current locale for  
      passing to pcre_compile().  
   
      The function pcre_fullinfo() is used to find out information  
      about a compiled pattern; pcre_info() is an obsolete version  
      which returns only some of the available information, but is  
      retained   for   backwards   compatibility.    The  function  
      pcre_version() returns a pointer to a string containing  the  
      version of PCRE and its date of release.  
   
      The global variables  pcre_malloc  and  pcre_free  initially  
      contain the entry points of the standard malloc() and free()  
      functions respectively. PCRE  calls  the  memory  management  
      functions  via  these  variables,  so  a calling program can  
      replace them if it  wishes  to  intercept  the  calls.  This  
      should be done before calling any PCRE functions.  
232    
233    UTF-8 SUPPORT
234    
235           To build PCRE with support for UTF-8 character strings, add
236    
237             --enable-utf8
238    
239           to  the  configure  command.  Of  itself, this does not make PCRE treat
240           strings as UTF-8. As well as compiling PCRE with this option, you  also
241           have  have to set the PCRE_UTF8 option when you call the pcre_compile()
242           function.
243    
244    
245    UNICODE CHARACTER PROPERTY SUPPORT
246    
247           UTF-8 support allows PCRE to process character values greater than  255
248           in  the  strings that it handles. On its own, however, it does not pro-
249           vide any facilities for accessing the properties of such characters. If
250           you  want  to  be able to use the pattern escapes \P, \p, and \X, which
251           refer to Unicode character properties, you must add
252    
253             --enable-unicode-properties
254    
255           to the configure command. This implies UTF-8 support, even if you  have
256           not explicitly requested it.
257    
258           Including  Unicode  property  support  adds around 90K of tables to the
259           PCRE library, approximately doubling its size. Only the  general  cate-
260           gory  properties  such as Lu and Nd are supported. Details are given in
261           the pcrepattern documentation.
262    
263    
264    CODE VALUE OF NEWLINE
265    
266           By default, PCRE treats character 10 (linefeed) as the newline  charac-
267           ter. This is the normal newline character on Unix-like systems. You can
268           compile PCRE to use character 13 (carriage return) instead by adding
269    
270             --enable-newline-is-cr
271    
272           to the configure command. For completeness there is  also  a  --enable-
273           newline-is-lf  option,  which explicitly specifies linefeed as the new-
274           line character.
275    
276    
277    BUILDING SHARED AND STATIC LIBRARIES
278    
279           The PCRE building process uses libtool to build both shared and  static
280           Unix  libraries by default. You can suppress one of these by adding one
281           of
282    
283             --disable-shared
284             --disable-static
285    
286           to the configure command, as required.
287    
288    
289    POSIX MALLOC USAGE
290    
291           When PCRE is called through the POSIX interface (see the pcreposix doc-
292           umentation),  additional  working  storage  is required for holding the
293           pointers to capturing substrings, because PCRE requires three  integers
294           per  substring,  whereas  the POSIX interface provides only two. If the
295           number of expected substrings is small, the wrapper function uses space
296           on the stack, because this is faster than using malloc() for each call.
297           The default threshold above which the stack is no longer used is 10; it
298           can be changed by adding a setting such as
299    
300             --with-posix-malloc-threshold=20
301    
302           to the configure command.
303    
304    
305    LIMITING PCRE RESOURCE USAGE
306    
307           Internally,  PCRE has a function called match(), which it calls repeat-
308           edly  (possibly  recursively)  when  matching  a   pattern   with   the
309           pcre_exec()  function.  By controlling the maximum number of times this
310           function may be called during a single matching operation, a limit  can
311           be  placed  on  the resources used by a single call to pcre_exec(). The
312           limit can be changed at run time, as described in the pcreapi  documen-
313           tation.  The default is 10 million, but this can be changed by adding a
314           setting such as
315    
316             --with-match-limit=500000
317    
318           to  the  configure  command.  This  setting  has  no  effect   on   the
319           pcre_dfa_exec() matching function.
320    
321    
322    HANDLING VERY LARGE PATTERNS
323    
324           Within  a  compiled  pattern,  offset values are used to point from one
325           part to another (for example, from an opening parenthesis to an  alter-
326           nation  metacharacter).  By default, two-byte values are used for these
327           offsets, leading to a maximum size for a  compiled  pattern  of  around
328           64K.  This  is sufficient to handle all but the most gigantic patterns.
329           Nevertheless, some people do want to process enormous patterns,  so  it
330           is  possible  to compile PCRE to use three-byte or four-byte offsets by
331           adding a setting such as
332    
333             --with-link-size=3
334    
335           to the configure command. The value given must be 2,  3,  or  4.  Using
336           longer  offsets slows down the operation of PCRE because it has to load
337           additional bytes when handling them.
338    
339           If you build PCRE with an increased link size, test 2 (and  test  5  if
340           you  are using UTF-8) will fail. Part of the output of these tests is a
341           representation of the compiled pattern, and this changes with the  link
342           size.
343    
344    
345    AVOIDING EXCESSIVE STACK USAGE
346    
347           When matching with the pcre_exec() function, PCRE implements backtrack-
348           ing by making recursive calls to an internal function  called  match().
349           In  environments  where  the size of the stack is limited, this can se-
350           verely limit PCRE's operation. (The Unix environment does  not  usually
351           suffer  from  this  problem.)  An alternative approach that uses memory
352           from the heap to remember data, instead  of  using  recursive  function
353           calls,  has been implemented to work round this problem. If you want to
354           build a version of PCRE that works this way, add
355    
356             --disable-stack-for-recursion
357    
358           to the configure command. With this configuration, PCRE  will  use  the
359           pcre_stack_malloc  and pcre_stack_free variables to call memory manage-
360           ment functions. Separate functions are provided because  the  usage  is
361           very  predictable:  the  block sizes requested are always the same, and
362           the blocks are always freed in reverse order. A calling  program  might
363           be  able  to implement optimized functions that perform better than the
364           standard malloc() and  free()  functions.  PCRE  runs  noticeably  more
365           slowly when built in this way. This option affects only the pcre_exec()
366           function; it is not relevant for the the pcre_dfa_exec() function.
367    
368    
369    USING EBCDIC CODE
370    
371           PCRE assumes by default that it will run in an  environment  where  the
372           character  code  is  ASCII  (or Unicode, which is a superset of ASCII).
373           PCRE can, however, be compiled to  run  in  an  EBCDIC  environment  by
374           adding
375    
376             --enable-ebcdic
377    
378           to the configure command.
379    
380    Last updated: 28 February 2005
381    Copyright (c) 1997-2005 University of Cambridge.
382    -----------------------------------------------------------------------------
383    
384    
385    
386    NAME
387           PCRE - Perl-compatible regular expressions
388    
389    
390    PCRE MATCHING ALGORITHMS
391    
392           This document describes the two different algorithms that are available
393           in PCRE for matching a compiled regular expression against a given sub-
394           ject  string.  The  "standard"  algorithm  is  the  one provided by the
395           pcre_exec() function.  This works in the same was  as  Perl's  matching
396           function, and provides a Perl-compatible matching operation.
397    
398           An  alternative  algorithm is provided by the pcre_dfa_exec() function;
399           this operates in a different way, and is not  Perl-compatible.  It  has
400           advantages  and disadvantages compared with the standard algorithm, and
401           these are described below.
402    
403           When there is only one possible way in which a given subject string can
404           match  a pattern, the two algorithms give the same answer. A difference
405           arises, however, when there are multiple possibilities. For example, if
406           the pattern
407    
408             ^<.*>
409    
410           is matched against the string
411    
412             <something> <something else> <something further>
413    
414           there are three possible answers. The standard algorithm finds only one
415           of them, whereas the DFA algorithm finds all three.
416    
417    
418    REGULAR EXPRESSIONS AS TREES
419    
420           The set of strings that are matched by a regular expression can be rep-
421           resented  as  a  tree structure. An unlimited repetition in the pattern
422           makes the tree of infinite size, but it is still a tree.  Matching  the
423           pattern  to a given subject string (from a given starting point) can be
424           thought of as a search of the tree.  There are  two  standard  ways  to
425           search  a  tree: depth-first and breadth-first, and these correspond to
426           the two matching algorithms provided by PCRE.
427    
428    
429    THE STANDARD MATCHING ALGORITHM
430    
431           In the terminology of Jeffrey Friedl's book Mastering  Regular  Expres-
432           sions,  the  standard  algorithm  is  an "NFA algorithm". It conducts a
433           depth-first search of the pattern tree. That is, it  proceeds  along  a
434           single path through the tree, checking that the subject matches what is
435           required. When there is a mismatch, the algorithm  tries  any  alterna-
436           tives  at  the  current point, and if they all fail, it backs up to the
437           previous branch point in the  tree,  and  tries  the  next  alternative
438           branch  at  that  level.  This often involves backing up (moving to the
439           left) in the subject string as well.  The  order  in  which  repetition
440           branches  are  tried  is controlled by the greedy or ungreedy nature of
441           the quantifier.
442    
443           If a leaf node is reached, a matching string has  been  found,  and  at
444           that  point the algorithm stops. Thus, if there is more than one possi-
445           ble match, this algorithm returns the first one that it finds.  Whether
446           this  is the shortest, the longest, or some intermediate length depends
447           on the way the greedy and ungreedy repetition quantifiers are specified
448           in the pattern.
449    
450           Because  it  ends  up  with a single path through the tree, it is rela-
451           tively straightforward for this algorithm to keep  track  of  the  sub-
452           strings  that  are  matched  by portions of the pattern in parentheses.
453           This provides support for capturing parentheses and back references.
454    
455    
456    THE DFA MATCHING ALGORITHM
457    
458           DFA stands for "deterministic finite automaton", but you do not need to
459           understand the origins of that name. This algorithm conducts a breadth-
460           first search of the tree. Starting from the first matching point in the
461           subject,  it scans the subject string from left to right, once, charac-
462           ter by character, and as it does  this,  it  remembers  all  the  paths
463           through the tree that represent valid matches.
464    
465           The  scan  continues until either the end of the subject is reached, or
466           there are no more unterminated paths. At this point,  terminated  paths
467           represent  the different matching possibilities (if there are none, the
468           match has failed).  Thus, if there is more  than  one  possible  match,
469           this algorithm finds all of them, and in particular, it finds the long-
470           est. In PCRE, there is an option to stop the algorithm after the  first
471           match (which is necessarily the shortest) has been found.
472    
473           Note that all the matches that are found start at the same point in the
474           subject. If the pattern
475    
476             cat(er(pillar)?)
477    
478           is matched against the string "the caterpillar catchment",  the  result
479           will  be the three strings "cat", "cater", and "caterpillar" that start
480           at the fourth character of the subject. The algorithm does not automat-
481           ically move on to find matches that start at later positions.
482    
483           There are a number of features of PCRE regular expressions that are not
484           supported by the DFA matching algorithm. They are as follows:
485    
486           1. Because the algorithm finds all  possible  matches,  the  greedy  or
487           ungreedy  nature  of repetition quantifiers is not relevant. Greedy and
488           ungreedy quantifiers are treated in exactly the same way.
489    
490           2. When dealing with multiple paths through the tree simultaneously, it
491           is  not  straightforward  to  keep track of captured substrings for the
492           different matching possibilities, and  PCRE's  implementation  of  this
493           algorithm does not attempt to do this. This means that no captured sub-
494           strings are available.
495    
496           3. Because no substrings are captured, back references within the  pat-
497           tern are not supported, and cause errors if encountered.
498    
499           4.  For  the same reason, conditional expressions that use a backrefer-
500           ence as the condition are not supported.
501    
502           5. Callouts are supported, but the value of the  capture_top  field  is
503           always 1, and the value of the capture_last field is always -1.
504    
505           6.  The \C escape sequence, which (in the standard algorithm) matches a
506           single byte, even in UTF-8 mode, is not supported because the DFA algo-
507           rithm moves through the subject string one character at a time, for all
508           active paths through the tree.
509    
510    
511    ADVANTAGES OF THE DFA ALGORITHM
512    
513           Using the DFA matching algorithm provides the following advantages:
514    
515           1. All possible matches (at a single point in the subject) are automat-
516           ically  found,  and  in particular, the longest match is found. To find
517           more than one match using the standard algorithm, you have to do kludgy
518           things with callouts.
519    
520           2.  There is much better support for partial matching. The restrictions
521           on the content of the pattern that apply when using the standard  algo-
522           rithm  for partial matching do not apply to the DFA algorithm. For non-
523           anchored patterns, the starting position of a partial match  is  avail-
524           able.
525    
526           3.  Because  the  DFA algorithm scans the subject string just once, and
527           never needs to backtrack, it is possible  to  pass  very  long  subject
528           strings  to  the matching function in several pieces, checking for par-
529           tial matching each time.
530    
531    
532    DISADVANTAGES OF THE DFA ALGORITHM
533    
534           The DFA algorithm suffers from a number of disadvantages:
535    
536           1. It is substantially slower than  the  standard  algorithm.  This  is
537           partly  because  it has to search for all possible matches, but is also
538           because it is less susceptible to optimization.
539    
540           2. Capturing parentheses and back references are not supported.
541    
542           3. The "atomic group" feature of PCRE regular expressions is supported,
543           but  does not provide the advantage that it does for the standard algo-
544           rithm.
545    
546    Last updated: 28 February 2005
547    Copyright (c) 1997-2005 University of Cambridge.
548    -----------------------------------------------------------------------------
549    
550    
551    
552    NAME
553           PCRE - Perl-compatible regular expressions
554    
555    
556    PCRE NATIVE API
557    
558           #include <pcre.h>
559    
560           pcre *pcre_compile(const char *pattern, int options,
561                const char **errptr, int *erroffset,
562                const unsigned char *tableptr);
563    
564           pcre *pcre_compile2(const char *pattern, int options,
565                int *errorcodeptr,
566                const char **errptr, int *erroffset,
567                const unsigned char *tableptr);
568    
569           pcre_extra *pcre_study(const pcre *code, int options,
570                const char **errptr);
571    
572           int pcre_exec(const pcre *code, const pcre_extra *extra,
573                const char *subject, int length, int startoffset,
574                int options, int *ovector, int ovecsize);
575    
576           int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
577                const char *subject, int length, int startoffset,
578                int options, int *ovector, int ovecsize,
579                int *workspace, int wscount);
580    
581           int pcre_copy_named_substring(const pcre *code,
582                const char *subject, int *ovector,
583                int stringcount, const char *stringname,
584                char *buffer, int buffersize);
585    
586           int pcre_copy_substring(const char *subject, int *ovector,
587                int stringcount, int stringnumber, char *buffer,
588                int buffersize);
589    
590           int pcre_get_named_substring(const pcre *code,
591                const char *subject, int *ovector,
592                int stringcount, const char *stringname,
593                const char **stringptr);
594    
595           int pcre_get_stringnumber(const pcre *code,
596                const char *name);
597    
598           int pcre_get_substring(const char *subject, int *ovector,
599                int stringcount, int stringnumber,
600                const char **stringptr);
601    
602           int pcre_get_substring_list(const char *subject,
603                int *ovector, int stringcount, const char ***listptr);
604    
605           void pcre_free_substring(const char *stringptr);
606    
607           void pcre_free_substring_list(const char **stringptr);
608    
609           const unsigned char *pcre_maketables(void);
610    
611  MULTI-THREADING         int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
612       The PCRE functions can be used in  multi-threading  applica-              int what, void *where);
      tions, with the proviso that the memory management functions  
      pointed to by pcre_malloc and pcre_free are  shared  by  all  
      threads.  
   
      The compiled form of a regular  expression  is  not  altered  
      during  matching, so the same compiled pattern can safely be  
      used by several threads at once.  
613    
614           int pcre_info(const pcre *code, int *optptr, int *firstcharptr);
615    
616           int pcre_refcount(pcre *code, int adjust);
617    
618           int pcre_config(int what, void *where);
619    
620           char *pcre_version(void);
621    
622           void *(*pcre_malloc)(size_t);
623    
624           void (*pcre_free)(void *);
625    
626           void *(*pcre_stack_malloc)(size_t);
627    
628           void (*pcre_stack_free)(void *);
629    
630           int (*pcre_callout)(pcre_callout_block *);
631    
632    
633    PCRE API OVERVIEW
634    
635           PCRE has its own native API, which is described in this document. There
636           is also a set of wrapper functions that correspond to the POSIX regular
637           expression  API.  These  are  described in the pcreposix documentation.
638           Both of these APIs define a set of C function calls. A C++  wrapper  is
639           distributed with PCRE. It is documented in the pcrecpp page.
640    
641           The  native  API  C  function prototypes are defined in the header file
642           pcre.h, and on Unix systems the library itself is called  libpcre.   It
643           can normally be accessed by adding -lpcre to the command for linking an
644           application  that  uses  PCRE.  The  header  file  defines  the  macros
645           PCRE_MAJOR  and  PCRE_MINOR to contain the major and minor release num-
646           bers for the library.  Applications can use these  to  include  support
647           for different releases of PCRE.
648    
649           The   functions   pcre_compile(),  pcre_compile2(),  pcre_study(),  and
650           pcre_exec() are used for compiling and matching regular expressions  in
651           a  Perl-compatible  manner. A sample program that demonstrates the sim-
652           plest way of using them is provided in the file  called  pcredemo.c  in
653           the  source distribution. The pcresample documentation describes how to
654           run it.
655    
656           A second matching function, pcre_dfa_exec(), which is not Perl-compati-
657           ble,  is  also provided. This uses a different algorithm for the match-
658           ing. This allows it to find all possible matches (at a given  point  in
659           the  subject),  not  just  one. However, this algorithm does not return
660           captured substrings. A description of the two matching  algorithms  and
661           their  advantages  and disadvantages is given in the pcrematching docu-
662           mentation.
663    
664           In addition to the main compiling and  matching  functions,  there  are
665           convenience functions for extracting captured substrings from a subject
666           string that is matched by pcre_exec(). They are:
667    
668             pcre_copy_substring()
669             pcre_copy_named_substring()
670             pcre_get_substring()
671             pcre_get_named_substring()
672             pcre_get_substring_list()
673             pcre_get_stringnumber()
674    
675           pcre_free_substring() and pcre_free_substring_list() are also provided,
676           to free the memory used for extracted strings.
677    
678           The  function  pcre_maketables()  is  used  to build a set of character
679           tables  in  the  current  locale   for   passing   to   pcre_compile(),
680           pcre_exec(),  or  pcre_dfa_exec(). This is an optional facility that is
681           provided for specialist use.  Most  commonly,  no  special  tables  are
682           passed,  in  which case internal tables that are generated when PCRE is
683           built are used.
684    
685           The function pcre_fullinfo() is used to find out  information  about  a
686           compiled  pattern; pcre_info() is an obsolete version that returns only
687           some of the available information, but is retained for  backwards  com-
688           patibility.   The function pcre_version() returns a pointer to a string
689           containing the version of PCRE and its date of release.
690    
691           The function pcre_refcount() maintains a  reference  count  in  a  data
692           block  containing  a compiled pattern. This is provided for the benefit
693           of object-oriented applications.
694    
695           The global variables pcre_malloc and pcre_free  initially  contain  the
696           entry  points  of  the  standard malloc() and free() functions, respec-
697           tively. PCRE calls the memory management functions via these variables,
698           so  a  calling  program  can replace them if it wishes to intercept the
699           calls. This should be done before calling any PCRE functions.
700    
701           The global variables pcre_stack_malloc  and  pcre_stack_free  are  also
702           indirections  to  memory  management functions. These special functions
703           are used only when PCRE is compiled to use  the  heap  for  remembering
704           data, instead of recursive function calls, when running the pcre_exec()
705           function. This is a non-standard way of building PCRE, for use in envi-
706           ronments that have limited stacks. Because of the greater use of memory
707           management, it runs more slowly.  Separate functions  are  provided  so
708           that  special-purpose  external  code  can  be used for this case. When
709           used, these functions are always called in a  stack-like  manner  (last
710           obtained,  first freed), and always for memory blocks of the same size.
711    
712           The global variable pcre_callout initially contains NULL. It can be set
713           by  the  caller  to  a "callout" function, which PCRE will then call at
714           specified points during a matching operation. Details are given in  the
715           pcrecallout documentation.
716    
717    
718    MULTITHREADING
719    
720           The  PCRE  functions  can be used in multi-threading applications, with
721           the  proviso  that  the  memory  management  functions  pointed  to  by
722           pcre_malloc, pcre_free, pcre_stack_malloc, and pcre_stack_free, and the
723           callout function pointed to by pcre_callout, are shared by all threads.
724    
725           The  compiled form of a regular expression is not altered during match-
726           ing, so the same compiled pattern can safely be used by several threads
727           at once.
728    
729    
730    SAVING PRECOMPILED PATTERNS FOR LATER USE
731    
732           The compiled form of a regular expression can be saved and re-used at a
733           later time, possibly by a different program, and even on a  host  other
734           than  the  one  on  which  it  was  compiled.  Details are given in the
735           pcreprecompile documentation.
736    
737    
738    CHECKING BUILD-TIME OPTIONS
739    
740           int pcre_config(int what, void *where);
741    
742           The function pcre_config() makes it possible for a PCRE client to  dis-
743           cover which optional features have been compiled into the PCRE library.
744           The pcrebuild documentation has more details about these optional  fea-
745           tures.
746    
747           The  first  argument  for pcre_config() is an integer, specifying which
748           information is required; the second argument is a pointer to a variable
749           into  which  the  information  is  placed. The following information is
750           available:
751    
752             PCRE_CONFIG_UTF8
753    
754           The output is an integer that is set to one if UTF-8 support is  avail-
755           able; otherwise it is set to zero.
756    
757             PCRE_CONFIG_UNICODE_PROPERTIES
758    
759           The  output  is  an  integer  that is set to one if support for Unicode
760           character properties is available; otherwise it is set to zero.
761    
762             PCRE_CONFIG_NEWLINE
763    
764           The output is an integer that is set to the value of the code  that  is
765           used  for the newline character. It is either linefeed (10) or carriage
766           return (13), and should normally be the  standard  character  for  your
767           operating system.
768    
769             PCRE_CONFIG_LINK_SIZE
770    
771           The  output  is  an  integer that contains the number of bytes used for
772           internal linkage in compiled regular expressions. The value is 2, 3, or
773           4.  Larger  values  allow larger regular expressions to be compiled, at
774           the expense of slower matching. The default value of  2  is  sufficient
775           for  all  but  the  most massive patterns, since it allows the compiled
776           pattern to be up to 64K in size.
777    
778             PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
779    
780           The output is an integer that contains the threshold  above  which  the
781           POSIX  interface  uses malloc() for output vectors. Further details are
782           given in the pcreposix documentation.
783    
784             PCRE_CONFIG_MATCH_LIMIT
785    
786           The output is an integer that gives the default limit for the number of
787           internal  matching  function  calls in a pcre_exec() execution. Further
788           details are given with pcre_exec() below.
789    
790             PCRE_CONFIG_STACKRECURSE
791    
792           The output is an integer that is set to one if internal recursion  when
793           running pcre_exec() is implemented by recursive function calls that use
794           the stack to remember their state. This is the usual way that  PCRE  is
795           compiled. The output is zero if PCRE was compiled to use blocks of data
796           on the  heap  instead  of  recursive  function  calls.  In  this  case,
797           pcre_stack_malloc  and  pcre_stack_free  are  called  to  manage memory
798           blocks on the heap, thus avoiding the use of the stack.
799    
800    
801  COMPILING A PATTERN  COMPILING A PATTERN
      The function pcre_compile() is called to compile  a  pattern  
      into  an internal form. The pattern is a C string terminated  
      by a binary zero, and is passed in the argument  pattern.  A  
      pointer  to  a  single  block of memory that is obtained via  
      pcre_malloc is returned. This contains the compiled code and  
      related data. The pcre type is defined for this for conveni-  
      ence, but in fact pcre is just a typedef for void, since the  
      contents  of  the block are not externally defined. It is up  
      to the caller to free  the  memory  when  it  is  no  longer  
      required.  
   
      The size of a compiled pattern is  roughly  proportional  to  
      the length of the pattern string, except that each character  
      class (other than those containing just a single  character,  
      negated  or  not)  requires 33 bytes, and repeat quantifiers  
      with a minimum greater than one or a bounded  maximum  cause  
      the  relevant  portions of the compiled pattern to be repli-  
      cated.  
   
      The options argument contains independent bits  that  affect  
      the  compilation.  It  should  be  zero  if  no  options are  
      required. Some of the options, in particular, those that are  
      compatible  with Perl, can also be set and unset from within  
      the pattern (see the detailed description of regular expres-  
      sions below). For these options, the contents of the options  
      argument specifies their initial settings at  the  start  of  
      compilation  and  execution. The PCRE_ANCHORED option can be  
      set at the time of matching as well as at compile time.  
   
      If errptr is NULL, pcre_compile() returns NULL  immediately.  
      Otherwise, if compilation of a pattern fails, pcre_compile()  
      returns NULL, and sets the variable pointed to by errptr  to  
      point  to a textual error message. The offset from the start  
      of  the  pattern  to  the  character  where  the  error  was  
      discovered   is   placed  in  the  variable  pointed  to  by  
      erroffset, which must not be NULL. If it  is,  an  immediate  
      error is given.  
   
      If the final  argument,  tableptr,  is  NULL,  PCRE  uses  a  
      default  set  of character tables which are built when it is  
      compiled, using the default C  locale.  Otherwise,  tableptr  
      must  be  the result of a call to pcre_maketables(). See the  
      section on locale support below.  
   
      The following option bits are defined in the header file:  
   
        PCRE_ANCHORED  
   
      If this bit is set, the pattern is forced to be  "anchored",  
      that is, it is constrained to match only at the start of the  
      string which is being searched (the "subject string").  This  
      effect can also be achieved by appropriate constructs in the  
      pattern itself, which is the only way to do it in Perl.  
   
        PCRE_CASELESS  
   
      If this bit is set, letters in the pattern match both  upper  
      and  lower  case  letters.  It  is  equivalent  to Perl's /i  
      option.  
   
        PCRE_DOLLAR_ENDONLY  
   
      If this bit is set, a dollar metacharacter  in  the  pattern  
      matches  only at the end of the subject string. Without this  
      option, a dollar also matches immediately before  the  final  
      character  if it is a newline (but not before any other new-  
      lines).  The  PCRE_DOLLAR_ENDONLY  option  is   ignored   if  
      PCRE_MULTILINE is set. There is no equivalent to this option  
      in Perl.  
   
        PCRE_DOTALL  
   
      If this bit is  set,  a  dot  metacharater  in  the  pattern  
      matches all characters, including newlines. Without it, new-  
      lines are excluded. This option is equivalent to  Perl's  /s  
      option.  A negative class such as [^a] always matches a new-  
      line character, independent of the setting of this option.  
   
        PCRE_EXTENDED  
   
      If this bit is set, whitespace data characters in  the  pat-  
      tern  are  totally  ignored  except when escaped or inside a  
      character class, and characters between an unescaped #  out-  
      side  a  character  class  and  the  next newline character,  
      inclusive, are also ignored. This is equivalent to Perl's /x  
      option,  and  makes  it  possible to include comments inside  
      complicated patterns. Note, however, that this applies  only  
      to  data  characters. Whitespace characters may never appear  
      within special character sequences in a pattern, for example  
      within  the sequence (?( which introduces a conditional sub-  
      pattern.  
   
        PCRE_EXTRA  
   
      This option was invented in  order  to  turn  on  additional  
      functionality of PCRE that is incompatible with Perl, but it  
      is currently of very little use. When set, any backslash  in  
      a  pattern  that is followed by a letter that has no special  
      meaning causes an error, thus reserving  these  combinations  
      for  future  expansion.  By default, as in Perl, a backslash  
      followed by a letter with no special meaning is treated as a  
      literal.  There  are at present no other features controlled  
      by this option. It can also be set by a (?X) option  setting  
      within a pattern.  
   
        PCRE_MULTILINE  
   
      By default, PCRE treats the subject string as consisting  of  
      a  single "line" of characters (even if it actually contains  
      several newlines). The "start  of  line"  metacharacter  (^)  
      matches  only  at the start of the string, while the "end of  
      line" metacharacter ($) matches  only  at  the  end  of  the  
      string,    or   before   a   terminating   newline   (unless  
      PCRE_DOLLAR_ENDONLY is set). This is the same as Perl.  
   
      When PCRE_MULTILINE it is set, the "start of line" and  "end  
      of  line"  constructs match immediately following or immedi-  
      ately before any newline  in  the  subject  string,  respec-  
      tively,  as  well  as  at  the  very  start and end. This is  
      equivalent to Perl's /m option. If there are no "\n" charac-  
      ters  in  a subject string, or no occurrences of ^ or $ in a  
      pattern, setting PCRE_MULTILINE has no effect.  
   
        PCRE_UNGREEDY  
   
      This option inverts the "greediness" of the  quantifiers  so  
      that  they  are  not greedy by default, but become greedy if  
      followed by "?". It is not compatible with Perl. It can also  
      be set by a (?U) option setting within the pattern.  
802    
803           pcre *pcre_compile(const char *pattern, int options,
804                const char **errptr, int *erroffset,
805                const unsigned char *tableptr);
806    
807           pcre *pcre_compile2(const char *pattern, int options,
808                int *errorcodeptr,
809                const char **errptr, int *erroffset,
810                const unsigned char *tableptr);
811    
812           Either of the functions pcre_compile() or pcre_compile2() can be called
813           to compile a pattern into an internal form. The only difference between
814           the two interfaces is that pcre_compile2() has an additional  argument,
815           errorcodeptr, via which a numerical error code can be returned.
816    
817           The pattern is a C string terminated by a binary zero, and is passed in
818           the pattern argument. A pointer to a single block  of  memory  that  is
819           obtained  via  pcre_malloc is returned. This contains the compiled code
820           and related data. The pcre type is defined for the returned block; this
821           is a typedef for a structure whose contents are not externally defined.
822           It is up to the caller  to  free  the  memory  when  it  is  no  longer
823           required.
824    
825           Although  the compiled code of a PCRE regex is relocatable, that is, it
826           does not depend on memory location, the complete pcre data block is not
827           fully  relocatable, because it may contain a copy of the tableptr argu-
828           ment, which is an address (see below).
829    
830           The options argument contains independent bits that affect the compila-
831           tion.  It  should  be  zero  if  no options are required. The available
832           options are described below. Some of them, in  particular,  those  that
833           are  compatible  with  Perl,  can also be set and unset from within the
834           pattern (see the detailed description  in  the  pcrepattern  documenta-
835           tion).  For  these options, the contents of the options argument speci-
836           fies their initial settings at the start of compilation and  execution.
837           The  PCRE_ANCHORED option can be set at the time of matching as well as
838           at compile time.
839    
840           If errptr is NULL, pcre_compile() returns NULL immediately.  Otherwise,
841           if  compilation  of  a  pattern fails, pcre_compile() returns NULL, and
842           sets the variable pointed to by errptr to point to a textual error mes-
843           sage.  The  offset from the start of the pattern to the character where
844           the error was discovered is  placed  in  the  variable  pointed  to  by
845           erroffset,  which  must  not  be  NULL. If it is, an immediate error is
846           given.
847    
848           If pcre_compile2() is used instead of pcre_compile(),  and  the  error-
849           codeptr  argument is not NULL, a non-zero error code number is returned
850           via this argument in the event of an error. This is in addition to  the
851           textual error message. Error codes and messages are listed below.
852    
853           If  the  final  argument, tableptr, is NULL, PCRE uses a default set of
854           character tables that are  built  when  PCRE  is  compiled,  using  the
855           default  C  locale.  Otherwise, tableptr must be an address that is the
856           result of a call to pcre_maketables(). This value is  stored  with  the
857           compiled  pattern,  and used again by pcre_exec(), unless another table
858           pointer is passed to it. For more discussion, see the section on locale
859           support below.
860    
861           This  code  fragment  shows a typical straightforward call to pcre_com-
862           pile():
863    
864             pcre *re;
865             const char *error;
866             int erroffset;
867             re = pcre_compile(
868               "^A.*Z",          /* the pattern */
869               0,                /* default options */
870               &error,           /* for error message */
871               &erroffset,       /* for error offset */
872               NULL);            /* use default character tables */
873    
874           The following names for option bits are defined in  the  pcre.h  header
875           file:
876    
877             PCRE_ANCHORED
878    
879           If this bit is set, the pattern is forced to be "anchored", that is, it
880           is constrained to match only at the first matching point in the  string
881           that  is being searched (the "subject string"). This effect can also be
882           achieved by appropriate constructs in the pattern itself, which is  the
883           only way to do it in Perl.
884    
885             PCRE_AUTO_CALLOUT
886    
887           If this bit is set, pcre_compile() automatically inserts callout items,
888           all with number 255, before each pattern item. For  discussion  of  the
889           callout facility, see the pcrecallout documentation.
890    
891             PCRE_CASELESS
892    
893           If  this  bit is set, letters in the pattern match both upper and lower
894           case letters. It is equivalent to Perl's  /i  option,  and  it  can  be
895           changed  within a pattern by a (?i) option setting. In UTF-8 mode, PCRE
896           always understands the concept of case for characters whose values  are
897           less  than 128, so caseless matching is always possible. For characters
898           with higher values, the concept of case is supported if  PCRE  is  com-
899           piled  with Unicode property support, but not otherwise. If you want to
900           use caseless matching for characters 128 and  above,  you  must  ensure
901           that  PCRE  is  compiled  with Unicode property support as well as with
902           UTF-8 support.
903    
904             PCRE_DOLLAR_ENDONLY
905    
906           If this bit is set, a dollar metacharacter in the pattern matches  only
907           at  the  end  of the subject string. Without this option, a dollar also
908           matches immediately before the final character if it is a newline  (but
909           not  before  any  other  newlines).  The  PCRE_DOLLAR_ENDONLY option is
910           ignored if PCRE_MULTILINE is set. There is no equivalent to this option
911           in Perl, and no way to set it within a pattern.
912    
913             PCRE_DOTALL
914    
915           If this bit is set, a dot metacharater in the pattern matches all char-
916           acters, including newlines. Without it,  newlines  are  excluded.  This
917           option  is equivalent to Perl's /s option, and it can be changed within
918           a pattern by a (?s) option setting.  A  negative  class  such  as  [^a]
919           always  matches a newline character, independent of the setting of this
920           option.
921    
922             PCRE_EXTENDED
923    
924           If this bit is set, whitespace  data  characters  in  the  pattern  are
925           totally ignored except when escaped or inside a character class. White-
926           space does not include the VT character (code 11). In addition, charac-
927           ters between an unescaped # outside a character class and the next new-
928           line character, inclusive, are also  ignored.  This  is  equivalent  to
929           Perl's  /x  option,  and  it  can be changed within a pattern by a (?x)
930           option setting.
931    
932           This option makes it possible to include  comments  inside  complicated
933           patterns.   Note,  however,  that this applies only to data characters.
934           Whitespace  characters  may  never  appear  within  special   character
935           sequences  in  a  pattern,  for  example  within the sequence (?( which
936           introduces a conditional subpattern.
937    
938             PCRE_EXTRA
939    
940           This option was invented in order to turn on  additional  functionality
941           of  PCRE  that  is  incompatible with Perl, but it is currently of very
942           little use. When set, any backslash in a pattern that is followed by  a
943           letter  that  has  no  special  meaning causes an error, thus reserving
944           these combinations for future expansion. By  default,  as  in  Perl,  a
945           backslash  followed by a letter with no special meaning is treated as a
946           literal. There are at present no  other  features  controlled  by  this
947           option. It can also be set by a (?X) option setting within a pattern.
948    
949             PCRE_FIRSTLINE
950    
951           If  this  option  is  set,  an  unanchored pattern is required to match
952           before or at the first newline character in the subject string,  though
953           the matched text may continue over the newline.
954    
955             PCRE_MULTILINE
956    
957           By  default,  PCRE  treats the subject string as consisting of a single
958           line of characters (even if it actually contains newlines). The  "start
959           of  line"  metacharacter  (^)  matches only at the start of the string,
960           while the "end of line" metacharacter ($) matches only at  the  end  of
961           the string, or before a terminating newline (unless PCRE_DOLLAR_ENDONLY
962           is set). This is the same as Perl.
963    
964           When PCRE_MULTILINE it is set, the "start of line" and  "end  of  line"
965           constructs  match  immediately following or immediately before any new-
966           line in the subject string, respectively, as well as at the very  start
967           and  end. This is equivalent to Perl's /m option, and it can be changed
968           within a pattern by a (?m) option setting. If there are no "\n" charac-
969           ters  in  a  subject  string, or no occurrences of ^ or $ in a pattern,
970           setting PCRE_MULTILINE has no effect.
971    
972             PCRE_NO_AUTO_CAPTURE
973    
974           If this option is set, it disables the use of numbered capturing paren-
975           theses  in the pattern. Any opening parenthesis that is not followed by
976           ? behaves as if it were followed by ?: but named parentheses can  still
977           be  used  for  capturing  (and  they acquire numbers in the usual way).
978           There is no equivalent of this option in Perl.
979    
980             PCRE_UNGREEDY
981    
982           This option inverts the "greediness" of the quantifiers  so  that  they
983           are  not greedy by default, but become greedy if followed by "?". It is
984           not compatible with Perl. It can also be set by a (?U)  option  setting
985           within the pattern.
986    
987             PCRE_UTF8
988    
989           This  option  causes PCRE to regard both the pattern and the subject as
990           strings of UTF-8 characters instead of single-byte  character  strings.
991           However,  it is available only when PCRE is built to include UTF-8 sup-
992           port. If not, the use of this option provokes an error. Details of  how
993           this  option  changes the behaviour of PCRE are given in the section on
994           UTF-8 support in the main pcre page.
995    
996             PCRE_NO_UTF8_CHECK
997    
998           When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
999           automatically  checked. If an invalid UTF-8 sequence of bytes is found,
1000           pcre_compile() returns an error. If you already know that your  pattern
1001           is  valid, and you want to skip this check for performance reasons, you
1002           can set the PCRE_NO_UTF8_CHECK option. When it is set,  the  effect  of
1003           passing an invalid UTF-8 string as a pattern is undefined. It may cause
1004           your program to crash.  Note that this option can  also  be  passed  to
1005           pcre_exec()  and pcre_dfa_exec(), to suppress the UTF-8 validity check-
1006           ing of subject strings.
1007    
1008    
1009    COMPILATION ERROR CODES
1010    
1011           The following table lists the error  codes  than  may  be  returned  by
1012           pcre_compile2(),  along with the error messages that may be returned by
1013           both compiling functions.
1014    
1015              0  no error
1016              1  \ at end of pattern
1017              2  \c at end of pattern
1018              3  unrecognized character follows \
1019              4  numbers out of order in {} quantifier
1020              5  number too big in {} quantifier
1021              6  missing terminating ] for character class
1022              7  invalid escape sequence in character class
1023              8  range out of order in character class
1024              9  nothing to repeat
1025             10  operand of unlimited repeat could match the empty string
1026             11  internal error: unexpected repeat
1027             12  unrecognized character after (?
1028             13  POSIX named classes are supported only within a class
1029             14  missing )
1030             15  reference to non-existent subpattern
1031             16  erroffset passed as NULL
1032             17  unknown option bit(s) set
1033             18  missing ) after comment
1034             19  parentheses nested too deeply
1035             20  regular expression too large
1036             21  failed to get memory
1037             22  unmatched parentheses
1038             23  internal error: code overflow
1039             24  unrecognized character after (?<
1040             25  lookbehind assertion is not fixed length
1041             26  malformed number after (?(
1042             27  conditional group contains more than two branches
1043             28  assertion expected after (?(
1044             29  (?R or (?digits must be followed by )
1045             30  unknown POSIX class name
1046             31  POSIX collating elements are not supported
1047             32  this version of PCRE is not compiled with PCRE_UTF8 support
1048             33  spare error
1049             34  character value in \x{...} sequence is too large
1050             35  invalid condition (?(0)
1051             36  \C not allowed in lookbehind assertion
1052             37  PCRE does not support \L, \l, \N, \U, or \u
1053             38  number after (?C is > 255
1054             39  closing ) for (?C expected
1055             40  recursive call could loop indefinitely
1056             41  unrecognized character after (?P
1057             42  syntax error after (?P
1058             43  two named groups have the same name
1059             44  invalid UTF-8 string
1060             45  support for \P, \p, and \X has not been compiled
1061             46  malformed \P or \p sequence
1062             47  unknown property name after \P or \p
1063    
1064    
1065  STUDYING A PATTERN  STUDYING A PATTERN
      When a pattern is going to be  used  several  times,  it  is  
      worth  spending  more time analyzing it in order to speed up  
      the time taken for matching. The function pcre_study() takes  
      a  pointer  to a compiled pattern as its first argument, and  
      returns a  pointer  to  a  pcre_extra  block  (another  void  
      typedef)  containing  additional  information about the pat-  
      tern; this can be passed to pcre_exec().  If  no  additional  
      information is available, NULL is returned.  
   
      The second argument contains option  bits.  At  present,  no  
      options  are  defined  for  pcre_study(),  and this argument  
      should always be zero.  
   
      The third argument for pcre_study() is a pointer to an error  
      message. If studying succeeds (even if no data is returned),  
      the variable it points to  is  set  to  NULL.  Otherwise  it  
      points to a textual error message.  
   
      At present, studying a  pattern  is  useful  only  for  non-  
      anchored  patterns  that do not have a single fixed starting  
      character. A  bitmap  of  possible  starting  characters  is  
      created.  
1066    
1067           pcre_extra *pcre_study(const pcre *code, int options
1068                const char **errptr);
1069    
1070           If a compiled pattern is going to be used several times,  it  is  worth
1071           spending more time analyzing it in order to speed up the time taken for
1072           matching. The function pcre_study() takes a pointer to a compiled  pat-
1073           tern as its first argument. If studying the pattern produces additional
1074           information that will help speed up matching,  pcre_study()  returns  a
1075           pointer  to a pcre_extra block, in which the study_data field points to
1076           the results of the study.
1077    
1078           The  returned  value  from  pcre_study()  can  be  passed  directly  to
1079           pcre_exec().  However,  a  pcre_extra  block also contains other fields
1080           that can be set by the caller before the block  is  passed;  these  are
1081           described below in the section on matching a pattern.
1082    
1083           If  studying  the  pattern  does not produce any additional information
1084           pcre_study() returns NULL. In that circumstance, if the calling program
1085           wants  to  pass  any of the other fields to pcre_exec(), it must set up
1086           its own pcre_extra block.
1087    
1088           The second argument of pcre_study() contains option bits.  At  present,
1089           no options are defined, and this argument should always be zero.
1090    
1091           The  third argument for pcre_study() is a pointer for an error message.
1092           If studying succeeds (even if no data is  returned),  the  variable  it
1093           points  to  is set to NULL. Otherwise it points to a textual error mes-
1094           sage. You should therefore test the error pointer for NULL after  call-
1095           ing pcre_study(), to be sure that it has run successfully.
1096    
1097           This is a typical call to pcre_study():
1098    
1099             pcre_extra *pe;
1100             pe = pcre_study(
1101               re,             /* result of pcre_compile() */
1102               0,              /* no options exist */
1103               &error);        /* set to NULL or points to a message */
1104    
1105           At present, studying a pattern is useful only for non-anchored patterns
1106           that do not have a single fixed starting character. A bitmap of  possi-
1107           ble starting bytes is created.
1108    
1109    
1110  LOCALE SUPPORT  LOCALE SUPPORT
      PCRE handles caseless matching, and determines whether char-  
      acters  are  letters, digits, or whatever, by reference to a  
      set of tables. The library contains a default set of  tables  
      which  is  created in the default C locale when PCRE is com-  
      piled.  This  is   used   when   the   final   argument   of  
      pcre_compile()  is NULL, and is sufficient for many applica-  
      tions.  
   
      An alternative set of tables can, however, be supplied. Such  
      tables  are built by calling the pcre_maketables() function,  
      which has no arguments, in the relevant locale.  The  result  
      can  then be passed to pcre_compile() as often as necessary.  
      For example, to build and use tables  that  are  appropriate  
      for  the French locale (where accented characters with codes  
      greater than 128 are treated as letters), the following code  
      could be used:  
   
        setlocale(LC_CTYPE, "fr");  
        tables = pcre_maketables();  
        re = pcre_compile(..., tables);  
   
      The  tables  are  built  in  memory  that  is  obtained  via  
      pcre_malloc.  The  pointer that is passed to pcre_compile is  
      saved with the compiled pattern, and  the  same  tables  are  
      used  via this pointer by pcre_study() and pcre_exec(). Thus  
      for any single pattern, compilation, studying  and  matching  
      all happen in the same locale, but different patterns can be  
      compiled in different locales. It is the caller's  responsi-  
      bility  to  ensure  that  the  memory  containing the tables  
      remains available for as long as it is needed.  
1111    
1112           PCRE  handles  caseless matching, and determines whether characters are
1113           letters digits, or whatever, by reference to a set of  tables,  indexed
1114           by  character  value.  When running in UTF-8 mode, this applies only to
1115           characters with codes less than 128. Higher-valued  codes  never  match
1116           escapes  such  as  \w or \d, but can be tested with \p if PCRE is built
1117           with Unicode character property support.
1118    
1119           An internal set of tables is created in the default C locale when  PCRE
1120           is  built.  This  is  used when the final argument of pcre_compile() is
1121           NULL, and is sufficient for many applications. An  alternative  set  of
1122           tables  can,  however, be supplied. These may be created in a different
1123           locale from the default. As more and more applications change to  using
1124           Unicode, the need for this locale support is expected to die away.
1125    
1126           External  tables  are  built by calling the pcre_maketables() function,
1127           which has no arguments, in the relevant locale. The result can then  be
1128           passed  to  pcre_compile()  or  pcre_exec()  as often as necessary. For
1129           example, to build and use tables that are appropriate  for  the  French
1130           locale  (where  accented  characters  with  values greater than 128 are
1131           treated as letters), the following code could be used:
1132    
1133             setlocale(LC_CTYPE, "fr_FR");
1134             tables = pcre_maketables();
1135             re = pcre_compile(..., tables);
1136    
1137           When pcre_maketables() runs, the tables are built  in  memory  that  is
1138           obtained  via  pcre_malloc. It is the caller's responsibility to ensure
1139           that the memory containing the tables remains available for as long  as
1140           it is needed.
1141    
1142           The pointer that is passed to pcre_compile() is saved with the compiled
1143           pattern, and the same tables are used via this pointer by  pcre_study()
1144           and normally also by pcre_exec(). Thus, by default, for any single pat-
1145           tern, compilation, studying and matching all happen in the same locale,
1146           but different patterns can be compiled in different locales.
1147    
1148           It  is  possible to pass a table pointer or NULL (indicating the use of
1149           the internal tables) to pcre_exec(). Although  not  intended  for  this
1150           purpose,  this facility could be used to match a pattern in a different
1151           locale from the one in which it was compiled. Passing table pointers at
1152           run time is discussed below in the section on matching a pattern.
1153    
1154    
1155  INFORMATION ABOUT A PATTERN  INFORMATION ABOUT A PATTERN
      The pcre_fullinfo() function  returns  information  about  a  
      compiled pattern. It replaces the obsolete pcre_info() func-  
      tion, which is nevertheless retained for backwards compabil-  
      ity (and is documented below).  
   
      The first argument for pcre_fullinfo() is a pointer  to  the  
      compiled  pattern.  The  second  argument  is  the result of  
      pcre_study(), or NULL if the pattern was  not  studied.  The  
      third  argument  specifies  which  piece  of  information is  
      required, while the fourth argument is a pointer to a  vari-  
      able  to receive the data. The yield of the function is zero  
      for success, or one of the following negative numbers:  
   
        PCRE_ERROR_NULL       the argument code was NULL  
                              the argument where was NULL  
        PCRE_ERROR_BADMAGIC   the "magic number" was not found  
        PCRE_ERROR_BADOPTION  the value of what was invalid  
   
      The possible values for the third argument  are  defined  in  
      pcre.h, and are as follows:  
   
        PCRE_INFO_OPTIONS  
   
      Return a copy of the options with which the pattern was com-  
      piled.  The fourth argument should point to au unsigned long  
      int variable. These option bits are those specified  in  the  
      call  to  pcre_compile(),  modified  by any top-level option  
      settings  within  the   pattern   itself,   and   with   the  
      PCRE_ANCHORED  bit  forcibly  set if the form of the pattern  
      implies that it can match only at the  start  of  a  subject  
      string.  
   
        PCRE_INFO_SIZE  
   
      Return the size of the compiled pattern, that is, the  value  
      that  was  passed as the argument to pcre_malloc() when PCRE  
      was getting memory in which to place the compiled data.  The  
      fourth argument should point to a size_t variable.  
   
        PCRE_INFO_CAPTURECOUNT  
   
      Return the number of capturing subpatterns in  the  pattern.  
      The fourth argument should point to an int variable.  
   
        PCRE_INFO_BACKREFMAX  
   
      Return the number of the highest back reference in the  pat-  
      tern.  The  fourth argument should point to an int variable.  
      Zero is returned if there are no back references.  
   
        PCRE_INFO_FIRSTCHAR  
   
      Return information about the first character of any  matched  
      string,  for  a  non-anchored  pattern.  If there is a fixed  
      first   character,   e.g.   from   a   pattern    such    as  
      (cat|cow|coyote),  it  is returned in the integer pointed to  
      by where. Otherwise, if either  
   
      (a) the pattern was compiled with the PCRE_MULTILINE option,  
      and every branch starts with "^", or  
   
      (b) every  branch  of  the  pattern  starts  with  ".*"  and  
      PCRE_DOTALL is not set (if it were set, the pattern would be  
      anchored),  
   
      -1 is returned, indicating that the pattern matches only  at  
      the  start  of a subject string or after any "\n" within the  
      string. Otherwise -2 is returned.  For anchored patterns, -2  
      is returned.  
   
        PCRE_INFO_FIRSTTABLE  
   
      If the pattern was studied, and this resulted  in  the  con-  
      struction of a 256-bit table indicating a fixed set of char-  
      acters for the first character in  any  matching  string,  a  
      pointer   to  the  table  is  returned.  Otherwise  NULL  is  
      returned. The fourth argument should point  to  an  unsigned  
      char * variable.  
   
        PCRE_INFO_LASTLITERAL  
   
      For a non-anchored pattern, return the value of  the  right-  
      most  literal  character  which  must  exist  in any matched  
      string, other than at its start. The fourth argument  should  
      point  to an int variable. If there is no such character, or  
      if the pattern is anchored, -1 is returned. For example, for  
      the pattern /a\d+z\d+/ the returned value is 'z'.  
   
      The pcre_info() function is now obsolete because its  inter-  
      face  is  too  restrictive  to return all the available data  
      about  a  compiled  pattern.   New   programs   should   use  
      pcre_fullinfo()  instead.  The  yield  of pcre_info() is the  
      number of capturing subpatterns, or  one  of  the  following  
      negative numbers:  
   
        PCRE_ERROR_NULL       the argument code was NULL  
        PCRE_ERROR_BADMAGIC   the "magic number" was not found  
   
      If the optptr argument is not NULL, a copy  of  the  options  
      with which the pattern was compiled is placed in the integer  
      it points to (see PCRE_INFO_OPTIONS above).  
   
      If the pattern is not anchored and the firstcharptr argument  
      is  not  NULL, it is used to pass back information about the  
      first    character    of    any    matched    string    (see  
      PCRE_INFO_FIRSTCHAR above).  
1156    
1157           int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
1158                int what, void *where);
1159    
1160           The  pcre_fullinfo() function returns information about a compiled pat-
1161           tern. It replaces the obsolete pcre_info() function, which is neverthe-
1162           less retained for backwards compability (and is documented below).
1163    
1164           The  first  argument  for  pcre_fullinfo() is a pointer to the compiled
1165           pattern. The second argument is the result of pcre_study(), or NULL  if
1166           the  pattern  was not studied. The third argument specifies which piece
1167           of information is required, and the fourth argument is a pointer  to  a
1168           variable  to  receive  the  data. The yield of the function is zero for
1169           success, or one of the following negative numbers:
1170    
1171             PCRE_ERROR_NULL       the argument code was NULL
1172                                   the argument where was NULL
1173             PCRE_ERROR_BADMAGIC   the "magic number" was not found
1174             PCRE_ERROR_BADOPTION  the value of what was invalid
1175    
1176           The "magic number" is placed at the start of each compiled  pattern  as
1177           an  simple check against passing an arbitrary memory pointer. Here is a
1178           typical call of pcre_fullinfo(), to obtain the length of  the  compiled
1179           pattern:
1180    
1181             int rc;
1182             unsigned long int length;
1183             rc = pcre_fullinfo(
1184               re,               /* result of pcre_compile() */
1185               pe,               /* result of pcre_study(), or NULL */
1186               PCRE_INFO_SIZE,   /* what is required */
1187               &length);         /* where to put the data */
1188    
1189           The  possible  values for the third argument are defined in pcre.h, and
1190           are as follows:
1191    
1192             PCRE_INFO_BACKREFMAX
1193    
1194           Return the number of the highest back reference  in  the  pattern.  The
1195           fourth  argument  should  point to an int variable. Zero is returned if
1196           there are no back references.
1197    
1198             PCRE_INFO_CAPTURECOUNT
1199    
1200           Return the number of capturing subpatterns in the pattern.  The  fourth
1201           argument should point to an int variable.
1202    
1203             PCRE_INFO_DEFAULT_TABLES
1204    
1205           Return  a pointer to the internal default character tables within PCRE.
1206           The fourth argument should point to an unsigned char *  variable.  This
1207           information call is provided for internal use by the pcre_study() func-
1208           tion. External callers can cause PCRE to use  its  internal  tables  by
1209           passing a NULL table pointer.
1210    
1211             PCRE_INFO_FIRSTBYTE
1212    
1213           Return  information  about  the first byte of any matched string, for a
1214           non-anchored   pattern.   (This    option    used    to    be    called
1215           PCRE_INFO_FIRSTCHAR;  the  old  name  is still recognized for backwards
1216           compatibility.)
1217    
1218           If there is a fixed first byte, for example, from  a  pattern  such  as
1219           (cat|cow|coyote),  it  is  returned in the integer pointed to by where.
1220           Otherwise, if either
1221    
1222           (a) the pattern was compiled with the PCRE_MULTILINE option, and  every
1223           branch starts with "^", or
1224    
1225           (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not
1226           set (if it were set, the pattern would be anchored),
1227    
1228           -1 is returned, indicating that the pattern matches only at  the  start
1229           of  a  subject string or after any newline within the string. Otherwise
1230           -2 is returned. For anchored patterns, -2 is returned.
1231    
1232             PCRE_INFO_FIRSTTABLE
1233    
1234           If the pattern was studied, and this resulted in the construction of  a
1235           256-bit table indicating a fixed set of bytes for the first byte in any
1236           matching string, a pointer to the table is returned. Otherwise NULL  is
1237           returned.  The fourth argument should point to an unsigned char * vari-
1238           able.
1239    
1240             PCRE_INFO_LASTLITERAL
1241    
1242           Return the value of the rightmost literal byte that must exist  in  any
1243           matched  string,  other  than  at  its  start,  if such a byte has been
1244           recorded. The fourth argument should point to an int variable. If there
1245           is  no such byte, -1 is returned. For anchored patterns, a last literal
1246           byte is recorded only if it follows something of variable  length.  For
1247           example, for the pattern /^a\d+z\d+/ the returned value is "z", but for
1248           /^a\dz\d/ the returned value is -1.
1249    
1250             PCRE_INFO_NAMECOUNT
1251             PCRE_INFO_NAMEENTRYSIZE
1252             PCRE_INFO_NAMETABLE
1253    
1254           PCRE supports the use of named as well as numbered capturing  parenthe-
1255           ses.  The names are just an additional way of identifying the parenthe-
1256           ses,  which  still  acquire  numbers.  A  convenience  function  called
1257           pcre_get_named_substring()  is  provided  for  extracting an individual
1258           captured substring by name. It is also possible  to  extract  the  data
1259           directly,  by  first converting the name to a number in order to access
1260           the correct pointers in the output vector (described  with  pcre_exec()
1261           below).  To  do the conversion, you need to use the name-to-number map,
1262           which is described by these three values.
1263    
1264           The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT
1265           gives the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size
1266           of each entry; both of these  return  an  int  value.  The  entry  size
1267           depends  on the length of the longest name. PCRE_INFO_NAMETABLE returns
1268           a pointer to the first entry of the table  (a  pointer  to  char).  The
1269           first two bytes of each entry are the number of the capturing parenthe-
1270           sis, most significant byte first. The rest of the entry is  the  corre-
1271           sponding  name,  zero  terminated. The names are in alphabetical order.
1272           For example, consider the following pattern  (assume  PCRE_EXTENDED  is
1273           set, so white space - including newlines - is ignored):
1274    
1275             (?P<date> (?P<year>(\d\d)?\d\d) -
1276             (?P<month>\d\d) - (?P<day>\d\d) )
1277    
1278           There  are  four  named subpatterns, so the table has four entries, and
1279           each entry in the table is eight bytes long. The table is  as  follows,
1280           with non-printing bytes shows in hexadecimal, and undefined bytes shown
1281           as ??:
1282    
1283             00 01 d  a  t  e  00 ??
1284             00 05 d  a  y  00 ?? ??
1285             00 04 m  o  n  t  h  00
1286             00 02 y  e  a  r  00 ??
1287    
1288           When writing code to extract data  from  named  subpatterns  using  the
1289           name-to-number map, remember that the length of each entry is likely to
1290           be different for each compiled pattern.
1291    
1292             PCRE_INFO_OPTIONS
1293    
1294           Return a copy of the options with which the pattern was  compiled.  The
1295           fourth  argument  should  point to an unsigned long int variable. These
1296           option bits are those specified in the call to pcre_compile(), modified
1297           by any top-level option settings within the pattern itself.
1298    
1299           A  pattern  is  automatically  anchored by PCRE if all of its top-level
1300           alternatives begin with one of the following:
1301    
1302             ^     unless PCRE_MULTILINE is set
1303             \A    always
1304             \G    always
1305             .*    if PCRE_DOTALL is set and there are no back
1306                     references to the subpattern in which .* appears
1307    
1308           For such patterns, the PCRE_ANCHORED bit is set in the options returned
1309           by pcre_fullinfo().
1310    
1311             PCRE_INFO_SIZE
1312    
1313           Return  the  size  of the compiled pattern, that is, the value that was
1314           passed as the argument to pcre_malloc() when PCRE was getting memory in
1315           which to place the compiled data. The fourth argument should point to a
1316           size_t variable.
1317    
1318             PCRE_INFO_STUDYSIZE
1319    
1320           Return the size of the data block pointed to by the study_data field in
1321           a  pcre_extra  block.  That  is,  it  is  the  value that was passed to
1322           pcre_malloc() when PCRE was getting memory into which to place the data
1323           created  by  pcre_study(). The fourth argument should point to a size_t
1324           variable.
1325    
1326    
1327    OBSOLETE INFO FUNCTION
1328    
1329           int pcre_info(const pcre *code, int *optptr, int *firstcharptr);
1330    
1331           The pcre_info() function is now obsolete because its interface  is  too
1332           restrictive  to return all the available data about a compiled pattern.
1333           New  programs  should  use  pcre_fullinfo()  instead.  The   yield   of
1334           pcre_info()  is the number of capturing subpatterns, or one of the fol-
1335           lowing negative numbers:
1336    
1337             PCRE_ERROR_NULL       the argument code was NULL
1338             PCRE_ERROR_BADMAGIC   the "magic number" was not found
1339    
1340           If the optptr argument is not NULL, a copy of the  options  with  which
1341           the  pattern  was  compiled  is placed in the integer it points to (see
1342           PCRE_INFO_OPTIONS above).
1343    
1344           If the pattern is not anchored and the  firstcharptr  argument  is  not
1345           NULL,  it is used to pass back information about the first character of
1346           any matched string (see PCRE_INFO_FIRSTBYTE above).
1347    
1348    
1349    REFERENCE COUNTS
1350    
1351           int pcre_refcount(pcre *code, int adjust);
1352    
1353           The pcre_refcount() function is used to maintain a reference  count  in
1354           the data block that contains a compiled pattern. It is provided for the
1355           benefit of applications that  operate  in  an  object-oriented  manner,
1356           where different parts of the application may be using the same compiled
1357           pattern, but you want to free the block when they are all done.
1358    
1359           When a pattern is compiled, the reference count field is initialized to
1360           zero.   It is changed only by calling this function, whose action is to
1361           add the adjust value (which may be positive or  negative)  to  it.  The
1362           yield of the function is the new value. However, the value of the count
1363           is constrained to lie between 0 and 65535, inclusive. If the new  value
1364           is outside these limits, it is forced to the appropriate limit value.
1365    
1366           Except  when it is zero, the reference count is not correctly preserved
1367           if a pattern is compiled on one host and then  transferred  to  a  host
1368           whose byte-order is different. (This seems a highly unlikely scenario.)
1369    
1370    
1371    MATCHING A PATTERN: THE TRADITIONAL FUNCTION
1372    
1373           int pcre_exec(const pcre *code, const pcre_extra *extra,
1374                const char *subject, int length, int startoffset,
1375                int options, int *ovector, int ovecsize);
1376    
1377           The function pcre_exec() is called to match a subject string against  a
1378           compiled  pattern, which is passed in the code argument. If the pattern
1379           has been studied, the result of the study should be passed in the extra
1380           argument.  This  function is the main matching facility of the library,
1381           and it operates in a Perl-like manner. For specialist use there is also
1382           an  alternative matching function, which is described below in the sec-
1383           tion about the pcre_dfa_exec() function.
1384    
1385           In most applications, the pattern will have been compiled (and  option-
1386           ally  studied)  in the same process that calls pcre_exec(). However, it
1387           is possible to save compiled patterns and study data, and then use them
1388           later  in  different processes, possibly even on different hosts. For a
1389           discussion about this, see the pcreprecompile documentation.
1390    
1391           Here is an example of a simple call to pcre_exec():
1392    
1393             int rc;
1394             int ovector[30];
1395             rc = pcre_exec(
1396               re,             /* result of pcre_compile() */
1397               NULL,           /* we didn't study the pattern */
1398               "some string",  /* the subject string */
1399               11,             /* the length of the subject string */
1400               0,              /* start at offset 0 in the subject */
1401               0,              /* default options */
1402               ovector,        /* vector of integers for substring information */
1403               30);            /* number of elements (NOT size in bytes) */
1404    
1405       Extra data for pcre_exec()
1406    
1407           If the extra argument is not NULL, it must point to a  pcre_extra  data
1408           block.  The pcre_study() function returns such a block (when it doesn't
1409           return NULL), but you can also create one for yourself, and pass  addi-
1410           tional  information in it. The fields in a pcre_extra block are as fol-
1411           lows:
1412    
1413             unsigned long int flags;
1414             void *study_data;
1415             unsigned long int match_limit;
1416             void *callout_data;
1417             const unsigned char *tables;
1418    
1419           The flags field is a bitmap that specifies which of  the  other  fields
1420           are set. The flag bits are:
1421    
1422             PCRE_EXTRA_STUDY_DATA
1423             PCRE_EXTRA_MATCH_LIMIT
1424             PCRE_EXTRA_CALLOUT_DATA
1425             PCRE_EXTRA_TABLES
1426    
1427           Other  flag  bits should be set to zero. The study_data field is set in
1428           the pcre_extra block that is returned by  pcre_study(),  together  with
1429           the appropriate flag bit. You should not set this yourself, but you may
1430           add to the block by setting the other fields  and  their  corresponding
1431           flag bits.
1432    
1433           The match_limit field provides a means of preventing PCRE from using up
1434           a vast amount of resources when running patterns that are not going  to
1435           match,  but  which  have  a very large number of possibilities in their
1436           search trees. The classic  example  is  the  use  of  nested  unlimited
1437           repeats.
1438    
1439           Internally,  PCRE uses a function called match() which it calls repeat-
1440           edly (sometimes recursively). The limit is imposed  on  the  number  of
1441           times  this  function is called during a match, which has the effect of
1442           limiting the amount of recursion and backtracking that can take  place.
1443           For patterns that are not anchored, the count starts from zero for each
1444           position in the subject string.
1445    
1446           The default limit for the library can be set when PCRE  is  built;  the
1447           default  default  is 10 million, which handles all but the most extreme
1448           cases. You can reduce  the  default  by  suppling  pcre_exec()  with  a
1449           pcre_extra  block  in  which match_limit is set to a smaller value, and
1450           PCRE_EXTRA_MATCH_LIMIT is set in the  flags  field.  If  the  limit  is
1451           exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1452    
1453           The  pcre_callout  field is used in conjunction with the "callout" fea-
1454           ture, which is described in the pcrecallout documentation.
1455    
1456           The tables field  is  used  to  pass  a  character  tables  pointer  to
1457           pcre_exec();  this overrides the value that is stored with the compiled
1458           pattern. A non-NULL value is stored with the compiled pattern  only  if
1459           custom  tables  were  supplied to pcre_compile() via its tableptr argu-
1460           ment.  If NULL is passed to pcre_exec() using this mechanism, it forces
1461           PCRE's  internal  tables  to be used. This facility is helpful when re-
1462           using patterns that have been saved after compiling  with  an  external
1463           set  of  tables,  because  the  external tables might be at a different
1464           address when pcre_exec() is called. See the  pcreprecompile  documenta-
1465           tion for a discussion of saving compiled patterns for later use.
1466    
1467       Option bits for pcre_exec()
1468    
1469           The  unused  bits of the options argument for pcre_exec() must be zero.
1470           The  only  bits  that  may  be  set  are  PCRE_ANCHORED,   PCRE_NOTBOL,
1471           PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NO_UTF8_CHECK and PCRE_PARTIAL.
1472    
1473             PCRE_ANCHORED
1474    
1475           The  PCRE_ANCHORED  option  limits pcre_exec() to matching at the first
1476           matching position. If a pattern was  compiled  with  PCRE_ANCHORED,  or
1477           turned  out to be anchored by virtue of its contents, it cannot be made
1478           unachored at matching time.
1479    
1480             PCRE_NOTBOL
1481    
1482           This option specifies that first character of the subject string is not
1483           the  beginning  of  a  line, so the circumflex metacharacter should not
1484           match before it. Setting this without PCRE_MULTILINE (at compile  time)
1485           causes  circumflex  never to match. This option affects only the behav-
1486           iour of the circumflex metacharacter. It does not affect \A.
1487    
1488             PCRE_NOTEOL
1489    
1490           This option specifies that the end of the subject string is not the end
1491           of  a line, so the dollar metacharacter should not match it nor (except
1492           in multiline mode) a newline immediately before it. Setting this  with-
1493           out PCRE_MULTILINE (at compile time) causes dollar never to match. This
1494           option affects only the behaviour of the dollar metacharacter. It  does
1495           not affect \Z or \z.
1496    
1497             PCRE_NOTEMPTY
1498    
1499           An empty string is not considered to be a valid match if this option is
1500           set. If there are alternatives in the pattern, they are tried.  If  all
1501           the  alternatives  match  the empty string, the entire match fails. For
1502           example, if the pattern
1503    
1504             a?b?
1505    
1506           is applied to a string not beginning with "a" or "b",  it  matches  the
1507           empty  string at the start of the subject. With PCRE_NOTEMPTY set, this
1508           match is not valid, so PCRE searches further into the string for occur-
1509           rences of "a" or "b".
1510    
1511           Perl has no direct equivalent of PCRE_NOTEMPTY, but it does make a spe-
1512           cial case of a pattern match of the empty  string  within  its  split()
1513           function,  and  when  using  the /g modifier. It is possible to emulate
1514           Perl's behaviour after matching a null string by first trying the match
1515           again at the same offset with PCRE_NOTEMPTY and PCRE_ANCHORED, and then
1516           if that fails by advancing the starting offset (see below)  and  trying
1517           an ordinary match again. There is some code that demonstrates how to do
1518           this in the pcredemo.c sample program.
1519    
1520             PCRE_NO_UTF8_CHECK
1521    
1522           When PCRE_UTF8 is set at compile time, the validity of the subject as a
1523           UTF-8  string is automatically checked when pcre_exec() is subsequently
1524           called.  The value of startoffset is also checked  to  ensure  that  it
1525           points  to the start of a UTF-8 character. If an invalid UTF-8 sequence
1526           of bytes is found, pcre_exec() returns the error PCRE_ERROR_BADUTF8. If
1527           startoffset  contains  an  invalid  value, PCRE_ERROR_BADUTF8_OFFSET is
1528           returned.
1529    
1530           If you already know that your subject is valid, and you  want  to  skip
1531           these    checks    for   performance   reasons,   you   can   set   the
1532           PCRE_NO_UTF8_CHECK option when calling pcre_exec(). You might  want  to
1533           do  this  for the second and subsequent calls to pcre_exec() if you are
1534           making repeated calls to find all  the  matches  in  a  single  subject
1535           string.  However,  you  should  be  sure  that the value of startoffset
1536           points to the start of a UTF-8 character.  When  PCRE_NO_UTF8_CHECK  is
1537           set,  the  effect of passing an invalid UTF-8 string as a subject, or a
1538           value of startoffset that does not point to the start of a UTF-8  char-
1539           acter, is undefined. Your program may crash.
1540    
1541             PCRE_PARTIAL
1542    
1543           This  option  turns  on  the  partial  matching feature. If the subject
1544           string fails to match the pattern, but at some point during the  match-
1545           ing  process  the  end of the subject was reached (that is, the subject
1546           partially matches the pattern and the failure to  match  occurred  only
1547           because  there were not enough subject characters), pcre_exec() returns
1548           PCRE_ERROR_PARTIAL instead of PCRE_ERROR_NOMATCH. When PCRE_PARTIAL  is
1549           used,  there  are restrictions on what may appear in the pattern. These
1550           are discussed in the pcrepartial documentation.
1551    
1552       The string to be matched by pcre_exec()
1553    
1554           The subject string is passed to pcre_exec() as a pointer in subject,  a
1555           length  in  length, and a starting byte offset in startoffset. In UTF-8
1556           mode, the byte offset must point to the start  of  a  UTF-8  character.
1557           Unlike  the  pattern string, the subject may contain binary zero bytes.
1558           When the starting offset is zero, the search for a match starts at  the
1559           beginning of the subject, and this is by far the most common case.
1560    
1561           A  non-zero  starting offset is useful when searching for another match
1562           in the same subject by calling pcre_exec() again after a previous  suc-
1563           cess.   Setting  startoffset differs from just passing over a shortened
1564           string and setting PCRE_NOTBOL in the case of  a  pattern  that  begins
1565           with any kind of lookbehind. For example, consider the pattern
1566    
1567             \Biss\B
1568    
1569           which  finds  occurrences  of "iss" in the middle of words. (\B matches
1570           only if the current position in the subject is not  a  word  boundary.)
1571           When  applied  to the string "Mississipi" the first call to pcre_exec()
1572           finds the first occurrence. If pcre_exec() is called  again  with  just
1573           the  remainder  of  the  subject,  namely  "issipi", it does not match,
1574           because \B is always false at the start of the subject, which is deemed
1575           to  be  a  word  boundary. However, if pcre_exec() is passed the entire
1576           string again, but with startoffset set to 4, it finds the second occur-
1577           rence  of "iss" because it is able to look behind the starting point to
1578           discover that it is preceded by a letter.
1579    
1580           If a non-zero starting offset is passed when the pattern  is  anchored,
1581           one attempt to match at the given offset is made. This can only succeed
1582           if the pattern does not require the match to be at  the  start  of  the
1583           subject.
1584    
1585       How pcre_exec() returns captured substrings
1586    
1587           In  general, a pattern matches a certain portion of the subject, and in
1588           addition, further substrings from the subject  may  be  picked  out  by
1589           parts  of  the  pattern.  Following the usage in Jeffrey Friedl's book,
1590           this is called "capturing" in what follows, and the  phrase  "capturing
1591           subpattern"  is  used for a fragment of a pattern that picks out a sub-
1592           string. PCRE supports several other kinds of  parenthesized  subpattern
1593           that do not cause substrings to be captured.
1594    
1595           Captured  substrings are returned to the caller via a vector of integer
1596           offsets whose address is passed in ovector. The number of  elements  in
1597           the  vector is passed in ovecsize, which must be a non-negative number.
1598           Note: this argument is NOT the size of ovector in bytes.
1599    
1600           The first two-thirds of the vector is used to pass back  captured  sub-
1601           strings,  each  substring using a pair of integers. The remaining third
1602           of the vector is used as workspace by pcre_exec() while  matching  cap-
1603           turing  subpatterns, and is not available for passing back information.
1604           The length passed in ovecsize should always be a multiple of three.  If
1605           it is not, it is rounded down.
1606    
1607           When  a  match  is successful, information about captured substrings is
1608           returned in pairs of integers, starting at the  beginning  of  ovector,
1609           and  continuing  up  to two-thirds of its length at the most. The first
1610           element of a pair is set to the offset of the first character in a sub-
1611           string,  and  the  second  is  set to the offset of the first character
1612           after the end of a substring. The  first  pair,  ovector[0]  and  ovec-
1613           tor[1],  identify  the  portion  of  the  subject string matched by the
1614           entire pattern. The next pair is used for the first  capturing  subpat-
1615           tern,  and  so  on.  The value returned by pcre_exec() is the number of
1616           pairs that have been set. If there are no  capturing  subpatterns,  the
1617           return  value  from  a  successful match is 1, indicating that just the
1618           first pair of offsets has been set.
1619    
1620           Some convenience functions are provided  for  extracting  the  captured
1621           substrings  as  separate  strings. These are described in the following
1622           section.
1623    
1624           It is possible for an capturing subpattern number  n+1  to  match  some
1625           part  of  the  subject  when subpattern n has not been used at all. For
1626           example, if the string "abc" is matched against the pattern (a|(z))(bc)
1627           subpatterns  1 and 3 are matched, but 2 is not. When this happens, both
1628           offset values corresponding to the unused subpattern are set to -1.
1629    
1630           If a capturing subpattern is matched repeatedly, it is the last portion
1631           of the string that it matched that is returned.
1632    
1633           If  the vector is too small to hold all the captured substring offsets,
1634           it is used as far as possible (up to two-thirds of its length), and the
1635           function  returns a value of zero. In particular, if the substring off-
1636           sets are not of interest, pcre_exec() may be called with ovector passed
1637           as  NULL  and  ovecsize  as zero. However, if the pattern contains back
1638           references and the ovector is not big enough to  remember  the  related
1639           substrings,  PCRE has to get additional memory for use during matching.
1640           Thus it is usually advisable to supply an ovector.
1641    
1642           Note that pcre_info() can be used to find out how many  capturing  sub-
1643           patterns there are in a compiled pattern. The smallest size for ovector
1644           that will allow for n captured substrings, in addition to  the  offsets
1645           of the substring matched by the whole pattern, is (n+1)*3.
1646    
1647       Return values from pcre_exec()
1648    
1649           If  pcre_exec()  fails, it returns a negative number. The following are
1650           defined in the header file:
1651    
1652             PCRE_ERROR_NOMATCH        (-1)
1653    
1654           The subject string did not match the pattern.
1655    
1656             PCRE_ERROR_NULL           (-2)
1657    
1658           Either code or subject was passed as NULL,  or  ovector  was  NULL  and
1659           ovecsize was not zero.
1660    
1661             PCRE_ERROR_BADOPTION      (-3)
1662    
1663           An unrecognized bit was set in the options argument.
1664    
1665             PCRE_ERROR_BADMAGIC       (-4)
1666    
1667           PCRE  stores a 4-byte "magic number" at the start of the compiled code,
1668           to catch the case when it is passed a junk pointer and to detect when a
1669           pattern that was compiled in an environment of one endianness is run in
1670           an environment with the other endianness. This is the error  that  PCRE
1671           gives when the magic number is not present.
1672    
1673             PCRE_ERROR_UNKNOWN_NODE   (-5)
1674    
1675           While running the pattern match, an unknown item was encountered in the
1676           compiled pattern. This error could be caused by a bug  in  PCRE  or  by
1677           overwriting of the compiled pattern.
1678    
1679             PCRE_ERROR_NOMEMORY       (-6)
1680    
1681           If  a  pattern contains back references, but the ovector that is passed
1682           to pcre_exec() is not big enough to remember the referenced substrings,
1683           PCRE  gets  a  block of memory at the start of matching to use for this
1684           purpose. If the call via pcre_malloc() fails, this error is given.  The
1685           memory is automatically freed at the end of matching.
1686    
1687             PCRE_ERROR_NOSUBSTRING    (-7)
1688    
1689           This  error is used by the pcre_copy_substring(), pcre_get_substring(),
1690           and  pcre_get_substring_list()  functions  (see  below).  It  is  never
1691           returned by pcre_exec().
1692    
1693             PCRE_ERROR_MATCHLIMIT     (-8)
1694    
1695           The  recursion  and backtracking limit, as specified by the match_limit
1696           field in a pcre_extra structure (or defaulted)  was  reached.  See  the
1697           description above.
1698    
1699             PCRE_ERROR_CALLOUT        (-9)
1700    
1701           This error is never generated by pcre_exec() itself. It is provided for
1702           use by callout functions that want to yield a distinctive  error  code.
1703           See the pcrecallout documentation for details.
1704    
1705             PCRE_ERROR_BADUTF8        (-10)
1706    
1707           A  string  that contains an invalid UTF-8 byte sequence was passed as a
1708           subject.
1709    
1710             PCRE_ERROR_BADUTF8_OFFSET (-11)
1711    
1712           The UTF-8 byte sequence that was passed as a subject was valid, but the
1713           value  of startoffset did not point to the beginning of a UTF-8 charac-
1714           ter.
1715    
1716             PCRE_ERROR_PARTIAL        (-12)
1717    
1718           The subject string did not match, but it did match partially.  See  the
1719           pcrepartial documentation for details of partial matching.
1720    
1721             PCRE_ERROR_BADPARTIAL     (-13)
1722    
1723           The  PCRE_PARTIAL  option  was  used with a compiled pattern containing
1724           items that are not supported for partial matching. See the  pcrepartial
1725           documentation for details of partial matching.
1726    
1727             PCRE_ERROR_INTERNAL       (-14)
1728    
1729           An  unexpected  internal error has occurred. This error could be caused
1730           by a bug in PCRE or by overwriting of the compiled pattern.
1731    
1732             PCRE_ERROR_BADCOUNT       (-15)
1733    
1734           This error is given if the value of the ovecsize argument is  negative.
1735    
1736    
1737    EXTRACTING CAPTURED SUBSTRINGS BY NUMBER
1738    
1739           int pcre_copy_substring(const char *subject, int *ovector,
1740                int stringcount, int stringnumber, char *buffer,
1741                int buffersize);
1742    
1743           int pcre_get_substring(const char *subject, int *ovector,
1744                int stringcount, int stringnumber,
1745                const char **stringptr);
1746    
1747           int pcre_get_substring_list(const char *subject,
1748                int *ovector, int stringcount, const char ***listptr);
1749    
1750           Captured  substrings  can  be  accessed  directly  by using the offsets
1751           returned by pcre_exec() in  ovector.  For  convenience,  the  functions
1752           pcre_copy_substring(),    pcre_get_substring(),    and    pcre_get_sub-
1753           string_list() are provided for extracting captured substrings  as  new,
1754           separate,  zero-terminated strings. These functions identify substrings
1755           by number. The next section describes functions  for  extracting  named
1756           substrings.  A  substring  that  contains  a  binary  zero is correctly
1757           extracted and has a further zero added on the end, but  the  result  is
1758           not, of course, a C string.
1759    
1760           The  first  three  arguments  are the same for all three of these func-
1761           tions: subject is the subject string that has  just  been  successfully
1762           matched, ovector is a pointer to the vector of integer offsets that was
1763           passed to pcre_exec(), and stringcount is the number of substrings that
1764           were  captured  by  the match, including the substring that matched the
1765           entire regular expression. This is the value returned by pcre_exec() if
1766           it  is greater than zero. If pcre_exec() returned zero, indicating that
1767           it ran out of space in ovector, the value passed as stringcount  should
1768           be the number of elements in the vector divided by three.
1769    
1770           The  functions pcre_copy_substring() and pcre_get_substring() extract a
1771           single substring, whose number is given as  stringnumber.  A  value  of
1772           zero  extracts  the  substring that matched the entire pattern, whereas
1773           higher values  extract  the  captured  substrings.  For  pcre_copy_sub-
1774           string(),  the  string  is  placed  in buffer, whose length is given by
1775           buffersize, while for pcre_get_substring() a new  block  of  memory  is
1776           obtained  via  pcre_malloc,  and its address is returned via stringptr.
1777           The yield of the function is the length of the  string,  not  including
1778           the terminating zero, or one of
1779    
1780             PCRE_ERROR_NOMEMORY       (-6)
1781    
1782           The  buffer  was too small for pcre_copy_substring(), or the attempt to
1783           get memory failed for pcre_get_substring().
1784    
1785             PCRE_ERROR_NOSUBSTRING    (-7)
1786    
1787           There is no substring whose number is stringnumber.
1788    
1789           The pcre_get_substring_list()  function  extracts  all  available  sub-
1790           strings  and  builds  a list of pointers to them. All this is done in a
1791           single block of memory that is obtained via pcre_malloc. The address of
1792           the  memory  block  is returned via listptr, which is also the start of
1793           the list of string pointers. The end of the list is marked  by  a  NULL
1794           pointer. The yield of the function is zero if all went well, or
1795    
1796             PCRE_ERROR_NOMEMORY       (-6)
1797    
1798           if the attempt to get the memory block failed.
1799    
1800           When  any of these functions encounter a substring that is unset, which
1801           can happen when capturing subpattern number n+1 matches  some  part  of
1802           the  subject, but subpattern n has not been used at all, they return an
1803           empty string. This can be distinguished from a genuine zero-length sub-
1804           string  by inspecting the appropriate offset in ovector, which is nega-
1805           tive for unset substrings.
1806    
1807           The two convenience functions pcre_free_substring() and  pcre_free_sub-
1808           string_list()  can  be  used  to free the memory returned by a previous
1809           call  of  pcre_get_substring()  or  pcre_get_substring_list(),  respec-
1810           tively.  They  do  nothing  more  than  call the function pointed to by
1811           pcre_free, which of course could be called directly from a  C  program.
1812           However,  PCRE is used in some situations where it is linked via a spe-
1813           cial  interface  to  another  programming  language  which  cannot  use
1814           pcre_free  directly;  it is for these cases that the functions are pro-
1815           vided.
1816    
1817    
1818    EXTRACTING CAPTURED SUBSTRINGS BY NAME
1819    
1820           int pcre_get_stringnumber(const pcre *code,
1821                const char *name);
1822    
1823           int pcre_copy_named_substring(const pcre *code,
1824                const char *subject, int *ovector,
1825                int stringcount, const char *stringname,
1826                char *buffer, int buffersize);
1827    
1828           int pcre_get_named_substring(const pcre *code,
1829                const char *subject, int *ovector,
1830                int stringcount, const char *stringname,
1831                const char **stringptr);
1832    
1833           To extract a substring by name, you first have to find associated  num-
1834           ber.  For example, for this pattern
1835    
1836             (a+)b(?<xxx>\d+)...
1837    
1838           the number of the subpattern called "xxx" is 2. You can find the number
1839           from the name by calling pcre_get_stringnumber(). The first argument is
1840           the  compiled  pattern,  and  the  second is the name. The yield of the
1841           function is the subpattern number, or  PCRE_ERROR_NOSUBSTRING  (-7)  if
1842           there is no subpattern of that name.
1843    
1844           Given the number, you can extract the substring directly, or use one of
1845           the functions described in the previous section. For convenience, there
1846           are also two functions that do the whole job.
1847    
1848           Most    of    the    arguments   of   pcre_copy_named_substring()   and
1849           pcre_get_named_substring() are the same  as  those  for  the  similarly
1850           named  functions  that extract by number. As these are described in the
1851           previous section, they are not re-described here. There  are  just  two
1852           differences:
1853    
1854           First,  instead  of a substring number, a substring name is given. Sec-
1855           ond, there is an extra argument, given at the start, which is a pointer
1856           to  the compiled pattern. This is needed in order to gain access to the
1857           name-to-number translation table.
1858    
1859           These functions call pcre_get_stringnumber(), and if it succeeds,  they
1860           then  call  pcre_copy_substring() or pcre_get_substring(), as appropri-
1861           ate.
1862    
1863    
1864    FINDING ALL POSSIBLE MATCHES
1865    
1866           The traditional matching function uses a  similar  algorithm  to  Perl,
1867           which stops when it finds the first match, starting at a given point in
1868           the subject. If you want to find all possible matches, or  the  longest
1869           possible  match,  consider using the alternative matching function (see
1870           below) instead. If you cannot use the alternative function,  but  still
1871           need  to  find all possible matches, you can kludge it up by making use
1872           of the callout facility, which is described in the pcrecallout documen-
1873           tation.
1874    
1875           What you have to do is to insert a callout right at the end of the pat-
1876           tern.  When your callout function is called, extract and save the  cur-
1877           rent  matched  substring.  Then  return  1, which forces pcre_exec() to
1878           backtrack and try other alternatives. Ultimately, when it runs  out  of
1879           matches, pcre_exec() will yield PCRE_ERROR_NOMATCH.
1880    
1881    
1882    MATCHING A PATTERN: THE ALTERNATIVE FUNCTION
1883    
1884           int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
1885                const char *subject, int length, int startoffset,
1886                int options, int *ovector, int ovecsize,
1887                int *workspace, int wscount);
1888    
1889           The  function  pcre_dfa_exec()  is  called  to  match  a subject string
1890           against a compiled pattern, using a "DFA" matching algorithm. This  has
1891           different  characteristics to the normal algorithm, and is not compati-
1892           ble with Perl. Some of the features of PCRE patterns are not supported.
1893           Nevertheless, there are times when this kind of matching can be useful.
1894           For a discussion of the two matching algorithms, see  the  pcrematching
1895           documentation.
1896    
1897           The  arguments  for  the  pcre_dfa_exec()  function are the same as for
1898           pcre_exec(), plus two extras. The ovector argument is used in a differ-
1899           ent  way,  and  this is described below. The other common arguments are
1900           used in the same way as for pcre_exec(), so their  description  is  not
1901           repeated here.
1902    
1903           The  two  additional  arguments provide workspace for the function. The
1904           workspace vector should contain at least 20 elements. It  is  used  for
1905           keeping  track  of  multiple  paths  through  the  pattern  tree.  More
1906           workspace will be needed for patterns and subjects where  there  are  a
1907           lot of possible matches.
1908    
1909           Here is an example of a simple call to pcre_exec():
1910    
1911             int rc;
1912             int ovector[10];
1913             int wspace[20];
1914             rc = pcre_exec(
1915               re,             /* result of pcre_compile() */
1916               NULL,           /* we didn't study the pattern */
1917               "some string",  /* the subject string */
1918               11,             /* the length of the subject string */
1919               0,              /* start at offset 0 in the subject */
1920               0,              /* default options */
1921               ovector,        /* vector of integers for substring information */
1922               10,             /* number of elements (NOT size in bytes) */
1923               wspace,         /* working space vector */
1924               20);            /* number of elements (NOT size in bytes) */
1925    
1926       Option bits for pcre_dfa_exec()
1927    
1928           The  unused  bits  of  the options argument for pcre_dfa_exec() must be
1929           zero. The only bits that may be  set  are  PCRE_ANCHORED,  PCRE_NOTBOL,
1930           PCRE_NOTEOL,     PCRE_NOTEMPTY,    PCRE_NO_UTF8_CHECK,    PCRE_PARTIAL,
1931           PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART. All  but  the  last  three  of
1932           these  are  the  same  as  for pcre_exec(), so their description is not
1933           repeated here.
1934    
1935             PCRE_PARTIAL
1936    
1937           This has the same general effect as it does for  pcre_exec(),  but  the
1938           details   are   slightly   different.  When  PCRE_PARTIAL  is  set  for
1939           pcre_dfa_exec(), the return code PCRE_ERROR_NOMATCH is  converted  into
1940           PCRE_ERROR_PARTIAL  if  the  end  of the subject is reached, there have
1941           been no complete matches, but there is still at least one matching pos-
1942           sibility.  The portion of the string that provided the partial match is
1943           set as the first matching string.
1944    
1945             PCRE_DFA_SHORTEST
1946    
1947           Setting the PCRE_DFA_SHORTEST option causes the matching  algorithm  to
1948           stop  as  soon  as  it  has found one match. Because of the way the DFA
1949           algorithm works, this is necessarily the shortest possible match at the
1950           first possible matching point in the subject string.
1951    
1952             PCRE_DFA_RESTART
1953    
1954           When  pcre_dfa_exec()  is  called  with  the  PCRE_PARTIAL  option, and
1955           returns a partial match, it is possible to call it  again,  with  addi-
1956           tional  subject  characters,  and have it continue with the same match.
1957           The PCRE_DFA_RESTART option requests this action; when it is  set,  the
1958           workspace  and wscount options must reference the same vector as before
1959           because data about the match so far is left in  them  after  a  partial
1960           match.  There  is  more  discussion of this facility in the pcrepartial
1961           documentation.
1962    
1963       Successful returns from pcre_dfa_exec()
1964    
1965           When pcre_dfa_exec() succeeds, it may have matched more than  one  sub-
1966           string in the subject. Note, however, that all the matches from one run
1967           of the function start at the same point in  the  subject.  The  shorter
1968           matches  are all initial substrings of the longer matches. For example,
1969           if the pattern
1970    
1971             <.*>
1972    
1973           is matched against the string
1974    
1975             This is <something> <something else> <something further> no more
1976    
1977           the three matched strings are
1978    
1979             <something>
1980             <something> <something else>
1981             <something> <something else> <something further>
1982    
1983           On success, the yield of the function is a number  greater  than  zero,
1984           which  is  the  number of matched substrings. The substrings themselves
1985           are returned in ovector. Each string uses two elements;  the  first  is
1986           the  offset  to the start, and the second is the offset to the end. All
1987           the strings have the same start offset. (Space could have been saved by
1988           giving  this only once, but it was decided to retain some compatibility
1989           with the way pcre_exec() returns data, even though the meaning  of  the
1990           strings is different.)
1991    
1992           The strings are returned in reverse order of length; that is, the long-
1993           est matching string is given first. If there were too many  matches  to
1994           fit  into ovector, the yield of the function is zero, and the vector is
1995           filled with the longest matches.
1996    
1997       Error returns from pcre_dfa_exec()
1998    
1999           The pcre_dfa_exec() function returns a negative number when  it  fails.
2000           Many  of  the  errors  are  the  same as for pcre_exec(), and these are
2001           described above.  There are in addition the following errors  that  are
2002           specific to pcre_dfa_exec():
2003    
2004             PCRE_ERROR_DFA_UITEM      (-16)
2005    
2006           This  return is given if pcre_dfa_exec() encounters an item in the pat-
2007           tern that it does not support, for instance, the use of \C  or  a  back
2008           reference.
2009    
2010             PCRE_ERROR_DFA_UCOND      (-17)
2011    
2012           This  return is given if pcre_dfa_exec() encounters a condition item in
2013           a pattern that uses a back reference for the  condition.  This  is  not
2014           supported.
2015    
2016             PCRE_ERROR_DFA_UMLIMIT    (-18)
2017    
2018           This  return  is given if pcre_dfa_exec() is called with an extra block
2019           that contains a setting of the match_limit field. This is not supported
2020           (it is meaningless).
2021    
2022             PCRE_ERROR_DFA_WSSIZE     (-19)
2023    
2024           This  return  is  given  if  pcre_dfa_exec()  runs  out of space in the
2025           workspace vector.
2026    
2027             PCRE_ERROR_DFA_RECURSE    (-20)
2028    
2029           When a recursive subpattern is processed, the matching  function  calls
2030           itself  recursively,  using  private vectors for ovector and workspace.
2031           This error is given if the output vector  is  not  large  enough.  This
2032           should be extremely rare, as a vector of size 1000 is used.
2033    
2034    Last updated: 16 May 2005
2035    Copyright (c) 1997-2005 University of Cambridge.
2036    -----------------------------------------------------------------------------
2037    
 MATCHING A PATTERN  
      The function pcre_exec() is called to match a subject string  
      against  a pre-compiled pattern, which is passed in the code  
      argument. If the pattern has been studied, the result of the  
      study should be passed in the extra argument. Otherwise this  
      must be NULL.  
   
      The PCRE_ANCHORED option can be passed in the options  argu-  
      ment,  whose unused bits must be zero. However, if a pattern  
      was  compiled  with  PCRE_ANCHORED,  or  turned  out  to  be  
      anchored  by  virtue  of  its  contents,  it  cannot be made  
      unachored at matching time.  
   
      There are also three further options that can be set only at  
      matching time:  
   
        PCRE_NOTBOL  
   
      The first character of the string is not the beginning of  a  
      line,  so  the  circumflex  metacharacter  should  not match  
      before it. Setting this without PCRE_MULTILINE  (at  compile  
      time) causes circumflex never to match.  
   
        PCRE_NOTEOL  
   
      The end of the string is not the end of a line, so the  dol-  
      lar  metacharacter should not match it nor (except in multi-  
      line mode) a newline immediately  before  it.  Setting  this  
      without PCRE_MULTILINE (at compile time) causes dollar never  
      to match.  
   
        PCRE_NOTEMPTY  
   
      An empty string is not considered to be  a  valid  match  if  
      this  option  is  set. If there are alternatives in the pat-  
      tern, they are tried. If  all  the  alternatives  match  the  
      empty  string,  the  entire match fails. For example, if the  
      pattern  
   
        a?b?  
   
      is applied to a string not beginning with  "a"  or  "b",  it  
      matches  the  empty string at the start of the subject. With  
      PCRE_NOTEMPTY set, this match is not valid, so PCRE searches  
      further into the string for occurrences of "a" or "b".  
   
      Perl has no direct equivalent of PCRE_NOTEMPTY, but it  does  
      make  a  special case of a pattern match of the empty string  
      within its split() function, and when using the /g modifier.  
      It  is possible to emulate Perl's behaviour after matching a  
      null string by first trying the  match  again  at  the  same  
      offset  with  PCRE_NOTEMPTY  set,  and then if that fails by  
      advancing the starting offset  (see  below)  and  trying  an  
      ordinary match again.  
   
      The subject string is passed as  a  pointer  in  subject,  a  
      length  in  length,  and  a  starting offset in startoffset.  
      Unlike the pattern string, it may contain binary zero  char-  
      acters.  When  the starting offset is zero, the search for a  
      match starts at the beginning of the subject, and this is by  
      far the most common case.  
   
      A non-zero starting offset  is  useful  when  searching  for  
      another  match  in  the  same subject by calling pcre_exec()  
      again after a previous success.  Setting startoffset differs  
      from  just  passing  over  a  shortened  string  and setting  
      PCRE_NOTBOL in the case of a pattern that  begins  with  any  
      kind of lookbehind. For example, consider the pattern  
   
        \Biss\B  
   
      which finds occurrences of "iss" in the middle of words. (\B  
      matches only if the current position in the subject is not a  
      word boundary.) When applied to the string "Mississipi"  the  
      first  call  to  pcre_exec()  finds the first occurrence. If  
      pcre_exec() is called again with just the remainder  of  the  
      subject,  namely  "issipi", it does not match, because \B is  
      always false at the start of the subject, which is deemed to  
      be  a  word  boundary. However, if pcre_exec() is passed the  
      entire string again, but with startoffset set to 4, it finds  
      the  second  occurrence  of "iss" because it is able to look  
      behind the starting point to discover that it is preceded by  
      a letter.  
   
      If a non-zero starting offset is passed when the pattern  is  
      anchored, one attempt to match at the given offset is tried.  
      This can only succeed if the pattern does  not  require  the  
      match to be at the start of the subject.  
   
      In general, a pattern matches a certain portion of the  sub-  
      ject,  and  in addition, further substrings from the subject  
      may be picked out by parts of  the  pattern.  Following  the  
      usage  in  Jeffrey Friedl's book, this is called "capturing"  
      in what follows, and the phrase  "capturing  subpattern"  is  
      used for a fragment of a pattern that picks out a substring.  
      PCRE supports several other kinds of  parenthesized  subpat-  
      tern that do not cause substrings to be captured.  
   
      Captured substrings are returned to the caller via a  vector  
      of  integer  offsets whose address is passed in ovector. The  
      number of elements in the vector is passed in ovecsize.  The  
      first two-thirds of the vector is used to pass back captured  
      substrings, each substring using a  pair  of  integers.  The  
      remaining  third  of  the  vector  is  used  as workspace by  
      pcre_exec() while matching capturing subpatterns, and is not  
      available for passing back information. The length passed in  
      ovecsize should always be a multiple of three. If it is not,  
      it is rounded down.  
   
      When a match has been successful, information about captured  
      substrings is returned in pairs of integers, starting at the  
      beginning of ovector, and continuing up to two-thirds of its  
      length  at  the  most. The first element of a pair is set to  
      the offset of the first character in a  substring,  and  the  
      second is set to the offset of the first character after the  
      end of a substring. The first  pair,  ovector[0]  and  ovec-  
      tor[1],  identify  the portion of the subject string matched  
      by the entire pattern. The next pair is used for  the  first  
      capturing  subpattern,  and  so  on.  The  value returned by  
      pcre_exec() is the number of pairs that have  been  set.  If  
      there  are no capturing subpatterns, the return value from a  
      successful match is 1, indicating that just the  first  pair  
      of offsets has been set.  
   
      Some convenience functions are provided for  extracting  the  
      captured substrings as separate strings. These are described  
      in the following section.  
   
      It is possible for an capturing  subpattern  number  n+1  to  
      match  some  part  of  the subject when subpattern n has not  
      been used at all.  For  example,  if  the  string  "abc"  is  
      matched  against the pattern (a|(z))(bc) subpatterns 1 and 3  
      are matched, but 2 is not. When this  happens,  both  offset  
      values corresponding to the unused subpattern are set to -1.  
   
      If a capturing subpattern is matched repeatedly, it  is  the  
      last  portion  of  the  string  that  it  matched  that gets  
      returned.  
   
      If the vector is too small to hold  all  the  captured  sub-  
      strings,  it is used as far as possible (up to two-thirds of  
      its length), and the function returns a value  of  zero.  In  
      particular,  if  the  substring offsets are not of interest,  
      pcre_exec() may be called with ovector passed  as  NULL  and  
      ovecsize  as  zero.  However,  if  the pattern contains back  
      references and the ovector isn't big enough to remember  the  
      related  substrings,  PCRE  has to get additional memory for  
      use during matching. Thus it is usually advisable to  supply  
      an ovector.  
   
      Note that pcre_info() can be used to find out how many  cap-  
      turing  subpatterns  there  are  in  a compiled pattern. The  
      smallest size for ovector that will  allow  for  n  captured  
      substrings  in  addition  to  the  offsets  of the substring  
      matched by the whole pattern is (n+1)*3.  
   
      If pcre_exec() fails, it returns a negative number. The fol-  
      lowing are defined in the header file:  
   
        PCRE_ERROR_NOMATCH        (-1)  
   
      The subject string did not match the pattern.  
   
        PCRE_ERROR_NULL           (-2)  
   
      Either code or subject was passed as NULL,  or  ovector  was  
      NULL and ovecsize was not zero.  
   
        PCRE_ERROR_BADOPTION      (-3)  
   
      An unrecognized bit was set in the options argument.  
   
        PCRE_ERROR_BADMAGIC       (-4)  
   
      PCRE stores a 4-byte "magic number" at the start of the com-  
      piled  code,  to  catch  the  case  when it is passed a junk  
      pointer. This is the error it gives when  the  magic  number  
      isn't present.  
   
        PCRE_ERROR_UNKNOWN_NODE   (-5)  
   
      While running the pattern match, an unknown item was encoun-  
      tered in the compiled pattern. This error could be caused by  
      a bug in PCRE or by overwriting of the compiled pattern.  
   
        PCRE_ERROR_NOMEMORY       (-6)  
   
      If a pattern contains back references, but the ovector  that  
      is  passed  to pcre_exec() is not big enough to remember the  
      referenced substrings, PCRE gets a block of  memory  at  the  
      start  of  matching to use for this purpose. If the call via  
      pcre_malloc() fails, this error  is  given.  The  memory  is  
      freed at the end of matching.  
   
   
   
 EXTRACTING CAPTURED SUBSTRINGS  
      Captured substrings can be accessed directly  by  using  the  
      offsets returned by pcre_exec() in ovector. For convenience,  
      the functions  pcre_copy_substring(),  pcre_get_substring(),  
      and  pcre_get_substring_list()  are  provided for extracting  
      captured  substrings  as  new,   separate,   zero-terminated  
      strings.   A  substring  that  contains  a  binary  zero  is  
      correctly extracted and has a further zero added on the end,  
      but the result does not, of course, function as a C string.  
   
      The first three arguments are the same for all  three  func-  
      tions:  subject  is  the  subject string which has just been  
      successfully matched, ovector is a pointer to the vector  of  
      integer   offsets   that  was  passed  to  pcre_exec(),  and  
      stringcount is the number of substrings that  were  captured  
      by  the  match,  including  the  substring  that matched the  
      entire regular expression. This is  the  value  returned  by  
      pcre_exec  if  it  is  greater  than  zero.  If  pcre_exec()  
      returned zero, indicating that it ran out of space in  ovec-  
      tor,  the  value passed as stringcount should be the size of  
      the vector divided by three.  
   
      The functions pcre_copy_substring() and pcre_get_substring()  
      extract a single substring, whose number is given as string-  
      number. A value of zero extracts the substring that  matched  
      the entire pattern, while higher values extract the captured  
      substrings. For pcre_copy_substring(), the string is  placed  
      in  buffer,  whose  length is given by buffersize, while for  
      pcre_get_substring() a new block of store  is  obtained  via  
      pcre_malloc,  and its address is returned via stringptr. The  
      yield of the function is  the  length  of  the  string,  not  
      including the terminating zero, or one of  
   
        PCRE_ERROR_NOMEMORY       (-6)  
   
      The buffer was too small for pcre_copy_substring(),  or  the  
      attempt to get memory failed for pcre_get_substring().  
   
        PCRE_ERROR_NOSUBSTRING    (-7)  
   
      There is no substring whose number is stringnumber.  
   
      The pcre_get_substring_list() function extracts  all  avail-  
      able  substrings  and builds a list of pointers to them. All  
      this is done in a single block of memory which  is  obtained  
      via pcre_malloc. The address of the memory block is returned  
      via listptr, which is also the start of the list  of  string  
      pointers.  The  end of the list is marked by a NULL pointer.  
      The yield of the function is zero if all went well, or  
   
        PCRE_ERROR_NOMEMORY       (-6)  
   
      if the attempt to get the memory block failed.  
   
      When any of these functions encounter a  substring  that  is  
      unset, which can happen when capturing subpattern number n+1  
      matches some part of the subject, but subpattern n  has  not  
      been  used  at all, they return an empty string. This can be  
      distinguished  from  a  genuine  zero-length  substring   by  
      inspecting the appropriate offset in ovector, which is nega-  
      tive for unset substrings.  
2038    
2039    
2040    NAME
2041           PCRE - Perl-compatible regular expressions
2042    
2043    
2044  LIMITATIONS  PCRE CALLOUTS
2045       There are some size limitations in PCRE but it is hoped that  
2046       they will never in practice be relevant.  The maximum length         int (*pcre_callout)(pcre_callout_block *);
2047       of a compiled pattern is 65539 (sic) bytes.  All  values  in  
2048       repeating  quantifiers must be less than 65536.  The maximum         PCRE provides a feature called "callout", which is a means of temporar-
2049       number of capturing subpatterns is 99.  The  maximum  number         ily passing control to the caller of PCRE  in  the  middle  of  pattern
2050       of  all  parenthesized subpatterns, including capturing sub-         matching.  The  caller of PCRE provides an external function by putting
2051       patterns, assertions, and other types of subpattern, is 200.         its entry point in the global variable pcre_callout. By  default,  this
2052           variable contains NULL, which disables all calling out.
2053       The maximum length of a subject string is the largest  posi-  
2054       tive number that an integer variable can hold. However, PCRE         Within  a  regular  expression,  (?C) indicates the points at which the
2055       uses recursion to handle subpatterns and indefinite  repeti-         external function is to be called.  Different  callout  points  can  be
2056       tion.  This  means  that the available stack space may limit         identified  by  putting  a number less than 256 after the letter C. The
2057       the size of a subject string that can be processed  by  cer-         default value is zero.  For  example,  this  pattern  has  two  callout
2058       tain patterns.         points:
2059    
2060             (?C1)eabc(?C2)def
2061    
2062  DIFFERENCES FROM PERL         If  the  PCRE_AUTO_CALLOUT  option  bit  is  set when pcre_compile() is
2063       The differences described here  are  with  respect  to  Perl         called, PCRE automatically  inserts  callouts,  all  with  number  255,
2064       5.005.         before  each  item in the pattern. For example, if PCRE_AUTO_CALLOUT is
2065           used with the pattern
2066       1. By default, a whitespace character is any character  that  
2067       the  C  library  function isspace() recognizes, though it is           A(\d{2}|--)
2068       possible to compile PCRE  with  alternative  character  type  
2069       tables. Normally isspace() matches space, formfeed, newline,         it is processed as if it were
2070       carriage return, horizontal tab, and vertical tab. Perl 5 no  
2071       longer  includes vertical tab in its set of whitespace char-         (?C255)A(?C255)((?C255)\d{2}(?C255)|(?C255)-(?C255)-(?C255))(?C255)
2072       acters. The \v escape that was in the Perl documentation for  
2073       a long time was never in fact recognized. However, the char-         Notice that there is a callout before and after  each  parenthesis  and
2074       acter itself was treated as whitespace at least up to 5.002.         alternation  bar.  Automatic  callouts  can  be  used  for tracking the
2075       In 5.004 and 5.005 it does not match \s.         progress of pattern matching. The pcretest command has an  option  that
2076           sets  automatic callouts; when it is used, the output indicates how the
2077       2. PCRE does  not  allow  repeat  quantifiers  on  lookahead         pattern is matched. This is useful information when you are  trying  to
2078       assertions. Perl permits them, but they do not mean what you         optimize the performance of a particular pattern.
2079       might think. For example, (?!a){3} does not assert that  the  
2080       next  three characters are not "a". It just asserts that the  
2081       next character is not "a" three times.  MISSING CALLOUTS
2082    
2083       3. Capturing subpatterns that occur inside  negative  looka-         You  should  be  aware  that,  because of optimizations in the way PCRE
2084       head  assertions  are  counted,  but  their  entries  in the         matches patterns, callouts sometimes do not happen. For example, if the
2085       offsets vector are never set. Perl sets its numerical  vari-         pattern is
2086       ables  from  any  such  patterns that are matched before the  
2087       assertion fails to match something (thereby succeeding), but           ab(?C4)cd
2088       only  if  the negative lookahead assertion contains just one  
2089       branch.         PCRE knows that any matching string must contain the letter "d". If the
2090           subject string is "abyz", the lack of "d" means that  matching  doesn't
2091       4. Though binary zero characters are supported in  the  sub-         ever  start,  and  the  callout is never reached. However, with "abyd",
2092       ject  string,  they  are  not  allowed  in  a pattern string         though the result is still no match, the callout is obeyed.
2093       because it is passed as a normal  C  string,  terminated  by  
2094       zero. The escape sequence "\0" can be used in the pattern to  
2095       represent a binary zero.  THE CALLOUT INTERFACE
2096    
2097       5. The following Perl escape sequences  are  not  supported:         During matching, when PCRE reaches a callout point, the external  func-
2098       \l,  \u,  \L,  \U,  \E, \Q. In fact these are implemented by         tion  defined by pcre_callout is called (if it is set). This applies to
2099       Perl's general string-handling and are not part of its  pat-         both the pcre_exec() and the pcre_dfa_exec()  matching  functions.  The
2100       tern matching engine.         only  argument  to  the callout function is a pointer to a pcre_callout
2101           block. This structure contains the following fields:
2102       6. The Perl \G assertion is  not  supported  as  it  is  not  
2103       relevant to single pattern matches.           int          version;
2104             int          callout_number;
2105       7. Fairly obviously, PCRE does not support the (?{code}) and           int         *offset_vector;
2106       (?p{code})  constructions. However, there is some experimen-           const char  *subject;
2107       tal support for recursive patterns using the  non-Perl  item           int          subject_length;
2108       (?R).           int          start_match;
2109       8. There are at the time of writing some  oddities  in  Perl           int          current_position;
2110       5.005_02  concerned  with  the  settings of captured strings           int          capture_top;
2111       when part of a pattern is repeated.  For  example,  matching           int          capture_last;
2112       "aba"  against the pattern /^(a(b)?)+$/ sets $2 to the value           void        *callout_data;
2113       "b", but matching "aabbaa" against /^(aa(bb)?)+$/ leaves  $2           int          pattern_position;
2114       unset.    However,    if   the   pattern   is   changed   to           int          next_item_length;
2115       /^(aa(b(b))?)+$/ then $2 (and $3) are set.  
2116           The version field is an integer containing the version  number  of  the
2117       In Perl 5.004 $2 is set in both cases, and that is also true         block  format. The initial version was 0; the current version is 1. The
2118       of PCRE. If in the future Perl changes to a consistent state         version number will change again in future  if  additional  fields  are
2119       that is different, PCRE may change to follow.         added, but the intention is never to remove any of the existing fields.
2120    
2121       9. Another as yet unresolved discrepancy  is  that  in  Perl         The callout_number field contains the number of the  callout,  as  com-
2122       5.005_02  the  pattern /^(a)?(?(1)a|b)+$/ matches the string         piled  into  the pattern (that is, the number after ?C for manual call-
2123       "a", whereas in PCRE it does not.  However, in both Perl and         outs, and 255 for automatically generated callouts).
2124       PCRE /^(a)?a/ matched against "a" leaves $1 unset.  
2125           The offset_vector field is a pointer to the vector of offsets that  was
2126       10. PCRE  provides  some  extensions  to  the  Perl  regular         passed   by   the   caller  to  pcre_exec()  or  pcre_dfa_exec().  When
2127       expression facilities:         pcre_exec() is used, the contents can be inspected in order to  extract
2128           substrings  that  have  been  matched  so  far,  in the same way as for
2129       (a) Although lookbehind assertions must match  fixed  length         extracting substrings after a match has completed. For  pcre_dfa_exec()
2130       strings,  each  alternative branch of a lookbehind assertion         this field is not useful.
2131       can match a different length of string. Perl 5.005  requires  
2132       them all to have the same length.         The subject and subject_length fields contain copies of the values that
2133           were passed to pcre_exec().
2134       (b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is  not  
2135       set,  the  $ meta- character matches only at the very end of         The start_match field contains the offset within the subject  at  which
2136       the string.         the  current match attempt started. If the pattern is not anchored, the
2137           callout function may be called several times from the same point in the
2138       (c) If PCRE_EXTRA is set, a backslash followed by  a  letter         pattern for different starting points in the subject.
2139       with no special meaning is faulted.  
2140           The  current_position  field  contains the offset within the subject of
2141       (d) If PCRE_UNGREEDY is set, the greediness of  the  repeti-         the current match pointer.
2142       tion  quantifiers  is inverted, that is, by default they are  
2143       not greedy, but if followed by a question mark they are.         When the pcre_exec() function is used, the capture_top  field  contains
2144           one  more than the number of the highest numbered captured substring so
2145       (e) PCRE_ANCHORED can be used to force a pattern to be tried         far. If no substrings have been captured, the value of  capture_top  is
2146       only at the start of the subject.         one.  This  is always the case when pcre_dfa_exec() is used, because it
2147           does not support captured substrings.
2148       (f) The PCRE_NOTBOL, PCRE_NOTEOL, and PCRE_NOTEMPTY  options  
2149       for pcre_exec() have no Perl equivalents.         The capture_last field contains the number of the  most  recently  cap-
2150           tured  substring. If no substrings have been captured, its value is -1.
2151       (g) The (?R) construct allows for recursive pattern matching         This is always the case when pcre_dfa_exec() is used.
2152       (Perl  5.6 can do this using the (?p{code}) construct, which  
2153       PCRE cannot of course support.)         The callout_data field contains a value that is passed  to  pcre_exec()
2154           or  pcre_dfa_exec() specifically so that it can be passed back in call-
2155           outs. It is passed in the pcre_callout field  of  the  pcre_extra  data
2156           structure.  If  no such data was passed, the value of callout_data in a
2157  REGULAR EXPRESSION DETAILS         pcre_callout block is NULL. There is a description  of  the  pcre_extra
2158       The syntax and semantics of  the  regular  expressions  sup-         structure in the pcreapi documentation.
2159       ported  by PCRE are described below. Regular expressions are  
2160       also described in the Perl documentation and in a number  of         The  pattern_position field is present from version 1 of the pcre_call-
2161           out structure. It contains the offset to the next item to be matched in
2162       other  books,  some  of which have copious examples. Jeffrey         the pattern string.
2163       Friedl's  "Mastering  Regular  Expressions",  published   by  
2164       O'Reilly  (ISBN  1-56592-257),  covers them in great detail.         The  next_item_length field is present from version 1 of the pcre_call-
2165       The description here is intended as reference documentation.         out structure. It contains the length of the next item to be matched in
2166           the  pattern  string. When the callout immediately precedes an alterna-
2167       A regular expression is a pattern that is matched against  a         tion bar, a closing parenthesis, or the end of the pattern, the  length
2168       subject string from left to right. Most characters stand for         is  zero.  When the callout precedes an opening parenthesis, the length
2169       themselves in a pattern, and match the corresponding charac-         is that of the entire subpattern.
2170       ters in the subject. As a trivial example, the pattern  
2171           The pattern_position and next_item_length fields are intended  to  help
2172         The quick brown fox         in  distinguishing between different automatic callouts, which all have
2173           the same callout number. However, they are set for all callouts.
2174       matches a portion of a subject string that is  identical  to  
2175       itself.  The  power  of  regular  expressions comes from the  
2176       ability to include alternatives and repetitions in the  pat-  RETURN VALUES
2177       tern.  These  are encoded in the pattern by the use of meta-  
2178       characters, which do not stand for  themselves  but  instead         The external callout function returns an integer to PCRE. If the  value
2179       are interpreted in some special way.         is  zero,  matching  proceeds  as  normal. If the value is greater than
2180           zero, matching fails at the current point, but  the  testing  of  other
2181       There are two different sets of meta-characters: those  that         matching possibilities goes ahead, just as if a lookahead assertion had
2182       are  recognized anywhere in the pattern except within square         failed. If the value is less than zero, the  match  is  abandoned,  and
2183       brackets, and those that are recognized in square  brackets.         pcre_exec() (or pcre_dfa_exec()) returns the negative value.
2184       Outside square brackets, the meta-characters are as follows:  
2185           Negative   values   should   normally   be   chosen  from  the  set  of
2186         \      general escape character with several uses         PCRE_ERROR_xxx values. In particular, PCRE_ERROR_NOMATCH forces a stan-
2187         ^      assert start of  subject  (or  line,  in  multiline         dard  "no  match"  failure.   The  error  number  PCRE_ERROR_CALLOUT is
2188       mode)         reserved for use by callout functions; it will never be  used  by  PCRE
2189         $      assert end of subject (or line, in multiline mode)         itself.
2190         .      match any character except newline (by default)  
2191         [      start character class definition  Last updated: 28 February 2005
2192         |      start of alternative branch  Copyright (c) 1997-2005 University of Cambridge.
2193         (      start subpattern  -----------------------------------------------------------------------------
2194         )      end subpattern  
2195         ?      extends the meaning of (  
2196                also 0 or 1 quantifier  
2197                also quantifier minimizer  NAME
2198         *      0 or more quantifier         PCRE - Perl-compatible regular expressions
2199         +      1 or more quantifier  
2200         {      start min/max quantifier  
2201    DIFFERENCES BETWEEN PCRE AND PERL
2202       Part of a pattern that is in square  brackets  is  called  a  
2203       "character  class".  In  a  character  class  the only meta-         This  document describes the differences in the ways that PCRE and Perl
2204       characters are:         handle regular expressions. The differences  described  here  are  with
2205           respect to Perl 5.8.
        \      general escape character  
        ^      negate the class, but only if the first character  
        -      indicates character range  
        ]      terminates the character class  
2206    
2207       The following sections describe  the  use  of  each  of  the         1.  PCRE does not have full UTF-8 support. Details of what it does have
2208       meta-characters.         are given in the section on UTF-8 support in the main pcre page.
2209    
2210           2. PCRE does not allow repeat quantifiers on lookahead assertions. Perl
2211           permits  them,  but they do not mean what you might think. For example,
2212           (?!a){3} does not assert that the next three characters are not "a". It
2213           just asserts that the next character is not "a" three times.
2214    
2215           3.  Capturing  subpatterns  that occur inside negative lookahead asser-
2216           tions are counted, but their entries in the offsets  vector  are  never
2217           set.  Perl sets its numerical variables from any such patterns that are
2218           matched before the assertion fails to match something (thereby succeed-
2219           ing),  but  only  if the negative lookahead assertion contains just one
2220           branch.
2221    
2222           4. Though binary zero characters are supported in the  subject  string,
2223           they are not allowed in a pattern string because it is passed as a nor-
2224           mal C string, terminated by zero. The escape sequence \0 can be used in
2225           the pattern to represent a binary zero.
2226    
2227           5.  The  following Perl escape sequences are not supported: \l, \u, \L,
2228           \U, and \N. In fact these are implemented by Perl's general string-han-
2229           dling  and are not part of its pattern matching engine. If any of these
2230           are encountered by PCRE, an error is generated.
2231    
2232           6. The Perl escape sequences \p, \P, and \X are supported only if  PCRE
2233           is  built  with Unicode character property support. The properties that
2234           can be tested with \p and \P are limited to the general category  prop-
2235           erties such as Lu and Nd.
2236    
2237           7. PCRE does support the \Q...\E escape for quoting substrings. Charac-
2238           ters in between are treated as literals.  This  is  slightly  different
2239           from  Perl  in  that  $  and  @ are also handled as literals inside the
2240           quotes. In Perl, they cause variable interpolation (but of course  PCRE
2241           does not have variables). Note the following examples:
2242    
2243               Pattern            PCRE matches      Perl matches
2244    
2245               \Qabc$xyz\E        abc$xyz           abc followed by the
2246                                                      contents of $xyz
2247               \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2248               \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2249    
2250           The  \Q...\E  sequence  is recognized both inside and outside character
2251           classes.
2252    
2253           8. Fairly obviously, PCRE does not support the (?{code}) and (?p{code})
2254           constructions.  However,  there is support for recursive patterns using
2255           the non-Perl items (?R),  (?number),  and  (?P>name).  Also,  the  PCRE
2256           "callout"  feature allows an external function to be called during pat-
2257           tern matching. See the pcrecallout documentation for details.
2258    
2259           9. There are some differences that are concerned with the  settings  of
2260           captured  strings  when  part  of  a  pattern is repeated. For example,
2261           matching "aba" against the  pattern  /^(a(b)?)+$/  in  Perl  leaves  $2
2262           unset, but in PCRE it is set to "b".
2263    
2264           10. PCRE provides some extensions to the Perl regular expression facil-
2265           ities:
2266    
2267           (a) Although lookbehind assertions must  match  fixed  length  strings,
2268           each alternative branch of a lookbehind assertion can match a different
2269           length of string. Perl requires them all to have the same length.
2270    
2271           (b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the  $
2272           meta-character matches only at the very end of the string.
2273    
2274           (c) If PCRE_EXTRA is set, a backslash followed by a letter with no spe-
2275           cial meaning is faulted.
2276    
2277           (d) If PCRE_UNGREEDY is set, the greediness of the  repetition  quanti-
2278           fiers is inverted, that is, by default they are not greedy, but if fol-
2279           lowed by a question mark they are.
2280    
2281           (e) PCRE_ANCHORED can be used at matching time to force a pattern to be
2282           tried only at the first matching position in the subject string.
2283    
2284           (f)  The PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, and PCRE_NO_AUTO_CAP-
2285           TURE options for pcre_exec() have no Perl equivalents.
2286    
2287           (g) The (?R), (?number), and (?P>name) constructs allows for  recursive
2288           pattern  matching  (Perl  can  do  this using the (?p{code}) construct,
2289           which PCRE cannot support.)
2290    
2291           (h) PCRE supports named capturing substrings, using the Python  syntax.
2292    
2293           (i)  PCRE  supports  the  possessive quantifier "++" syntax, taken from
2294           Sun's Java package.
2295    
2296           (j) The (R) condition, for testing recursion, is a PCRE extension.
2297    
2298           (k) The callout facility is PCRE-specific.
2299    
2300           (l) The partial matching facility is PCRE-specific.
2301    
2302           (m) Patterns compiled by PCRE can be saved and re-used at a later time,
2303           even on different hosts that have the other endianness.
2304    
2305           (n)  The  alternative  matching function (pcre_dfa_exec()) matches in a
2306           different way and is not Perl-compatible.
2307    
2308    Last updated: 28 February 2005
2309    Copyright (c) 1997-2005 University of Cambridge.
2310    -----------------------------------------------------------------------------
2311    
2312    
2313    
2314    NAME
2315           PCRE - Perl-compatible regular expressions
2316    
2317    
2318    PCRE REGULAR EXPRESSION DETAILS
2319    
2320           The  syntax  and semantics of the regular expressions supported by PCRE
2321           are described below. Regular expressions are also described in the Perl
2322           documentation  and  in  a  number  of books, some of which have copious
2323           examples.  Jeffrey Friedl's "Mastering Regular Expressions",  published
2324           by  O'Reilly, covers regular expressions in great detail. This descrip-
2325           tion of PCRE's regular expressions is intended as reference material.
2326    
2327           The original operation of PCRE was on strings of  one-byte  characters.
2328           However,  there is now also support for UTF-8 character strings. To use
2329           this, you must build PCRE to  include  UTF-8  support,  and  then  call
2330           pcre_compile()  with  the  PCRE_UTF8  option.  How this affects pattern
2331           matching is mentioned in several places below. There is also a  summary
2332           of  UTF-8  features  in  the  section on UTF-8 support in the main pcre
2333           page.
2334    
2335           The remainder of this document discusses the  patterns  that  are  sup-
2336           ported  by  PCRE when its main matching function, pcre_exec(), is used.
2337           From  release  6.0,   PCRE   offers   a   second   matching   function,
2338           pcre_dfa_exec(),  which matches using a different algorithm that is not
2339           Perl-compatible. The advantages and disadvantages  of  the  alternative
2340           function, and how it differs from the normal function, are discussed in
2341           the pcrematching page.
2342    
2343           A regular expression is a pattern that is  matched  against  a  subject
2344           string  from  left  to right. Most characters stand for themselves in a
2345           pattern, and match the corresponding characters in the  subject.  As  a
2346           trivial example, the pattern
2347    
2348             The quick brown fox
2349    
2350           matches a portion of a subject string that is identical to itself. When
2351           caseless matching is specified (the PCRE_CASELESS option), letters  are
2352           matched  independently  of case. In UTF-8 mode, PCRE always understands
2353           the concept of case for characters whose values are less than  128,  so
2354           caseless  matching  is always possible. For characters with higher val-
2355           ues, the concept of case is supported if PCRE is compiled with  Unicode
2356           property  support,  but  not  otherwise.   If  you want to use caseless
2357           matching for characters 128 and above, you must  ensure  that  PCRE  is
2358           compiled with Unicode property support as well as with UTF-8 support.
2359    
2360           The  power  of  regular  expressions  comes from the ability to include
2361           alternatives and repetitions in the pattern. These are encoded  in  the
2362           pattern by the use of metacharacters, which do not stand for themselves
2363           but instead are interpreted in some special way.
2364    
2365           There are two different sets of metacharacters: those that  are  recog-
2366           nized  anywhere in the pattern except within square brackets, and those
2367           that are recognized in square brackets. Outside  square  brackets,  the
2368           metacharacters are as follows:
2369    
2370             \      general escape character with several uses
2371             ^      assert start of string (or line, in multiline mode)
2372             $      assert end of string (or line, in multiline mode)
2373             .      match any character except newline (by default)
2374             [      start character class definition
2375             |      start of alternative branch
2376             (      start subpattern
2377             )      end subpattern
2378             ?      extends the meaning of (
2379                    also 0 or 1 quantifier
2380                    also quantifier minimizer
2381             *      0 or more quantifier
2382             +      1 or more quantifier
2383                    also "possessive quantifier"
2384             {      start min/max quantifier
2385    
2386           Part  of  a  pattern  that is in square brackets is called a "character
2387           class". In a character class the only metacharacters are:
2388    
2389             \      general escape character
2390             ^      negate the class, but only if the first character
2391             -      indicates character range
2392             [      POSIX character class (only if followed by POSIX
2393                      syntax)
2394             ]      terminates the character class
2395    
2396           The following sections describe the use of each of the  metacharacters.
2397    
2398    
2399  BACKSLASH  BACKSLASH
      The backslash character has several uses. Firstly, if it  is  
      followed  by  a  non-alphameric character, it takes away any  
      special  meaning  that  character  may  have.  This  use  of  
      backslash  as  an  escape  character applies both inside and  
      outside character classes.  
   
      For example, if you want to match a "*" character, you write  
      "\*" in the pattern. This applies whether or not the follow-  
      ing character would otherwise  be  interpreted  as  a  meta-  
      character,  so it is always safe to precede a non-alphameric  
      with "\" to specify that it stands for itself.  In  particu-  
      lar, if you want to match a backslash, you write "\\".  
   
      If a pattern is compiled with the PCRE_EXTENDED option, whi-  
      tespace in the pattern (other than in a character class) and  
      characters between a "#" outside a character class  and  the  
      next  newline  character  are ignored. An escaping backslash  
      can be used to include a whitespace or "#" character as part  
      of the pattern.  
   
      A second use of backslash provides a way  of  encoding  non-  
      printing  characters  in patterns in a visible manner. There  
      is no restriction on the appearance of non-printing  charac-  
      ters,  apart from the binary zero that terminates a pattern,  
      but when a pattern is being prepared by text editing, it  is  
      usually  easier to use one of the following escape sequences  
      than the binary character it represents:  
   
        \a     alarm, that is, the BEL character (hex 07)  
        \cx    "control-x", where x is any character  
        \e     escape (hex 1B)  
        \f     formfeed (hex 0C)  
        \n     newline (hex 0A)  
        \r     carriage return (hex 0D)  
        \t     tab (hex 09)  
        \xhh   character with hex code hh  
        \ddd   character with octal code ddd, or backreference  
   
      The precise effect of "\cx" is as follows: if "x" is a lower  
      case  letter,  it  is converted to upper case. Then bit 6 of  
      the character (hex 40) is inverted.  Thus "\cz" becomes  hex  
      1A, but "\c{" becomes hex 3B, while "\c;" becomes hex 7B.  
   
      After "\x", up to two hexadecimal digits are  read  (letters  
      can be in upper or lower case).  
   
      After "\0" up to two further octal digits are read. In  both  
      cases,  if  there are fewer than two digits, just those that  
      are present are used. Thus the sequence "\0\x\07"  specifies  
      two binary zeros followed by a BEL character.  Make sure you  
      supply two digits after the initial zero  if  the  character  
      that follows is itself an octal digit.  
   
      The handling of a backslash followed by a digit other than 0  
      is  complicated.   Outside  a character class, PCRE reads it  
      and any following digits as a decimal number. If the  number  
      is  less  than  10, or if there have been at least that many  
      previous capturing left parentheses in the  expression,  the  
      entire  sequence is taken as a back reference. A description  
      of how this works is given later, following  the  discussion  
      of parenthesized subpatterns.  
   
      Inside a character  class,  or  if  the  decimal  number  is  
      greater  than  9 and there have not been that many capturing  
      subpatterns, PCRE re-reads up to three octal digits  follow-  
      ing  the  backslash,  and  generates  a single byte from the  
      least significant 8 bits of the value. Any subsequent digits  
      stand for themselves.  For example:  
   
        \040   is another way of writing a space  
        \40    is the same, provided there are fewer than 40  
                  previous capturing subpatterns  
        \7     is always a back reference  
        \11    might be a back reference, or another way of  
                  writing a tab  
        \011   is always a tab  
        \0113  is a tab followed by the character "3"  
        \113   is the character with octal code 113 (since there  
                  can be no more than 99 back references)  
        \377   is a byte consisting entirely of 1 bits  
        \81    is either a back reference, or a binary zero  
                  followed by the two characters "8" and "1"  
   
      Note that octal values of 100 or greater must not be  intro-  
      duced  by  a  leading zero, because no more than three octal  
      digits are ever read.  
   
      All the sequences that define a single  byte  value  can  be  
      used both inside and outside character classes. In addition,  
      inside a character class, the sequence "\b"  is  interpreted  
      as  the  backspace  character  (hex 08). Outside a character  
      class it has a different meaning (see below).  
   
      The third use of backslash is for specifying generic charac-  
      ter types:  
   
        \d     any decimal digit  
        \D     any character that is not a decimal digit  
        \s     any whitespace character  
        \S     any character that is not a whitespace character  
        \w     any "word" character  
        \W     any "non-word" character  
   
      Each pair of escape sequences partitions the complete set of  
      characters  into  two  disjoint  sets.  Any  given character  
      matches one, and only one, of each pair.  
   
      A "word" character is any letter or digit or the  underscore  
      character,  that  is,  any  character which can be part of a  
      Perl "word". The definition of letters and  digits  is  con-  
      trolled  by PCRE's character tables, and may vary if locale-  
      specific matching is  taking  place  (see  "Locale  support"  
      above). For example, in the "fr" (French) locale, some char-  
      acter codes greater than 128 are used for accented  letters,  
      and these are matched by \w.  
   
      These character type sequences can appear  both  inside  and  
      outside  character classes. They each match one character of  
      the appropriate type. If the current matching  point  is  at  
      the end of the subject string, all of them fail, since there  
      is no character to match.  
   
      The fourth use of backslash is  for  certain  simple  asser-  
      tions. An assertion specifies a condition that has to be met  
      at a particular point in  a  match,  without  consuming  any  
      characters  from  the subject string. The use of subpatterns  
      for more complicated  assertions  is  described  below.  The  
      backslashed assertions are  
   
        \b     word boundary  
        \B     not a word boundary  
        \A     start of subject (independent of multiline mode)  
        \Z     end of subject or newline at  end  (independent  of  
      multiline mode)  
        \z     end of subject (independent of multiline mode)  
   
      These assertions may not appear in  character  classes  (but  
      note that "\b" has a different meaning, namely the backspace  
      character, inside a character class).  
   
      A word boundary is a position in the  subject  string  where  
      the current character and the previous character do not both  
      match \w or \W (i.e. one matches \w and  the  other  matches  
      \W),  or the start or end of the string if the first or last  
      character matches \w, respectively.  
   
      The \A, \Z, and \z assertions differ  from  the  traditional  
      circumflex  and  dollar  (described below) in that they only  
      ever match at the very start and end of the subject  string,  
      whatever  options  are  set.  They  are  not affected by the  
      PCRE_NOTBOL or PCRE_NOTEOL options. If the startoffset argu-  
      ment  of  pcre_exec()  is  non-zero, \A can never match. The  
      difference between \Z and \z is that  \Z  matches  before  a  
      newline  that is the last character of the string as well as  
      at the end of the string, whereas \z  matches  only  at  the  
      end.  
2400    
2401           The backslash character has several uses. Firstly, if it is followed by
2402           a non-alphanumeric character, it takes away any  special  meaning  that
2403           character  may  have.  This  use  of  backslash  as an escape character
2404           applies both inside and outside character classes.
2405    
2406           For example, if you want to match a * character, you write  \*  in  the
2407           pattern.   This  escaping  action  applies whether or not the following
2408           character would otherwise be interpreted as a metacharacter, so  it  is
2409           always  safe  to  precede  a non-alphanumeric with backslash to specify
2410           that it stands for itself. In particular, if you want to match a  back-
2411           slash, you write \\.
2412    
2413           If  a  pattern is compiled with the PCRE_EXTENDED option, whitespace in
2414           the pattern (other than in a character class) and characters between  a
2415           # outside a character class and the next newline character are ignored.
2416           An escaping backslash can be used to include a whitespace or #  charac-
2417           ter as part of the pattern.
2418    
2419           If  you  want  to remove the special meaning from a sequence of charac-
2420           ters, you can do so by putting them between \Q and \E. This is  differ-
2421           ent  from  Perl  in  that  $  and  @ are handled as literals in \Q...\E
2422           sequences in PCRE, whereas in Perl, $ and @ cause  variable  interpola-
2423           tion. Note the following examples:
2424    
2425             Pattern            PCRE matches   Perl matches
2426    
2427             \Qabc$xyz\E        abc$xyz        abc followed by the
2428                                                 contents of $xyz
2429             \Qabc\$xyz\E       abc\$xyz       abc\$xyz
2430             \Qabc\E\$\Qxyz\E   abc$xyz        abc$xyz
2431    
2432           The  \Q...\E  sequence  is recognized both inside and outside character
2433           classes.
2434    
2435       Non-printing characters
2436    
2437           A second use of backslash provides a way of encoding non-printing char-
2438           acters  in patterns in a visible manner. There is no restriction on the
2439           appearance of non-printing characters, apart from the binary zero  that
2440           terminates  a  pattern,  but  when  a pattern is being prepared by text
2441           editing, it is usually easier  to  use  one  of  the  following  escape
2442           sequences than the binary character it represents:
2443    
2444             \a        alarm, that is, the BEL character (hex 07)
2445             \cx       "control-x", where x is any character
2446             \e        escape (hex 1B)
2447             \f        formfeed (hex 0C)
2448             \n        newline (hex 0A)
2449             \r        carriage return (hex 0D)
2450             \t        tab (hex 09)
2451             \ddd      character with octal code ddd, or backreference
2452             \xhh      character with hex code hh
2453             \x{hhh..} character with hex code hhh... (UTF-8 mode only)
2454    
2455           The  precise  effect of \cx is as follows: if x is a lower case letter,
2456           it is converted to upper case. Then bit 6 of the character (hex 40)  is
2457           inverted.   Thus  \cz becomes hex 1A, but \c{ becomes hex 3B, while \c;
2458           becomes hex 7B.
2459    
2460           After \x, from zero to two hexadecimal digits are read (letters can  be
2461           in  upper or lower case). In UTF-8 mode, any number of hexadecimal dig-
2462           its may appear between \x{ and }, but the value of the  character  code
2463           must  be  less  than  2**31  (that is, the maximum hexadecimal value is
2464           7FFFFFFF). If characters other than hexadecimal digits  appear  between
2465           \x{  and }, or if there is no terminating }, this form of escape is not
2466           recognized. Instead, the initial \x will  be  interpreted  as  a  basic
2467           hexadecimal  escape, with no following digits, giving a character whose
2468           value is zero.
2469    
2470           Characters whose value is less than 256 can be defined by either of the
2471           two  syntaxes for \x when PCRE is in UTF-8 mode. There is no difference
2472           in the way they are handled. For example, \xdc is exactly the  same  as
2473           \x{dc}.
2474    
2475           After  \0  up  to  two further octal digits are read. In both cases, if
2476           there are fewer than two digits, just those that are present are  used.
2477           Thus  the sequence \0\x\07 specifies two binary zeros followed by a BEL
2478           character (code value 7). Make sure you supply  two  digits  after  the
2479           initial  zero  if the pattern character that follows is itself an octal
2480           digit.
2481    
2482           The handling of a backslash followed by a digit other than 0 is compli-
2483           cated.  Outside a character class, PCRE reads it and any following dig-
2484           its as a decimal number. If the number is less than  10,  or  if  there
2485           have been at least that many previous capturing left parentheses in the
2486           expression, the entire  sequence  is  taken  as  a  back  reference.  A
2487           description  of how this works is given later, following the discussion
2488           of parenthesized subpatterns.
2489    
2490           Inside a character class, or if the decimal number is  greater  than  9
2491           and  there have not been that many capturing subpatterns, PCRE re-reads
2492           up to three octal digits following the backslash, and generates a  sin-
2493           gle byte from the least significant 8 bits of the value. Any subsequent
2494           digits stand for themselves.  For example:
2495    
2496             \040   is another way of writing a space
2497             \40    is the same, provided there are fewer than 40
2498                       previous capturing subpatterns
2499             \7     is always a back reference
2500             \11    might be a back reference, or another way of
2501                       writing a tab
2502             \011   is always a tab
2503             \0113  is a tab followed by the character "3"
2504             \113   might be a back reference, otherwise the
2505                       character with octal code 113
2506             \377   might be a back reference, otherwise
2507                       the byte consisting entirely of 1 bits
2508             \81    is either a back reference, or a binary zero
2509                       followed by the two characters "8" and "1"
2510    
2511           Note that octal values of 100 or greater must not be  introduced  by  a
2512           leading zero, because no more than three octal digits are ever read.
2513    
2514           All  the  sequences  that  define a single byte value or a single UTF-8
2515           character (in UTF-8 mode) can be used both inside and outside character
2516           classes.  In  addition,  inside  a  character class, the sequence \b is
2517           interpreted as the backspace character (hex 08), and the sequence \X is
2518           interpreted  as  the  character  "X".  Outside a character class, these
2519           sequences have different meanings (see below).
2520    
2521       Generic character types
2522    
2523           The third use of backslash is for specifying generic  character  types.
2524           The following are always recognized:
2525    
2526             \d     any decimal digit
2527             \D     any character that is not a decimal digit
2528             \s     any whitespace character
2529             \S     any character that is not a whitespace character
2530             \w     any "word" character
2531             \W     any "non-word" character
2532    
2533           Each pair of escape sequences partitions the complete set of characters
2534           into two disjoint sets. Any given character matches one, and only  one,
2535           of each pair.
2536    
2537           These character type sequences can appear both inside and outside char-
2538           acter classes. They each match one character of the  appropriate  type.
2539           If  the current matching point is at the end of the subject string, all
2540           of them fail, since there is no character to match.
2541    
2542           For compatibility with Perl, \s does not match the VT  character  (code
2543           11).   This makes it different from the the POSIX "space" class. The \s
2544           characters are HT (9), LF (10), FF (12), CR (13), and space (32).
2545    
2546           A "word" character is an underscore or any character less than 256 that
2547           is  a  letter  or  digit.  The definition of letters and digits is con-
2548           trolled by PCRE's low-valued character tables, and may vary if  locale-
2549           specific  matching is taking place (see "Locale support" in the pcreapi
2550           page). For example, in the  "fr_FR"  (French)  locale,  some  character
2551           codes  greater  than  128  are used for accented letters, and these are
2552           matched by \w.
2553    
2554           In UTF-8 mode, characters with values greater than 128 never match  \d,
2555           \s, or \w, and always match \D, \S, and \W. This is true even when Uni-
2556           code character property support is available.
2557    
2558       Unicode character properties
2559    
2560           When PCRE is built with Unicode character property support, three addi-
2561           tional  escape sequences to match generic character types are available
2562           when UTF-8 mode is selected. They are:
2563    
2564            \p{xx}   a character with the xx property
2565            \P{xx}   a character without the xx property
2566            \X       an extended Unicode sequence
2567    
2568           The property names represented by xx above are limited to  the  Unicode
2569           general  category properties. Each character has exactly one such prop-
2570           erty, specified by a two-letter abbreviation.  For  compatibility  with
2571           Perl,  negation  can be specified by including a circumflex between the
2572           opening brace and the property name. For example, \p{^Lu} is  the  same
2573           as \P{Lu}.
2574    
2575           If  only  one  letter  is  specified with \p or \P, it includes all the
2576           properties that start with that letter. In this case, in the absence of
2577           negation, the curly brackets in the escape sequence are optional; these
2578           two examples have the same effect:
2579    
2580             \p{L}
2581             \pL
2582    
2583           The following property codes are supported:
2584    
2585             C     Other
2586             Cc    Control
2587             Cf    Format
2588             Cn    Unassigned
2589             Co    Private use
2590             Cs    Surrogate
2591    
2592             L     Letter
2593             Ll    Lower case letter
2594             Lm    Modifier letter
2595             Lo    Other letter
2596             Lt    Title case letter
2597             Lu    Upper case letter
2598    
2599             M     Mark
2600             Mc    Spacing mark
2601             Me    Enclosing mark
2602             Mn    Non-spacing mark
2603    
2604             N     Number
2605             Nd    Decimal number
2606             Nl    Letter number
2607             No    Other number
2608    
2609             P     Punctuation
2610             Pc    Connector punctuation
2611             Pd    Dash punctuation
2612             Pe    Close punctuation
2613             Pf    Final punctuation
2614             Pi    Initial punctuation
2615             Po    Other punctuation
2616             Ps    Open punctuation
2617    
2618             S     Symbol
2619             Sc    Currency symbol
2620             Sk    Modifier symbol
2621             Sm    Mathematical symbol
2622             So    Other symbol
2623    
2624             Z     Separator
2625             Zl    Line separator
2626             Zp    Paragraph separator
2627             Zs    Space separator
2628    
2629           Extended properties such as "Greek" or "InMusicalSymbols" are not  sup-
2630           ported by PCRE.
2631    
2632           Specifying  caseless  matching  does not affect these escape sequences.
2633           For example, \p{Lu} always matches only upper case letters.
2634    
2635           The \X escape matches any number of Unicode  characters  that  form  an
2636           extended Unicode sequence. \X is equivalent to
2637    
2638             (?>\PM\pM*)
2639    
2640           That  is,  it matches a character without the "mark" property, followed
2641           by zero or more characters with the "mark"  property,  and  treats  the
2642           sequence  as  an  atomic group (see below).  Characters with the "mark"
2643           property are typically accents that affect the preceding character.
2644    
2645           Matching characters by Unicode property is not fast, because  PCRE  has
2646           to  search  a  structure  that  contains data for over fifteen thousand
2647           characters. That is why the traditional escape sequences such as \d and
2648           \w do not use Unicode properties in PCRE.
2649    
2650       Simple assertions
2651    
2652           The fourth use of backslash is for certain simple assertions. An asser-
2653           tion specifies a condition that has to be met at a particular point  in
2654           a  match, without consuming any characters from the subject string. The
2655           use of subpatterns for more complicated assertions is described  below.
2656           The backslashed assertions are:
2657    
2658             \b     matches at a word boundary
2659             \B     matches when not at a word boundary
2660             \A     matches at start of subject
2661             \Z     matches at end of subject or before newline at end
2662             \z     matches at end of subject
2663             \G     matches at first matching position in subject
2664    
2665           These  assertions may not appear in character classes (but note that \b
2666           has a different meaning, namely the backspace character, inside a char-
2667           acter class).
2668    
2669           A  word  boundary is a position in the subject string where the current
2670           character and the previous character do not both match \w or  \W  (i.e.
2671           one  matches  \w  and the other matches \W), or the start or end of the
2672           string if the first or last character matches \w, respectively.
2673    
2674           The \A, \Z, and \z assertions differ from  the  traditional  circumflex
2675           and dollar (described in the next section) in that they only ever match
2676           at the very start and end of the subject string, whatever  options  are
2677           set.  Thus,  they are independent of multiline mode. These three asser-
2678           tions are not affected by the PCRE_NOTBOL or PCRE_NOTEOL options, which
2679           affect  only the behaviour of the circumflex and dollar metacharacters.
2680           However, if the startoffset argument of pcre_exec() is non-zero,  indi-
2681           cating that matching is to start at a point other than the beginning of
2682           the subject, \A can never match. The difference between \Z  and  \z  is
2683           that  \Z  matches  before  a  newline that is the last character of the
2684           string as well as at the end of the string, whereas \z matches only  at
2685           the end.
2686    
2687           The  \G assertion is true only when the current matching position is at
2688           the start point of the match, as specified by the startoffset  argument
2689           of  pcre_exec().  It  differs  from \A when the value of startoffset is
2690           non-zero. By calling pcre_exec() multiple times with appropriate  argu-
2691           ments, you can mimic Perl's /g option, and it is in this kind of imple-
2692           mentation where \G can be useful.
2693    
2694           Note, however, that PCRE's interpretation of \G, as the  start  of  the
2695           current match, is subtly different from Perl's, which defines it as the
2696           end of the previous match. In Perl, these can  be  different  when  the
2697           previously  matched  string was empty. Because PCRE does just one match
2698           at a time, it cannot reproduce this behaviour.
2699    
2700           If all the alternatives of a pattern begin with \G, the  expression  is
2701           anchored to the starting match position, and the "anchored" flag is set
2702           in the compiled regular expression.
2703    
2704    
2705  CIRCUMFLEX AND DOLLAR  CIRCUMFLEX AND DOLLAR
      Outside a character class, in the default matching mode, the  
      circumflex  character  is an assertion which is true only if  
      the current matching point is at the start  of  the  subject  
      string.  If  the startoffset argument of pcre_exec() is non-  
      zero, circumflex can never match. Inside a character  class,  
      circumflex has an entirely different meaning (see below).  
   
      Circumflex need not be the first character of the pattern if  
      a  number of alternatives are involved, but it should be the  
      first thing in each alternative in which it appears  if  the  
      pattern is ever to match that branch. If all possible alter-  
      natives start with a circumflex, that is, if the pattern  is  
      constrained to match only at the start of the subject, it is  
      said to be an "anchored" pattern. (There are also other con-  
      structs that can cause a pattern to be anchored.)  
   
      A dollar character is an assertion which is true only if the  
      current  matching point is at the end of the subject string,  
      or immediately before a newline character that is  the  last  
      character in the string (by default). Dollar need not be the  
      last character of the pattern if a  number  of  alternatives  
      are  involved,  but it should be the last item in any branch  
      in which it appears.  Dollar has no  special  meaning  in  a  
      character class.  
   
      The meaning of dollar can be changed so that it matches only  
      at   the   very   end   of   the   string,  by  setting  the  
      PCRE_DOLLAR_ENDONLY option at compile or matching time. This  
      does not affect the \Z assertion.  
   
      The meanings of the circumflex  and  dollar  characters  are  
      changed  if  the  PCRE_MULTILINE option is set. When this is  
      the case,  they  match  immediately  after  and  immediately  
      before an internal "\n" character, respectively, in addition  
      to matching at the start and end of the subject string.  For  
      example,  the  pattern  /^abc$/  matches  the subject string  
      "def\nabc" in multiline  mode,  but  not  otherwise.  Conse-  
      quently,  patterns  that  are  anchored  in single line mode  
      because all branches start with "^" are not anchored in mul-  
      tiline mode, and a match for circumflex is possible when the  
      startoffset  argument  of  pcre_exec()  is   non-zero.   The  
      PCRE_DOLLAR_ENDONLY  option  is ignored if PCRE_MULTILINE is  
      set.  
   
      Note that the sequences \A, \Z, and \z can be used to  match  
      the  start  and end of the subject in both modes, and if all  
      branches of a pattern start with \A is it  always  anchored,  
      whether PCRE_MULTILINE is set or not.  
2706    
2707           Outside a character class, in the default matching mode, the circumflex
2708           character  is  an  assertion  that is true only if the current matching
2709           point is at the start of the subject string. If the  startoffset  argu-
2710           ment  of  pcre_exec()  is  non-zero,  circumflex can never match if the
2711           PCRE_MULTILINE option is unset. Inside a  character  class,  circumflex
2712           has an entirely different meaning (see below).
2713    
2714           Circumflex  need  not be the first character of the pattern if a number
2715           of alternatives are involved, but it should be the first thing in  each
2716           alternative  in  which  it appears if the pattern is ever to match that
2717           branch. If all possible alternatives start with a circumflex, that  is,
2718           if  the  pattern  is constrained to match only at the start of the sub-
2719           ject, it is said to be an "anchored" pattern.  (There  are  also  other
2720           constructs that can cause a pattern to be anchored.)
2721    
2722           A  dollar  character  is  an assertion that is true only if the current
2723           matching point is at the end of  the  subject  string,  or  immediately
2724           before a newline character that is the last character in the string (by
2725           default). Dollar need not be the last character of  the  pattern  if  a
2726           number  of alternatives are involved, but it should be the last item in
2727           any branch in which it appears.  Dollar has no  special  meaning  in  a
2728           character class.
2729    
2730           The  meaning  of  dollar  can be changed so that it matches only at the
2731           very end of the string, by setting the  PCRE_DOLLAR_ENDONLY  option  at
2732           compile time. This does not affect the \Z assertion.
2733    
2734           The meanings of the circumflex and dollar characters are changed if the
2735           PCRE_MULTILINE option is set. When this is the case, they match immedi-
2736           ately  after  and  immediately  before  an  internal newline character,
2737           respectively, in addition to matching at the start and end of the  sub-
2738           ject  string.  For  example,  the  pattern  /^abc$/ matches the subject
2739           string "def\nabc" (where \n represents a newline character)  in  multi-
2740           line mode, but not otherwise.  Consequently, patterns that are anchored
2741           in single line mode because all branches start with ^ are not  anchored
2742           in  multiline  mode,  and  a  match for circumflex is possible when the
2743           startoffset  argument  of  pcre_exec()  is  non-zero.   The   PCRE_DOL-
2744           LAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
2745    
2746           Note  that  the sequences \A, \Z, and \z can be used to match the start
2747           and end of the subject in both modes, and if all branches of a  pattern
2748           start  with  \A it is always anchored, whether PCRE_MULTILINE is set or
2749           not.
2750    
2751    
2752  FULL STOP (PERIOD, DOT)  FULL STOP (PERIOD, DOT)
      Outside a character class, a dot in the pattern matches  any  
      one character in the subject, including a non-printing char-  
      acter, but not (by default)  newline.   If  the  PCRE_DOTALL  
      option  is set, dots match newlines as well. The handling of  
      dot is entirely independent of the  handling  of  circumflex  
      and  dollar,  the  only  relationship  being  that they both  
      involve newline characters. Dot has no special meaning in  a  
      character class.  
   
   
   
 SQUARE BRACKETS  
      An opening square bracket introduces a character class, ter-  
      minated  by  a  closing  square  bracket.  A  closing square  
      bracket on its own is  not  special.  If  a  closing  square  
      bracket  is  required as a member of the class, it should be  
      the first data character in the class (after an initial cir-  
      cumflex, if present) or escaped with a backslash.  
   
      A character class matches a single character in the subject;  
      the  character  must  be in the set of characters defined by  
      the class, unless the first character in the class is a cir-  
      cumflex,  in which case the subject character must not be in  
      the set defined by the class. If a  circumflex  is  actually  
      required  as  a  member  of  the class, ensure it is not the  
      first character, or escape it with a backslash.  
   
      For example, the character class [aeiou] matches  any  lower  
      case vowel, while [^aeiou] matches any character that is not  
      a lower case vowel. Note that a circumflex is  just  a  con-  
      venient  notation for specifying the characters which are in  
      the class by enumerating those that are not. It  is  not  an  
      assertion:  it  still  consumes a character from the subject  
      string, and fails if the current pointer is at  the  end  of  
      the string.  
   
      When caseless matching  is  set,  any  letters  in  a  class  
      represent  both their upper case and lower case versions, so  
      for example, a caseless [aeiou] matches "A" as well as  "a",  
      and  a caseless [^aeiou] does not match "A", whereas a case-  
      ful version would.  
   
      The newline character is never treated in any special way in  
      character  classes,  whatever the setting of the PCRE_DOTALL  
      or PCRE_MULTILINE options is. A  class  such  as  [^a]  will  
      always match a newline.  
   
      The minus (hyphen) character can be used to specify a  range  
      of  characters  in  a  character  class.  For example, [d-m]  
      matches any letter between d and m, inclusive.  If  a  minus  
      character  is required in a class, it must be escaped with a  
      backslash or appear in a position where it cannot be  inter-  
      preted as indicating a range, typically as the first or last  
      character in the class.  
   
      It is not possible to have the literal character "]" as  the  
      end  character  of  a  range.  A  pattern such as [W-]46] is  
      interpreted as a class of two characters ("W" and "-")  fol-  
      lowed by a literal string "46]", so it would match "W46]" or  
      "-46]". However, if the "]" is escaped with a  backslash  it  
      is  interpreted  as  the end of range, so [W-\]46] is inter-  
      preted as a single class containing a range followed by  two  
      separate characters. The octal or hexadecimal representation  
      of "]" can also be used to end a range.  
   
      Ranges operate in ASCII collating sequence. They can also be  
      used  for  characters  specified  numerically,  for  example  
      [\000-\037]. If a range that includes letters is  used  when  
      caseless  matching  is set, it matches the letters in either  
      case. For example, [W-c] is equivalent  to  [][\^_`wxyzabc],  
      matched  caselessly,  and  if  character tables for the "fr"  
      locale are in use, [\xc8-\xcb] matches accented E characters  
      in both cases.  
   
      The character types \d, \D, \s, \S,  \w,  and  \W  may  also  
      appear  in  a  character  class, and add the characters that  
      they match to the class. For example, [\dABCDEF] matches any  
      hexadecimal  digit.  A  circumflex  can conveniently be used  
      with the upper case character types to specify a  more  res-  
      tricted set of characters than the matching lower case type.  
      For example, the class [^\W_] matches any letter  or  digit,  
      but not underscore.  
   
      All non-alphameric characters other than \,  -,  ^  (at  the  
      start)  and  the  terminating ] are non-special in character  
      classes, but it does no harm if they are escaped.  
2753    
2754           Outside a character class, a dot in the pattern matches any one charac-
2755           ter  in  the  subject,  including a non-printing character, but not (by
2756           default) newline.  In UTF-8 mode, a dot matches  any  UTF-8  character,
2757           which might be more than one byte long, except (by default) newline. If
2758           the PCRE_DOTALL option is set, dots match newlines as  well.  The  han-
2759           dling  of dot is entirely independent of the handling of circumflex and
2760           dollar, the only relationship being  that  they  both  involve  newline
2761           characters. Dot has no special meaning in a character class.
2762    
2763    
2764    MATCHING A SINGLE BYTE
2765    
2766           Outside a character class, the escape sequence \C matches any one byte,
2767           both in and out of UTF-8 mode. Unlike a dot, it can  match  a  newline.
2768           The  feature  is provided in Perl in order to match individual bytes in
2769           UTF-8 mode. Because it  breaks  up  UTF-8  characters  into  individual
2770           bytes,  what remains in the string may be a malformed UTF-8 string. For
2771           this reason, the \C escape sequence is best avoided.
2772    
2773           PCRE does not allow \C to appear in  lookbehind  assertions  (described
2774           below),  because  in UTF-8 mode this would make it impossible to calcu-
2775           late the length of the lookbehind.
2776    
2777    
2778    SQUARE BRACKETS AND CHARACTER CLASSES
2779    
2780           An opening square bracket introduces a character class, terminated by a
2781           closing square bracket. A closing square bracket on its own is not spe-
2782           cial. If a closing square bracket is required as a member of the class,
2783           it  should  be  the first data character in the class (after an initial
2784           circumflex, if present) or escaped with a backslash.
2785    
2786           A character class matches a single character in the subject.  In  UTF-8
2787           mode,  the character may occupy more than one byte. A matched character
2788           must be in the set of characters defined by the class, unless the first
2789           character  in  the  class definition is a circumflex, in which case the
2790           subject character must not be in the set defined by  the  class.  If  a
2791           circumflex  is actually required as a member of the class, ensure it is
2792           not the first character, or escape it with a backslash.
2793    
2794           For example, the character class [aeiou] matches any lower case  vowel,
2795           while  [^aeiou]  matches  any character that is not a lower case vowel.
2796           Note that a circumflex is just a convenient notation for specifying the
2797           characters  that  are in the class by enumerating those that are not. A
2798           class that starts with a circumflex is not an assertion: it still  con-
2799           sumes  a  character  from the subject string, and therefore it fails if
2800           the current pointer is at the end of the string.
2801    
2802           In UTF-8 mode, characters with values greater than 255 can be  included
2803           in  a  class as a literal string of bytes, or by using the \x{ escaping
2804           mechanism.
2805    
2806           When caseless matching is set, any letters in a  class  represent  both
2807           their  upper  case  and lower case versions, so for example, a caseless
2808           [aeiou] matches "A" as well as "a", and a caseless  [^aeiou]  does  not
2809           match  "A", whereas a caseful version would. In UTF-8 mode, PCRE always
2810           understands the concept of case for characters whose  values  are  less
2811           than  128, so caseless matching is always possible. For characters with
2812           higher values, the concept of case is supported  if  PCRE  is  compiled
2813           with  Unicode  property support, but not otherwise.  If you want to use
2814           caseless matching for characters 128 and above, you  must  ensure  that
2815           PCRE  is  compiled  with Unicode property support as well as with UTF-8
2816           support.
2817    
2818           The newline character is never treated in any special way in  character
2819           classes,  whatever  the  setting  of  the PCRE_DOTALL or PCRE_MULTILINE
2820           options is. A class such as [^a] will always match a newline.
2821    
2822           The minus (hyphen) character can be used to specify a range of  charac-
2823           ters  in  a  character  class.  For  example,  [d-m] matches any letter
2824           between d and m, inclusive. If a  minus  character  is  required  in  a
2825           class,  it  must  be  escaped  with a backslash or appear in a position
2826           where it cannot be interpreted as indicating a range, typically as  the
2827           first or last character in the class.
2828    
2829           It is not possible to have the literal character "]" as the end charac-
2830           ter of a range. A pattern such as [W-]46] is interpreted as a class  of
2831           two  characters ("W" and "-") followed by a literal string "46]", so it
2832           would match "W46]" or "-46]". However, if the "]"  is  escaped  with  a
2833           backslash  it is interpreted as the end of range, so [W-\]46] is inter-
2834           preted as a class containing a range followed by two other  characters.
2835           The  octal or hexadecimal representation of "]" can also be used to end
2836           a range.
2837    
2838           Ranges operate in the collating sequence of character values. They  can
2839           also   be  used  for  characters  specified  numerically,  for  example
2840           [\000-\037]. In UTF-8 mode, ranges can include characters whose  values
2841           are greater than 255, for example [\x{100}-\x{2ff}].
2842    
2843           If a range that includes letters is used when caseless matching is set,
2844           it matches the letters in either case. For example, [W-c] is equivalent
2845           to  [][\\^_`wxyzabc],  matched  caselessly,  and  in non-UTF-8 mode, if
2846           character tables for the "fr_FR" locale are in use, [\xc8-\xcb] matches
2847           accented  E  characters in both cases. In UTF-8 mode, PCRE supports the
2848           concept of case for characters with values greater than 128  only  when
2849           it is compiled with Unicode property support.
2850    
2851           The  character types \d, \D, \p, \P, \s, \S, \w, and \W may also appear
2852           in a character class, and add the characters that  they  match  to  the
2853           class. For example, [\dABCDEF] matches any hexadecimal digit. A circum-
2854           flex can conveniently be used with the upper case  character  types  to
2855           specify  a  more  restricted  set of characters than the matching lower
2856           case type. For example, the class [^\W_] matches any letter  or  digit,
2857           but not underscore.
2858    
2859           The  only  metacharacters  that are recognized in character classes are
2860           backslash, hyphen (only where it can be  interpreted  as  specifying  a
2861           range),  circumflex  (only  at the start), opening square bracket (only
2862           when it can be interpreted as introducing a POSIX class name - see  the
2863           next  section),  and  the  terminating closing square bracket. However,
2864           escaping other non-alphanumeric characters does no harm.
2865    
2866    
2867  POSIX CHARACTER CLASSES  POSIX CHARACTER CLASSES
      Perl 5.6 (not yet released at the time of writing) is  going  
      to  support  the POSIX notation for character classes, which  
      uses names enclosed by  [:  and  :]   within  the  enclosing  
      square brackets. PCRE supports this notation. For example,  
   
        [01[:alpha:]%]  
   
      matches "0", "1", any alphabetic character, or "%". The sup-  
      ported class names are  
   
        alnum    letters and digits  
        alpha    letters  
        ascii    character codes 0 - 127  
        cntrl    control characters  
        digit    decimal digits (same as \d)  
        graph    printing characters, excluding space  
        lower    lower case letters  
        print    printing characters, including space  
        punct    printing characters, excluding letters and digits  
        space    white space (same as \s)  
        upper    upper case letters  
        word     "word" characters (same as \w)  
        xdigit   hexadecimal digits  
   
      The names "ascii" and "word" are  Perl  extensions.  Another  
      Perl  extension is negation, which is indicated by a ^ char-  
      acter after the colon. For example,  
   
        [12[:^digit:]]  
   
      matches "1", "2", or any non-digit.  PCRE  (and  Perl)  also  
      recogize  the POSIX syntax [.ch.] and [=ch=] where "ch" is a  
      "collating element", but these are  not  supported,  and  an  
      error is given if they are encountered.  
2868    
2869           Perl supports the POSIX notation for character classes. This uses names
2870           enclosed  by  [: and :] within the enclosing square brackets. PCRE also
2871           supports this notation. For example,
2872    
2873             [01[:alpha:]%]
2874    
2875           matches "0", "1", any alphabetic character, or "%". The supported class
2876           names are
2877    
2878             alnum    letters and digits
2879             alpha    letters
2880             ascii    character codes 0 - 127
2881             blank    space or tab only
2882             cntrl    control characters