/[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 69 by nigel, Sat Feb 24 21:40:18 2007 UTC revision 1194 by ph10, Wed Oct 31 17:42:29 2012 UTC
# Line 1  Line 1 
1    -----------------------------------------------------------------------------
2  This file contains a concatenation of the PCRE man pages, converted to plain  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  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  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  synopses of each function in the library have not been included. Neither has
6  separate text files for the pcregrep and pcretest commands.  the pcredemo program. There are separate text files for the pcregrep and
7    pcretest commands.
8  -----------------------------------------------------------------------------  -----------------------------------------------------------------------------
9    
10    
11    PCRE(3)                                                                PCRE(3)
12    
13    
14    NAME
15           PCRE - Perl-compatible regular expressions
16    
17    
18    INTRODUCTION
19    
20           The  PCRE  library is a set of functions that implement regular expres-
21           sion pattern matching using the same syntax and semantics as Perl, with
22           just  a few differences. Some features that appeared in Python and PCRE
23           before they appeared in Perl are also available using the  Python  syn-
24           tax,  there  is  some  support for one or two .NET and Oniguruma syntax
25           items, and there is an option for requesting some  minor  changes  that
26           give better JavaScript compatibility.
27    
28           Starting with release 8.30, it is possible to compile two separate PCRE
29           libraries:  the  original,  which  supports  8-bit  character   strings
30           (including  UTF-8  strings),  and a second library that supports 16-bit
31           character strings (including UTF-16 strings). The build process  allows
32           either  one  or both to be built. The majority of the work to make this
33           possible was done by Zoltan Herczeg.
34    
35           Starting with release 8.32 it is possible to compile a  third  separate
36           PCRE library, which supports 32-bit character strings (including UTF-32
37           strings). The build process allows any set of the 8-,  16-  and  32-bit
38           libraries. The work to make this possible was done by Christian Persch.
39    
40           The  three  libraries  contain identical sets of functions, except that
41           the names in the 16-bit library start with pcre16_  instead  of  pcre_,
42           and  the  names  in  the  32-bit  library start with pcre32_ instead of
43           pcre_. To avoid over-complication and reduce the documentation  mainte-
44           nance load, most of the documentation describes the 8-bit library, with
45           the differences for the 16-bit and  32-bit  libraries  described  sepa-
46           rately  in  the  pcre16  and  pcre32  pages. References to functions or
47           structures of the  form  pcre[16|32]_xxx  should  be  read  as  meaning
48           "pcre_xxx  when  using  the  8-bit  library,  pcre16_xxx when using the
49           16-bit library, or pcre32_xxx when using the 32-bit library".
50    
51           The current implementation of PCRE corresponds approximately with  Perl
52           5.12,  including  support  for  UTF-8/16/32 encoded strings and Unicode
53           general category properties. However, UTF-8/16/32 and  Unicode  support
54           has to be explicitly enabled; it is not the default. The Unicode tables
55           correspond to Unicode release 6.2.0.
56    
57           In addition to the Perl-compatible matching function, PCRE contains  an
58           alternative  function that matches the same compiled patterns in a dif-
59           ferent way. In certain circumstances, the alternative function has some
60           advantages.   For  a discussion of the two matching algorithms, see the
61           pcrematching page.
62    
63           PCRE is written in C and released as a C library. A  number  of  people
64           have  written  wrappers and interfaces of various kinds. In particular,
65           Google Inc.  have provided a comprehensive C++ wrapper  for  the  8-bit
66           library.  This  is  now  included as part of the PCRE distribution. The
67           pcrecpp page has details of this interface.  Other  people's  contribu-
68           tions  can  be  found in the Contrib directory at the primary FTP site,
69           which is:
70    
71           ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre
72    
73           Details of exactly which Perl regular expression features are  and  are
74           not supported by PCRE are given in separate documents. See the pcrepat-
75           tern and pcrecompat pages. There is a syntax summary in the  pcresyntax
76           page.
77    
78           Some  features  of  PCRE can be included, excluded, or changed when the
79           library is built. The pcre_config() function makes it  possible  for  a
80           client  to  discover  which  features are available. The features them-
81           selves are described in the pcrebuild page. Documentation about  build-
82           ing  PCRE  for various operating systems can be found in the README and
83           NON-AUTOTOOLS_BUILD files in the source distribution.
84    
85           The libraries contains a number of undocumented internal functions  and
86           data  tables  that  are  used by more than one of the exported external
87           functions, but which are not intended  for  use  by  external  callers.
88           Their  names all begin with "_pcre_" or "_pcre16_" or "_pcre32_", which
89           hopefully will not provoke any name clashes. In some  environments,  it
90           is  possible  to  control  which  external  symbols are exported when a
91           shared library is built, and in these cases  the  undocumented  symbols
92           are not exported.
93    
94    
95    SECURITY CONSIDERATIONS
96    
97           If  you  are  using PCRE in a non-UTF application that permits users to
98           supply arbitrary patterns for compilation, you should  be  aware  of  a
99           feature that allows users to turn on UTF support from within a pattern,
100           provided that PCRE was built with UTF support. For  example,  an  8-bit
101           pattern  that  begins  with  "(*UTF8)" turns on UTF-8 mode. This causes
102           both the pattern and any data against which it is matched to be checked
103           for UTF-8 validity. If the data string is very long, such a check might
104           use sufficiently many resources as to cause your  application  to  lose
105           performance.
106    
107           The  best  way  of  guarding  against  this  possibility  is to use the
108           pcre_fullinfo() function to check the compiled  pattern's  options  for
109           UTF.
110    
111           If  your  application  is one that supports UTF, be aware that validity
112           checking can take time. If the same data string is to be  matched  many
113           times, you can use the PCRE_NO_UTF[8|16|32]_CHECK option for the second
114           and subsequent matches to save redundant checks.
115    
116           Another way that performance can be hit is by running  a  pattern  that
117           has  a  very  large search tree against a string that will never match.
118           Nested unlimited repeats in a pattern are a common example.  PCRE  pro-
119           vides some protection against this: see the PCRE_EXTRA_MATCH_LIMIT fea-
120           ture in the pcreapi page.
121    
122    
123    USER DOCUMENTATION
124    
125           The user documentation for PCRE comprises a number  of  different  sec-
126           tions.  In the "man" format, each of these is a separate "man page". In
127           the HTML format, each is a separate page, linked from the  index  page.
128           In  the  plain  text format, all the sections, except the pcredemo sec-
129           tion, are concatenated, for ease of searching. The sections are as fol-
130           lows:
131    
132             pcre              this document
133             pcre16            details of the 16-bit library
134             pcre32            details of the 32-bit library
135             pcre-config       show PCRE installation configuration information
136             pcreapi           details of PCRE's native C API
137             pcrebuild         options for building PCRE
138             pcrecallout       details of the callout feature
139             pcrecompat        discussion of Perl compatibility
140             pcrecpp           details of the C++ wrapper for the 8-bit library
141             pcredemo          a demonstration C program that uses PCRE
142             pcregrep          description of the pcregrep command (8-bit only)
143             pcrejit           discussion of the just-in-time optimization support
144             pcrelimits        details of size and other limits
145             pcrematching      discussion of the two matching algorithms
146             pcrepartial       details of the partial matching facility
147             pcrepattern       syntax and semantics of supported
148                                 regular expressions
149             pcreperform       discussion of performance issues
150             pcreposix         the POSIX-compatible C API for the 8-bit library
151             pcreprecompile    details of saving and re-using precompiled patterns
152             pcresample        discussion of the pcredemo program
153             pcrestack         discussion of stack usage
154             pcresyntax        quick syntax reference
155             pcretest          description of the pcretest testing command
156             pcreunicode       discussion of Unicode and UTF-8/16/32 support
157    
158           In  addition,  in the "man" and HTML formats, there is a short page for
159           each C library function, listing its arguments and results.
160    
161    
162    AUTHOR
163    
164           Philip Hazel
165           University Computing Service
166           Cambridge CB2 3QH, England.
167    
168           Putting an actual email address here seems to have been a spam  magnet,
169           so  I've  taken  it away. If you want to email me, use my two initials,
170           followed by the two digits 10, at the domain cam.ac.uk.
171    
172    
173    REVISION
174    
175           Last updated: 30 October 2012
176           Copyright (c) 1997-2012 University of Cambridge.
177    ------------------------------------------------------------------------------
178    
179    
180    PCRE(3)                                                                PCRE(3)
181    
182    
183    NAME
184           PCRE - Perl-compatible regular expressions
185    
186           #include <pcre.h>
187    
188    
189    PCRE 16-BIT API BASIC FUNCTIONS
190    
191           pcre16 *pcre16_compile(PCRE_SPTR16 pattern, int options,
192                const char **errptr, int *erroffset,
193                const unsigned char *tableptr);
194    
195           pcre16 *pcre16_compile2(PCRE_SPTR16 pattern, int options,
196                int *errorcodeptr,
197                const char **errptr, int *erroffset,
198                const unsigned char *tableptr);
199    
200           pcre16_extra *pcre16_study(const pcre16 *code, int options,
201                const char **errptr);
202    
203           void pcre16_free_study(pcre16_extra *extra);
204    
205           int pcre16_exec(const pcre16 *code, const pcre16_extra *extra,
206                PCRE_SPTR16 subject, int length, int startoffset,
207                int options, int *ovector, int ovecsize);
208    
209           int pcre16_dfa_exec(const pcre16 *code, const pcre16_extra *extra,
210                PCRE_SPTR16 subject, int length, int startoffset,
211                int options, int *ovector, int ovecsize,
212                int *workspace, int wscount);
213    
214    
215    PCRE 16-BIT API STRING EXTRACTION FUNCTIONS
216    
217           int pcre16_copy_named_substring(const pcre16 *code,
218                PCRE_SPTR16 subject, int *ovector,
219                int stringcount, PCRE_SPTR16 stringname,
220                PCRE_UCHAR16 *buffer, int buffersize);
221    
222           int pcre16_copy_substring(PCRE_SPTR16 subject, int *ovector,
223                int stringcount, int stringnumber, PCRE_UCHAR16 *buffer,
224                int buffersize);
225    
226           int pcre16_get_named_substring(const pcre16 *code,
227                PCRE_SPTR16 subject, int *ovector,
228                int stringcount, PCRE_SPTR16 stringname,
229                PCRE_SPTR16 *stringptr);
230    
231           int pcre16_get_stringnumber(const pcre16 *code,
232                PCRE_SPTR16 name);
233    
234           int pcre16_get_stringtable_entries(const pcre16 *code,
235                PCRE_SPTR16 name, PCRE_UCHAR16 **first, PCRE_UCHAR16 **last);
236    
237           int pcre16_get_substring(PCRE_SPTR16 subject, int *ovector,
238                int stringcount, int stringnumber,
239                PCRE_SPTR16 *stringptr);
240    
241           int pcre16_get_substring_list(PCRE_SPTR16 subject,
242                int *ovector, int stringcount, PCRE_SPTR16 **listptr);
243    
244           void pcre16_free_substring(PCRE_SPTR16 stringptr);
245    
246           void pcre16_free_substring_list(PCRE_SPTR16 *stringptr);
247    
248    
249    PCRE 16-BIT API AUXILIARY FUNCTIONS
250    
251           pcre16_jit_stack *pcre16_jit_stack_alloc(int startsize, int maxsize);
252    
253           void pcre16_jit_stack_free(pcre16_jit_stack *stack);
254    
255           void pcre16_assign_jit_stack(pcre16_extra *extra,
256                pcre16_jit_callback callback, void *data);
257    
258           const unsigned char *pcre16_maketables(void);
259    
260           int pcre16_fullinfo(const pcre16 *code, const pcre16_extra *extra,
261                int what, void *where);
262    
263           int pcre16_refcount(pcre16 *code, int adjust);
264    
265           int pcre16_config(int what, void *where);
266    
267           const char *pcre16_version(void);
268    
269           int pcre16_pattern_to_host_byte_order(pcre16 *code,
270                pcre16_extra *extra, const unsigned char *tables);
271    
272    
273    PCRE 16-BIT API INDIRECTED FUNCTIONS
274    
275           void *(*pcre16_malloc)(size_t);
276    
277           void (*pcre16_free)(void *);
278    
279           void *(*pcre16_stack_malloc)(size_t);
280    
281           void (*pcre16_stack_free)(void *);
282    
283           int (*pcre16_callout)(pcre16_callout_block *);
284    
285    
286    PCRE 16-BIT API 16-BIT-ONLY FUNCTION
287    
288           int pcre16_utf16_to_host_byte_order(PCRE_UCHAR16 *output,
289                PCRE_SPTR16 input, int length, int *byte_order,
290                int keep_boms);
291    
292    
293    THE PCRE 16-BIT LIBRARY
294    
295           Starting  with  release  8.30, it is possible to compile a PCRE library
296           that supports 16-bit character strings, including  UTF-16  strings,  as
297           well  as  or instead of the original 8-bit library. The majority of the
298           work to make  this  possible  was  done  by  Zoltan  Herczeg.  The  two
299           libraries contain identical sets of functions, used in exactly the same
300           way. Only the names of the functions and the data types of their  argu-
301           ments  and results are different. To avoid over-complication and reduce
302           the documentation maintenance load,  most  of  the  PCRE  documentation
303           describes  the  8-bit  library,  with only occasional references to the
304           16-bit library. This page describes what is different when you use  the
305           16-bit library.
306    
307           WARNING:  A  single  application can be linked with both libraries, but
308           you must take care when processing any particular pattern to use  func-
309           tions  from  just one library. For example, if you want to study a pat-
310           tern that was compiled with  pcre16_compile(),  you  must  do  so  with
311           pcre16_study(), not pcre_study(), and you must free the study data with
312           pcre16_free_study().
313    
314    
315    THE HEADER FILE
316    
317           There is only one header file, pcre.h. It contains prototypes  for  all
318           the  functions  in  both  libraries,  as  well as definitions of flags,
319           structures, error codes, etc.
320    
321    
322    THE LIBRARY NAME
323    
324           In Unix-like systems, the 16-bit library is called libpcre16,  and  can
325           normally  be  accesss  by adding -lpcre16 to the command for linking an
326           application that uses PCRE.
327    
328    
329    STRING TYPES
330    
331           In the 8-bit library, strings are passed to PCRE library  functions  as
332           vectors  of  bytes  with  the  C  type "char *". In the 16-bit library,
333           strings are passed as vectors of unsigned 16-bit quantities. The  macro
334           PCRE_UCHAR16  specifies  an  appropriate  data type, and PCRE_SPTR16 is
335           defined as "const PCRE_UCHAR16 *". In very  many  environments,  "short
336           int" is a 16-bit data type. When PCRE is built, it defines PCRE_UCHAR16
337           as "short int", but checks that it really is a 16-bit data type. If  it
338           is not, the build fails with an error message telling the maintainer to
339           modify the definition appropriately.
340    
341    
342    STRUCTURE TYPES
343    
344           The types of the opaque structures that are used  for  compiled  16-bit
345           patterns  and  JIT stacks are pcre16 and pcre16_jit_stack respectively.
346           The  type  of  the  user-accessible  structure  that  is  returned   by
347           pcre16_study()  is  pcre16_extra, and the type of the structure that is
348           used for passing data to a callout  function  is  pcre16_callout_block.
349           These structures contain the same fields, with the same names, as their
350           8-bit counterparts. The only difference is that pointers  to  character
351           strings are 16-bit instead of 8-bit types.
352    
353    
354    16-BIT FUNCTIONS
355    
356           For  every function in the 8-bit library there is a corresponding func-
357           tion in the 16-bit library with a name that starts with pcre16_ instead
358           of  pcre_.  The  prototypes are listed above. In addition, there is one
359           extra function, pcre16_utf16_to_host_byte_order(). This  is  a  utility
360           function  that converts a UTF-16 character string to host byte order if
361           necessary. The other 16-bit  functions  expect  the  strings  they  are
362           passed to be in host byte order.
363    
364           The input and output arguments of pcre16_utf16_to_host_byte_order() may
365           point to the same address, that is, conversion in place  is  supported.
366           The output buffer must be at least as long as the input.
367    
368           The  length  argument  specifies the number of 16-bit data units in the
369           input string; a negative value specifies a zero-terminated string.
370    
371           If byte_order is NULL, it is assumed that the string starts off in host
372           byte  order. This may be changed by byte-order marks (BOMs) anywhere in
373           the string (commonly as the first character).
374    
375           If byte_order is not NULL, a non-zero value of the integer to which  it
376           points  means  that  the input starts off in host byte order, otherwise
377           the opposite order is assumed. Again, BOMs in  the  string  can  change
378           this. The final byte order is passed back at the end of processing.
379    
380           If  keep_boms  is  not  zero,  byte-order  mark characters (0xfeff) are
381           copied into the output string. Otherwise they are discarded.
382    
383           The result of the function is the number of 16-bit  units  placed  into
384           the  output  buffer,  including  the  zero terminator if the string was
385           zero-terminated.
386    
387    
388    SUBJECT STRING OFFSETS
389    
390           The offsets within subject strings that are returned  by  the  matching
391           functions are in 16-bit units rather than bytes.
392    
393    
394    NAMED SUBPATTERNS
395    
396           The  name-to-number translation table that is maintained for named sub-
397           patterns uses 16-bit characters.  The  pcre16_get_stringtable_entries()
398           function returns the length of each entry in the table as the number of
399           16-bit data units.
400    
401    
402    OPTION NAMES
403    
404           There   are   two   new   general   option   names,   PCRE_UTF16    and
405           PCRE_NO_UTF16_CHECK,     which     correspond    to    PCRE_UTF8    and
406           PCRE_NO_UTF8_CHECK in the 8-bit library. In  fact,  these  new  options
407           define  the  same bits in the options word. There is a discussion about
408           the validity of UTF-16 strings in the pcreunicode page.
409    
410           For the pcre16_config() function there is an  option  PCRE_CONFIG_UTF16
411           that  returns  1  if UTF-16 support is configured, otherwise 0. If this
412           option  is  given  to  pcre_config()  or  pcre32_config(),  or  if  the
413           PCRE_CONFIG_UTF8  or  PCRE_CONFIG_UTF32  option is given to pcre16_con-
414           fig(), the result is the PCRE_ERROR_BADOPTION error.
415    
416    
417    CHARACTER CODES
418    
419           In 16-bit mode, when  PCRE_UTF16  is  not  set,  character  values  are
420           treated in the same way as in 8-bit, non UTF-8 mode, except, of course,
421           that they can range from 0 to 0xffff instead of 0  to  0xff.  Character
422           types  for characters less than 0xff can therefore be influenced by the
423           locale in the same way as before.  Characters greater  than  0xff  have
424           only one case, and no "type" (such as letter or digit).
425    
426           In  UTF-16  mode,  the  character  code  is  Unicode, in the range 0 to
427           0x10ffff, with the exception of values in the range  0xd800  to  0xdfff
428           because  those  are "surrogate" values that are used in pairs to encode
429           values greater than 0xffff.
430    
431           A UTF-16 string can indicate its endianness by special code knows as  a
432           byte-order mark (BOM). The PCRE functions do not handle this, expecting
433           strings  to  be  in  host  byte  order.  A  utility   function   called
434           pcre16_utf16_to_host_byte_order()  is  provided  to help with this (see
435           above).
436    
437    
438    ERROR NAMES
439    
440           The errors PCRE_ERROR_BADUTF16_OFFSET and PCRE_ERROR_SHORTUTF16  corre-
441           spond  to  their  8-bit  counterparts.  The error PCRE_ERROR_BADMODE is
442           given when a compiled pattern is passed to a  function  that  processes
443           patterns  in  the  other  mode, for example, if a pattern compiled with
444           pcre_compile() is passed to pcre16_exec().
445    
446           There are new error codes whose names  begin  with  PCRE_UTF16_ERR  for
447           invalid  UTF-16  strings,  corresponding to the PCRE_UTF8_ERR codes for
448           UTF-8 strings that are described in the section entitled "Reason  codes
449           for  invalid UTF-8 strings" in the main pcreapi page. The UTF-16 errors
450           are:
451    
452             PCRE_UTF16_ERR1  Missing low surrogate at end of string
453             PCRE_UTF16_ERR2  Invalid low surrogate follows high surrogate
454             PCRE_UTF16_ERR3  Isolated low surrogate
455             PCRE_UTF16_ERR4  Non-character
456    
457    
458    ERROR TEXTS
459    
460           If there is an error while compiling a pattern, the error text that  is
461           passed  back by pcre16_compile() or pcre16_compile2() is still an 8-bit
462           character string, zero-terminated.
463    
464    
465    CALLOUTS
466    
467           The subject and mark fields in the callout block that is  passed  to  a
468           callout function point to 16-bit vectors.
469    
470    
471    TESTING
472    
473           The  pcretest  program continues to operate with 8-bit input and output
474           files, but it can be used for testing the 16-bit library. If it is  run
475           with the command line option -16, patterns and subject strings are con-
476           verted from 8-bit to 16-bit before being passed to PCRE, and the 16-bit
477           library  functions  are used instead of the 8-bit ones. Returned 16-bit
478           strings are converted to 8-bit for output. If both the  8-bit  and  the
479           32-bit libraries were not compiled, pcretest defaults to 16-bit and the
480           -16 option is ignored.
481    
482           When PCRE is being built, the RunTest script that is  called  by  "make
483           check"  uses  the  pcretest  -C  option to discover which of the 8-bit,
484           16-bit and 32-bit libraries has been built, and runs the  tests  appro-
485           priately.
486    
487    
488    NOT SUPPORTED IN 16-BIT MODE
489    
490           Not all the features of the 8-bit library are available with the 16-bit
491           library. The C++ and POSIX wrapper functions  support  only  the  8-bit
492           library, and the pcregrep program is at present 8-bit only.
493    
494    
495    AUTHOR
496    
497           Philip Hazel
498           University Computing Service
499           Cambridge CB2 3QH, England.
500    
501    
502    REVISION
503    
504           Last updated: 14 April 2012
505           Copyright (c) 1997-2012 University of Cambridge.
506    ------------------------------------------------------------------------------
507    
508    
509    PCRE(3)                                                                PCRE(3)
510    
511    
512    NAME
513           PCRE - Perl-compatible regular expressions
514    
515           #include <pcre.h>
516    
517    
518    PCRE 32-BIT API BASIC FUNCTIONS
519    
520           pcre32 *pcre32_compile(PCRE_SPTR32 pattern, int options,
521                const char **errptr, int *erroffset,
522                const unsigned char *tableptr);
523    
524           pcre32 *pcre32_compile2(PCRE_SPTR32 pattern, int options,
525                int *errorcodeptr,
526                const char **errptr, int *erroffset,
527                const unsigned char *tableptr);
528    
529           pcre32_extra *pcre32_study(const pcre32 *code, int options,
530                const char **errptr);
531    
532           void pcre32_free_study(pcre32_extra *extra);
533    
534           int pcre32_exec(const pcre32 *code, const pcre32_extra *extra,
535                PCRE_SPTR32 subject, int length, int startoffset,
536                int options, int *ovector, int ovecsize);
537    
538           int pcre32_dfa_exec(const pcre32 *code, const pcre32_extra *extra,
539                PCRE_SPTR32 subject, int length, int startoffset,
540                int options, int *ovector, int ovecsize,
541                int *workspace, int wscount);
542    
543    
544    PCRE 32-BIT API STRING EXTRACTION FUNCTIONS
545    
546           int pcre32_copy_named_substring(const pcre32 *code,
547                PCRE_SPTR32 subject, int *ovector,
548                int stringcount, PCRE_SPTR32 stringname,
549                PCRE_UCHAR32 *buffer, int buffersize);
550    
551           int pcre32_copy_substring(PCRE_SPTR32 subject, int *ovector,
552                int stringcount, int stringnumber, PCRE_UCHAR32 *buffer,
553                int buffersize);
554    
555           int pcre32_get_named_substring(const pcre32 *code,
556                PCRE_SPTR32 subject, int *ovector,
557                int stringcount, PCRE_SPTR32 stringname,
558                PCRE_SPTR32 *stringptr);
559    
560           int pcre32_get_stringnumber(const pcre32 *code,
561                PCRE_SPTR32 name);
562    
563           int pcre32_get_stringtable_entries(const pcre32 *code,
564                PCRE_SPTR32 name, PCRE_UCHAR32 **first, PCRE_UCHAR32 **last);
565    
566           int pcre32_get_substring(PCRE_SPTR32 subject, int *ovector,
567                int stringcount, int stringnumber,
568                PCRE_SPTR32 *stringptr);
569    
570           int pcre32_get_substring_list(PCRE_SPTR32 subject,
571                int *ovector, int stringcount, PCRE_SPTR32 **listptr);
572    
573           void pcre32_free_substring(PCRE_SPTR32 stringptr);
574    
575           void pcre32_free_substring_list(PCRE_SPTR32 *stringptr);
576    
577    
578    PCRE 32-BIT API AUXILIARY FUNCTIONS
579    
580           pcre32_jit_stack *pcre32_jit_stack_alloc(int startsize, int maxsize);
581    
582           void pcre32_jit_stack_free(pcre32_jit_stack *stack);
583    
584           void pcre32_assign_jit_stack(pcre32_extra *extra,
585                pcre32_jit_callback callback, void *data);
586    
587           const unsigned char *pcre32_maketables(void);
588    
589           int pcre32_fullinfo(const pcre32 *code, const pcre32_extra *extra,
590                int what, void *where);
591    
592           int pcre32_refcount(pcre32 *code, int adjust);
593    
594           int pcre32_config(int what, void *where);
595    
596           const char *pcre32_version(void);
597    
598           int pcre32_pattern_to_host_byte_order(pcre32 *code,
599                pcre32_extra *extra, const unsigned char *tables);
600    
601    
602    PCRE 32-BIT API INDIRECTED FUNCTIONS
603    
604           void *(*pcre32_malloc)(size_t);
605    
606           void (*pcre32_free)(void *);
607    
608           void *(*pcre32_stack_malloc)(size_t);
609    
610           void (*pcre32_stack_free)(void *);
611    
612           int (*pcre32_callout)(pcre32_callout_block *);
613    
614    
615    PCRE 32-BIT API 32-BIT-ONLY FUNCTION
616    
617           int pcre32_utf32_to_host_byte_order(PCRE_UCHAR32 *output,
618                PCRE_SPTR32 input, int length, int *byte_order,
619                int keep_boms);
620    
621    
622    THE PCRE 32-BIT LIBRARY
623    
624           Starting  with  release  8.32, it is possible to compile a PCRE library
625           that supports 32-bit character strings, including  UTF-32  strings,  as
626           well as or instead of the original 8-bit library. This work was done by
627           Christian Persch, based on the work done  by  Zoltan  Herczeg  for  the
628           16-bit  library.  All  three  libraries contain identical sets of func-
629           tions, used in exactly the same way.  Only the names of  the  functions
630           and  the  data  types  of their arguments and results are different. To
631           avoid over-complication and reduce the documentation maintenance  load,
632           most  of  the PCRE documentation describes the 8-bit library, with only
633           occasional references to the 16-bit and  32-bit  libraries.  This  page
634           describes what is different when you use the 32-bit library.
635    
636           WARNING:  A  single  application  can  be linked with all or any of the
637           three libraries, but you must take care when processing any  particular
638           pattern  to  use  functions  from just one library. For example, if you
639           want to study a pattern that was compiled  with  pcre32_compile(),  you
640           must do so with pcre32_study(), not pcre_study(), and you must free the
641           study data with pcre32_free_study().
642    
643    
644    THE HEADER FILE
645    
646           There is only one header file, pcre.h. It contains prototypes  for  all
647           the  functions  in  both  libraries,  as  well as definitions of flags,
648           structures, error codes, etc.
649    
650    
651    THE LIBRARY NAME
652    
653           In Unix-like systems, the 32-bit library is called libpcre32,  and  can
654           normally  be  accesss  by adding -lpcre32 to the command for linking an
655           application that uses PCRE.
656    
657    
658    STRING TYPES
659    
660           In the 8-bit library, strings are passed to PCRE library  functions  as
661           vectors  of  bytes  with  the  C  type "char *". In the 32-bit library,
662           strings are passed as vectors of unsigned 32-bit quantities. The  macro
663           PCRE_UCHAR32  specifies  an  appropriate  data type, and PCRE_SPTR32 is
664           defined as "const PCRE_UCHAR32 *". In very many environments, "unsigned
665           int" is a 32-bit data type. When PCRE is built, it defines PCRE_UCHAR32
666           as "unsigned int", but checks that it really is a 32-bit data type.  If
667           it is not, the build fails with an error message telling the maintainer
668           to modify the definition appropriately.
669    
670    
671    STRUCTURE TYPES
672    
673           The types of the opaque structures that are used  for  compiled  32-bit
674           patterns  and  JIT stacks are pcre32 and pcre32_jit_stack respectively.
675           The  type  of  the  user-accessible  structure  that  is  returned   by
676           pcre32_study()  is  pcre32_extra, and the type of the structure that is
677           used for passing data to a callout  function  is  pcre32_callout_block.
678           These structures contain the same fields, with the same names, as their
679           8-bit counterparts. The only difference is that pointers  to  character
680           strings are 32-bit instead of 8-bit types.
681    
682    
683    32-BIT FUNCTIONS
684    
685           For  every function in the 8-bit library there is a corresponding func-
686           tion in the 32-bit library with a name that starts with pcre32_ instead
687           of  pcre_.  The  prototypes are listed above. In addition, there is one
688           extra function, pcre32_utf32_to_host_byte_order(). This  is  a  utility
689           function  that converts a UTF-32 character string to host byte order if
690           necessary. The other 32-bit  functions  expect  the  strings  they  are
691           passed to be in host byte order.
692    
693           The input and output arguments of pcre32_utf32_to_host_byte_order() may
694           point to the same address, that is, conversion in place  is  supported.
695           The output buffer must be at least as long as the input.
696    
697           The  length  argument  specifies the number of 32-bit data units in the
698           input string; a negative value specifies a zero-terminated string.
699    
700           If byte_order is NULL, it is assumed that the string starts off in host
701           byte  order. This may be changed by byte-order marks (BOMs) anywhere in
702           the string (commonly as the first character).
703    
704           If byte_order is not NULL, a non-zero value of the integer to which  it
705           points  means  that  the input starts off in host byte order, otherwise
706           the opposite order is assumed. Again, BOMs in  the  string  can  change
707           this. The final byte order is passed back at the end of processing.
708    
709           If  keep_boms  is  not  zero,  byte-order  mark characters (0xfeff) are
710           copied into the output string. Otherwise they are discarded.
711    
712           The result of the function is the number of 32-bit  units  placed  into
713           the  output  buffer,  including  the  zero terminator if the string was
714           zero-terminated.
715    
716    
717    SUBJECT STRING OFFSETS
718    
719           The offsets within subject strings that are returned  by  the  matching
720           functions are in 32-bit units rather than bytes.
721    
722    
723    NAMED SUBPATTERNS
724    
725           The  name-to-number translation table that is maintained for named sub-
726           patterns uses 32-bit characters.  The  pcre32_get_stringtable_entries()
727           function returns the length of each entry in the table as the number of
728           32-bit data units.
729    
730    
731    OPTION NAMES
732    
733           There   are   two   new   general   option   names,   PCRE_UTF32    and
734           PCRE_NO_UTF32_CHECK,     which     correspond    to    PCRE_UTF8    and
735           PCRE_NO_UTF8_CHECK in the 8-bit library. In  fact,  these  new  options
736           define  the  same bits in the options word. There is a discussion about
737           the validity of UTF-32 strings in the pcreunicode page.
738    
739           For the pcre32_config() function there is an  option  PCRE_CONFIG_UTF32
740           that  returns  1  if UTF-32 support is configured, otherwise 0. If this
741           option  is  given  to  pcre_config()  or  pcre16_config(),  or  if  the
742           PCRE_CONFIG_UTF8  or  PCRE_CONFIG_UTF16  option is given to pcre32_con-
743           fig(), the result is the PCRE_ERROR_BADOPTION error.
744    
745    
746    CHARACTER CODES
747    
748           In 32-bit mode, when  PCRE_UTF32  is  not  set,  character  values  are
749           treated in the same way as in 8-bit, non UTF-8 mode, except, of course,
750           that they can range from 0 to 0x7fffffff instead of 0 to 0xff.  Charac-
751           ter  types for characters less than 0xff can therefore be influenced by
752           the locale in the same way as before.   Characters  greater  than  0xff
753           have only one case, and no "type" (such as letter or digit).
754    
755           In  UTF-32  mode,  the  character  code  is  Unicode, in the range 0 to
756           0x10ffff, with the exception of values in the range  0xd800  to  0xdfff
757           because those are "surrogate" values that are ill-formed in UTF-32.
758    
759           A  UTF-32 string can indicate its endianness by special code knows as a
760           byte-order mark (BOM). The PCRE functions do not handle this, expecting
761           strings   to   be  in  host  byte  order.  A  utility  function  called
762           pcre32_utf32_to_host_byte_order() is provided to help  with  this  (see
763           above).
764    
765    
766    ERROR NAMES
767    
768           The  error  PCRE_ERROR_BADUTF32  corresponds  to its 8-bit counterpart.
769           The error PCRE_ERROR_BADMODE is given when a compiled pattern is passed
770           to  a  function that processes patterns in the other mode, for example,
771           if a pattern compiled with pcre_compile() is passed to pcre32_exec().
772    
773           There are new error codes whose names  begin  with  PCRE_UTF32_ERR  for
774           invalid  UTF-32  strings,  corresponding to the PCRE_UTF8_ERR codes for
775           UTF-8 strings that are described in the section entitled "Reason  codes
776           for  invalid UTF-8 strings" in the main pcreapi page. The UTF-32 errors
777           are:
778    
779             PCRE_UTF32_ERR1  Surrogate character (range from 0xd800 to 0xdfff)
780             PCRE_UTF32_ERR2  Non-character
781             PCRE_UTF32_ERR3  Character > 0x10ffff
782    
783    
784    ERROR TEXTS
785    
786           If there is an error while compiling a pattern, the error text that  is
787           passed  back by pcre32_compile() or pcre32_compile2() is still an 8-bit
788           character string, zero-terminated.
789    
790    
791    CALLOUTS
792    
793           The subject and mark fields in the callout block that is  passed  to  a
794           callout function point to 32-bit vectors.
795    
796    
797    TESTING
798    
799           The  pcretest  program continues to operate with 8-bit input and output
800           files, but it can be used for testing the 32-bit library. If it is  run
801           with the command line option -32, patterns and subject strings are con-
802           verted from 8-bit to 32-bit before being passed to PCRE, and the 32-bit
803           library  functions  are used instead of the 8-bit ones. Returned 32-bit
804           strings are converted to 8-bit for output. If both the  8-bit  and  the
805           16-bit libraries were not compiled, pcretest defaults to 32-bit and the
806           -32 option is ignored.
807    
808           When PCRE is being built, the RunTest script that is  called  by  "make
809           check"  uses  the  pcretest  -C  option to discover which of the 8-bit,
810           16-bit and 32-bit libraries has been built, and runs the  tests  appro-
811           priately.
812    
813    
814    NOT SUPPORTED IN 32-BIT MODE
815    
816           Not all the features of the 8-bit library are available with the 32-bit
817           library. The C++ and POSIX wrapper functions  support  only  the  8-bit
818           library, and the pcregrep program is at present 8-bit only.
819    
820    
821    AUTHOR
822    
823           Philip Hazel
824           University Computing Service
825           Cambridge CB2 3QH, England.
826    
827    
828    REVISION
829    
830           Last updated: 24 June 2012
831           Copyright (c) 1997-2012 University of Cambridge.
832    ------------------------------------------------------------------------------
833    
834    
835    PCREBUILD(3)                                                      PCREBUILD(3)
836    
837    
838    NAME
839           PCRE - Perl-compatible regular expressions
840    
841    
842    PCRE BUILD-TIME OPTIONS
843    
844           This  document  describes  the  optional  features  of PCRE that can be
845           selected when the library is compiled. It assumes use of the  configure
846           script,  where the optional features are selected or deselected by pro-
847           viding options to configure before running the make  command.  However,
848           the  same  options  can be selected in both Unix-like and non-Unix-like
849           environments using the GUI facility of cmake-gui if you are using CMake
850           instead of configure to build PCRE.
851    
852           There  is a lot more information about building PCRE without using con-
853           figure (including information about using CMake or building "by  hand")
854           in  the file called NON-AUTOTOOLS-BUILD, which is part of the PCRE dis-
855           tribution. You should consult this file as well as the README  file  if
856           you are building in a non-Unix-like environment.
857    
858           The complete list of options for configure (which includes the standard
859           ones such as the  selection  of  the  installation  directory)  can  be
860           obtained by running
861    
862             ./configure --help
863    
864           The  following  sections  include  descriptions  of options whose names
865           begin with --enable or --disable. These settings specify changes to the
866           defaults  for  the configure command. Because of the way that configure
867           works, --enable and --disable always come in pairs, so  the  complemen-
868           tary  option always exists as well, but as it specifies the default, it
869           is not described.
870    
871    
872    BUILDING 8-BIT, 16-BIT AND 32-BIT LIBRARIES
873    
874           By default, a library called libpcre  is  built,  containing  functions
875           that  take  string  arguments  contained in vectors of bytes, either as
876           single-byte characters, or interpreted as UTF-8 strings. You  can  also
877           build  a  separate library, called libpcre16, in which strings are con-
878           tained in vectors of 16-bit data units and interpreted either  as  sin-
879           gle-unit characters or UTF-16 strings, by adding
880    
881             --enable-pcre16
882    
883           to the configure command. You can also build a separate library, called
884           libpcre32, in which strings are contained in  vectors  of  32-bit  data
885           units  and  interpreted  either  as  single-unit  characters  or UTF-32
886           strings, by adding
887    
888             --enable-pcre32
889    
890           to the configure command. If you do not want the 8-bit library, add
891    
892             --disable-pcre8
893    
894           as well. At least one of the three libraries must be built.  Note  that
895           the  C++  and  POSIX  wrappers are for the 8-bit library only, and that
896           pcregrep is an 8-bit program. None of these are  built  if  you  select
897           only the 16-bit or 32-bit libraries.
898    
899    
900    BUILDING SHARED AND STATIC LIBRARIES
901    
902           The  PCRE building process uses libtool to build both shared and static
903           Unix libraries by default. You can suppress one of these by adding  one
904           of
905    
906             --disable-shared
907             --disable-static
908    
909           to the configure command, as required.
910    
911    
912    C++ SUPPORT
913    
914           By  default,  if the 8-bit library is being built, the configure script
915           will search for a C++ compiler and C++ header files. If it finds  them,
916           it  automatically  builds  the C++ wrapper library (which supports only
917           8-bit strings). You can disable this by adding
918    
919             --disable-cpp
920    
921           to the configure command.
922    
923    
924    UTF-8, UTF-16 AND UTF-32 SUPPORT
925    
926           To build PCRE with support for UTF Unicode character strings, add
927    
928             --enable-utf
929    
930           to the configure command. This setting applies to all three  libraries,
931           adding  support  for  UTF-8 to the 8-bit library, support for UTF-16 to
932           the 16-bit library, and  support  for  UTF-32  to  the  to  the  32-bit
933           library.  There  are no separate options for enabling UTF-8, UTF-16 and
934           UTF-32 independently because that would allow ridiculous settings  such
935           as  requesting UTF-16 support while building only the 8-bit library. It
936           is not possible to build one library with UTF support and another with-
937           out  in the same configuration. (For backwards compatibility, --enable-
938           utf8 is a synonym of --enable-utf.)
939    
940           Of itself, this setting does not make  PCRE  treat  strings  as  UTF-8,
941           UTF-16  or UTF-32. As well as compiling PCRE with this option, you also
942           have have to set the PCRE_UTF8, PCRE_UTF16  or  PCRE_UTF32  option  (as
943           appropriate) when you call one of the pattern compiling functions.
944    
945           If  you  set --enable-utf when compiling in an EBCDIC environment, PCRE
946           expects its input to be either ASCII or UTF-8 (depending  on  the  run-
947           time option). It is not possible to support both EBCDIC and UTF-8 codes
948           in the same version of  the  library.  Consequently,  --enable-utf  and
949           --enable-ebcdic are mutually exclusive.
950    
951    
952    UNICODE CHARACTER PROPERTY SUPPORT
953    
954           UTF  support allows the libraries to process character codepoints up to
955           0x10ffff in the strings that they handle. On its own, however, it  does
956           not provide any facilities for accessing the properties of such charac-
957           ters. If you want to be able to use the pattern escapes \P, \p, and \X,
958           which refer to Unicode character properties, you must add
959    
960             --enable-unicode-properties
961    
962           to  the  configure  command. This implies UTF support, even if you have
963           not explicitly requested it.
964    
965           Including Unicode property support adds around 30K  of  tables  to  the
966           PCRE  library.  Only  the general category properties such as Lu and Nd
967           are supported. Details are given in the pcrepattern documentation.
968    
969    
970    JUST-IN-TIME COMPILER SUPPORT
971    
972           Just-in-time compiler support is included in the build by specifying
973    
974             --enable-jit
975    
976           This support is available only for certain hardware  architectures.  If
977           this  option  is  set  for  an unsupported architecture, a compile time
978           error occurs.  See the pcrejit documentation for a  discussion  of  JIT
979           usage. When JIT support is enabled, pcregrep automatically makes use of
980           it, unless you add
981    
982             --disable-pcregrep-jit
983    
984           to the "configure" command.
985    
986    
987    CODE VALUE OF NEWLINE
988    
989           By default, PCRE interprets the linefeed (LF) character  as  indicating
990           the  end  of  a line. This is the normal newline character on Unix-like
991           systems. You can compile PCRE to use carriage return (CR)  instead,  by
992           adding
993    
994             --enable-newline-is-cr
995    
996           to  the  configure  command.  There  is  also  a --enable-newline-is-lf
997           option, which explicitly specifies linefeed as the newline character.
998    
999           Alternatively, you can specify that line endings are to be indicated by
1000           the two character sequence CRLF. If you want this, add
1001    
1002             --enable-newline-is-crlf
1003    
1004           to the configure command. There is a fourth option, specified by
1005    
1006             --enable-newline-is-anycrlf
1007    
1008           which  causes  PCRE  to recognize any of the three sequences CR, LF, or
1009           CRLF as indicating a line ending. Finally, a fifth option, specified by
1010    
1011             --enable-newline-is-any
1012    
1013           causes PCRE to recognize any Unicode newline sequence.
1014    
1015           Whatever line ending convention is selected when PCRE is built  can  be
1016           overridden  when  the library functions are called. At build time it is
1017           conventional to use the standard for your operating system.
1018    
1019    
1020    WHAT \R MATCHES
1021    
1022           By default, the sequence \R in a pattern matches  any  Unicode  newline
1023           sequence,  whatever  has  been selected as the line ending sequence. If
1024           you specify
1025    
1026             --enable-bsr-anycrlf
1027    
1028           the default is changed so that \R matches only CR, LF, or  CRLF.  What-
1029           ever  is selected when PCRE is built can be overridden when the library
1030           functions are called.
1031    
1032    
1033    POSIX MALLOC USAGE
1034    
1035           When the 8-bit library is called through the POSIX interface  (see  the
1036           pcreposix  documentation),  additional  working storage is required for
1037           holding the pointers to capturing  substrings,  because  PCRE  requires
1038           three integers per substring, whereas the POSIX interface provides only
1039           two. If the number of expected substrings is small, the  wrapper  func-
1040           tion  uses  space  on the stack, because this is faster than using mal-
1041           loc() for each call. The default threshold above which the stack is  no
1042           longer used is 10; it can be changed by adding a setting such as
1043    
1044             --with-posix-malloc-threshold=20
1045    
1046           to the configure command.
1047    
1048    
1049    HANDLING VERY LARGE PATTERNS
1050    
1051           Within  a  compiled  pattern,  offset values are used to point from one
1052           part to another (for example, from an opening parenthesis to an  alter-
1053           nation  metacharacter).  By default, in the 8-bit and 16-bit libraries,
1054           two-byte values are used for these offsets, leading to a  maximum  size
1055           for  a compiled pattern of around 64K. This is sufficient to handle all
1056           but the most gigantic patterns.  Nevertheless, some people do  want  to
1057           process  truly  enormous patterns, so it is possible to compile PCRE to
1058           use three-byte or four-byte offsets by adding a setting such as
1059    
1060             --with-link-size=3
1061    
1062           to the configure command. The value given must be 2, 3, or 4.  For  the
1063           16-bit  library,  a  value of 3 is rounded up to 4. In these libraries,
1064           using longer offsets slows down the operation of PCRE because it has to
1065           load  additional  data  when  handling them. For the 32-bit library the
1066           value is always 4 and cannot be overridden; the value  of  --with-link-
1067           size is ignored.
1068    
1069    
1070    AVOIDING EXCESSIVE STACK USAGE
1071    
1072           When matching with the pcre_exec() function, PCRE implements backtrack-
1073           ing by making recursive calls to an internal function  called  match().
1074           In  environments  where  the size of the stack is limited, this can se-
1075           verely limit PCRE's operation. (The Unix environment does  not  usually
1076           suffer from this problem, but it may sometimes be necessary to increase
1077           the maximum stack size.  There is a discussion in the  pcrestack  docu-
1078           mentation.)  An alternative approach to recursion that uses memory from
1079           the heap to remember data, instead of using recursive  function  calls,
1080           has  been  implemented to work round the problem of limited stack size.
1081           If you want to build a version of PCRE that works this way, add
1082    
1083             --disable-stack-for-recursion
1084    
1085           to the configure command. With this configuration, PCRE  will  use  the
1086           pcre_stack_malloc  and pcre_stack_free variables to call memory manage-
1087           ment functions. By default these point to malloc() and free(), but  you
1088           can replace the pointers so that your own functions are used instead.
1089    
1090           Separate  functions  are  provided  rather  than  using pcre_malloc and
1091           pcre_free because the  usage  is  very  predictable:  the  block  sizes
1092           requested  are  always  the  same,  and  the blocks are always freed in
1093           reverse order. A calling program might be able to  implement  optimized
1094           functions  that  perform  better  than  malloc()  and free(). PCRE runs
1095           noticeably more slowly when built in this way. This option affects only
1096           the pcre_exec() function; it is not relevant for pcre_dfa_exec().
1097    
1098    
1099    LIMITING PCRE RESOURCE USAGE
1100    
1101           Internally,  PCRE has a function called match(), which it calls repeat-
1102           edly  (sometimes  recursively)  when  matching  a  pattern   with   the
1103           pcre_exec()  function.  By controlling the maximum number of times this
1104           function may be called during a single matching operation, a limit  can
1105           be  placed  on  the resources used by a single call to pcre_exec(). The
1106           limit can be changed at run time, as described in the pcreapi  documen-
1107           tation.  The default is 10 million, but this can be changed by adding a
1108           setting such as
1109    
1110             --with-match-limit=500000
1111    
1112           to  the  configure  command.  This  setting  has  no  effect   on   the
1113           pcre_dfa_exec() matching function.
1114    
1115           In  some  environments  it is desirable to limit the depth of recursive
1116           calls of match() more strictly than the total number of calls, in order
1117           to  restrict  the maximum amount of stack (or heap, if --disable-stack-
1118           for-recursion is specified) that is used. A second limit controls this;
1119           it  defaults  to  the  value  that is set for --with-match-limit, which
1120           imposes no additional constraints. However, you can set a  lower  limit
1121           by adding, for example,
1122    
1123             --with-match-limit-recursion=10000
1124    
1125           to  the  configure  command.  This  value can also be overridden at run
1126           time.
1127    
1128    
1129    CREATING CHARACTER TABLES AT BUILD TIME
1130    
1131           PCRE uses fixed tables for processing characters whose code values  are
1132           less  than 256. By default, PCRE is built with a set of tables that are
1133           distributed in the file pcre_chartables.c.dist. These  tables  are  for
1134           ASCII codes only. If you add
1135    
1136             --enable-rebuild-chartables
1137    
1138           to  the  configure  command, the distributed tables are no longer used.
1139           Instead, a program called dftables is compiled and  run.  This  outputs
1140           the source for new set of tables, created in the default locale of your
1141           C run-time system. (This method of replacing the tables does  not  work
1142           if  you are cross compiling, because dftables is run on the local host.
1143           If you need to create alternative tables when cross compiling, you will
1144           have to do so "by hand".)
1145    
1146    
1147    USING EBCDIC CODE
1148    
1149           PCRE  assumes  by  default that it will run in an environment where the
1150           character code is ASCII (or Unicode, which is  a  superset  of  ASCII).
1151           This  is  the  case for most computer operating systems. PCRE can, how-
1152           ever, be compiled to run in an EBCDIC environment by adding
1153    
1154             --enable-ebcdic
1155    
1156           to the configure command. This setting implies --enable-rebuild-charta-
1157           bles.  You  should  only  use  it if you know that you are in an EBCDIC
1158           environment (for example,  an  IBM  mainframe  operating  system).  The
1159           --enable-ebcdic option is incompatible with --enable-utf.
1160    
1161           The EBCDIC character that corresponds to an ASCII LF is assumed to have
1162           the value 0x15 by default. However, in some EBCDIC  environments,  0x25
1163           is used. In such an environment you should use
1164    
1165             --enable-ebcdic-nl25
1166    
1167           as well as, or instead of, --enable-ebcdic. The EBCDIC character for CR
1168           has the same value as in ASCII, namely, 0x0d.  Whichever  of  0x15  and
1169           0x25 is not chosen as LF is made to correspond to the Unicode NEL char-
1170           acter (which, in Unicode, is 0x85).
1171    
1172           The options that select newline behaviour, such as --enable-newline-is-
1173           cr, and equivalent run-time options, refer to these character values in
1174           an EBCDIC environment.
1175    
1176    
1177    PCREGREP OPTIONS FOR COMPRESSED FILE SUPPORT
1178    
1179           By default, pcregrep reads all files as plain text. You can build it so
1180           that it recognizes files whose names end in .gz or .bz2, and reads them
1181           with libz or libbz2, respectively, by adding one or both of
1182    
1183             --enable-pcregrep-libz
1184             --enable-pcregrep-libbz2
1185    
1186           to the configure command. These options naturally require that the rel-
1187           evant  libraries  are installed on your system. Configuration will fail
1188           if they are not.
1189    
1190    
1191    PCREGREP BUFFER SIZE
1192    
1193           pcregrep uses an internal buffer to hold a "window" on the file  it  is
1194           scanning, in order to be able to output "before" and "after" lines when
1195           it finds a match. The size of the buffer is controlled by  a  parameter
1196           whose default value is 20K. The buffer itself is three times this size,
1197           but because of the way it is used for holding "before" lines, the long-
1198           est  line  that  is guaranteed to be processable is the parameter size.
1199           You can change the default parameter value by adding, for example,
1200    
1201             --with-pcregrep-bufsize=50K
1202    
1203           to the configure command. The caller of pcregrep can, however, override
1204           this value by specifying a run-time option.
1205    
1206    
1207    PCRETEST OPTION FOR LIBREADLINE SUPPORT
1208    
1209           If you add
1210    
1211             --enable-pcretest-libreadline
1212    
1213           to  the  configure  command,  pcretest  is  linked with the libreadline
1214           library, and when its input is from a terminal, it reads it  using  the
1215           readline() function. This provides line-editing and history facilities.
1216           Note that libreadline is GPL-licensed, so if you distribute a binary of
1217           pcretest linked in this way, there may be licensing issues.
1218    
1219           Setting  this  option  causes  the -lreadline option to be added to the
1220           pcretest build. In many operating environments with  a  sytem-installed
1221           libreadline this is sufficient. However, in some environments (e.g.  if
1222           an unmodified distribution version of readline is in use),  some  extra
1223           configuration  may  be necessary. The INSTALL file for libreadline says
1224           this:
1225    
1226             "Readline uses the termcap functions, but does not link with the
1227             termcap or curses library itself, allowing applications which link
1228             with readline the to choose an appropriate library."
1229    
1230           If your environment has not been set up so that an appropriate  library
1231           is automatically included, you may need to add something like
1232    
1233             LIBS="-ncurses"
1234    
1235           immediately before the configure command.
1236    
1237    
1238    DEBUGGING WITH VALGRIND SUPPORT
1239    
1240           By adding the
1241    
1242             --enable-valgrind
1243    
1244           option  to to the configure command, PCRE will use valgrind annotations
1245           to mark certain memory regions as  unaddressable.  This  allows  it  to
1246           detect invalid memory accesses, and is mostly useful for debugging PCRE
1247           itself.
1248    
1249    
1250    CODE COVERAGE REPORTING
1251    
1252           If your C compiler is gcc, you can build a version  of  PCRE  that  can
1253           generate a code coverage report for its test suite. To enable this, you
1254           must install lcov version 1.6 or above. Then specify
1255    
1256             --enable-coverage
1257    
1258           to the configure command and build PCRE in the usual way.
1259    
1260           Note that using ccache (a caching C compiler) is incompatible with code
1261           coverage  reporting. If you have configured ccache to run automatically
1262           on your system, you must set the environment variable
1263    
1264             CCACHE_DISABLE=1
1265    
1266           before running make to build PCRE, so that ccache is not used.
1267    
1268           When --enable-coverage is used,  the  following  addition  targets  are
1269           added to the Makefile:
1270    
1271             make coverage
1272    
1273           This  creates  a  fresh  coverage report for the PCRE test suite. It is
1274           equivalent to running "make coverage-reset", "make  coverage-baseline",
1275           "make check", and then "make coverage-report".
1276    
1277             make coverage-reset
1278    
1279           This zeroes the coverage counters, but does nothing else.
1280    
1281             make coverage-baseline
1282    
1283           This captures baseline coverage information.
1284    
1285             make coverage-report
1286    
1287           This creates the coverage report.
1288    
1289             make coverage-clean-report
1290    
1291           This  removes the generated coverage report without cleaning the cover-
1292           age data itself.
1293    
1294             make coverage-clean-data
1295    
1296           This removes the captured coverage data without removing  the  coverage
1297           files created at compile time (*.gcno).
1298    
1299             make coverage-clean
1300    
1301           This  cleans all coverage data including the generated coverage report.
1302           For more information about code coverage, see the gcov and  lcov  docu-
1303           mentation.
1304    
1305    
1306    SEE ALSO
1307    
1308           pcreapi(3), pcre16, pcre32, pcre_config(3).
1309    
1310    
1311    AUTHOR
1312    
1313           Philip Hazel
1314           University Computing Service
1315           Cambridge CB2 3QH, England.
1316    
1317    
1318    REVISION
1319    
1320           Last updated: 30 October 2012
1321           Copyright (c) 1997-2012 University of Cambridge.
1322    ------------------------------------------------------------------------------
1323    
1324    
1325    PCREMATCHING(3)                                                PCREMATCHING(3)
1326    
1327    
1328    NAME
1329           PCRE - Perl-compatible regular expressions
1330    
1331    
1332    PCRE MATCHING ALGORITHMS
1333    
1334           This document describes the two different algorithms that are available
1335           in PCRE for matching a compiled regular expression against a given sub-
1336           ject  string.  The  "standard"  algorithm  is  the  one provided by the
1337           pcre_exec(), pcre16_exec() and pcre32_exec() functions. These  work  in
1338           the  same as as Perl's matching function, and provide a Perl-compatible
1339           matching  operation.   The  just-in-time  (JIT)  optimization  that  is
1340           described  in  the pcrejit documentation is compatible with these func-
1341           tions.
1342    
1343           An  alternative  algorithm  is   provided   by   the   pcre_dfa_exec(),
1344           pcre16_dfa_exec()  and  pcre32_dfa_exec()  functions; they operate in a
1345           different way, and are not Perl-compatible. This alternative has advan-
1346           tages and disadvantages compared with the standard algorithm, and these
1347           are described below.
1348    
1349           When there is only one possible way in which a given subject string can
1350           match  a pattern, the two algorithms give the same answer. A difference
1351           arises, however, when there are multiple possibilities. For example, if
1352           the pattern
1353    
1354             ^<.*>
1355    
1356           is matched against the string
1357    
1358             <something> <something else> <something further>
1359    
1360           there are three possible answers. The standard algorithm finds only one
1361           of them, whereas the alternative algorithm finds all three.
1362    
1363    
1364    REGULAR EXPRESSIONS AS TREES
1365    
1366           The set of strings that are matched by a regular expression can be rep-
1367           resented  as  a  tree structure. An unlimited repetition in the pattern
1368           makes the tree of infinite size, but it is still a tree.  Matching  the
1369           pattern  to a given subject string (from a given starting point) can be
1370           thought of as a search of the tree.  There are two  ways  to  search  a
1371           tree:  depth-first  and  breadth-first, and these correspond to the two
1372           matching algorithms provided by PCRE.
1373    
1374    
1375    THE STANDARD MATCHING ALGORITHM
1376    
1377           In the terminology of Jeffrey Friedl's book "Mastering Regular  Expres-
1378           sions",  the  standard  algorithm  is an "NFA algorithm". It conducts a
1379           depth-first search of the pattern tree. That is, it  proceeds  along  a
1380           single path through the tree, checking that the subject matches what is
1381           required. When there is a mismatch, the algorithm  tries  any  alterna-
1382           tives  at  the  current point, and if they all fail, it backs up to the
1383           previous branch point in the  tree,  and  tries  the  next  alternative
1384           branch  at  that  level.  This often involves backing up (moving to the
1385           left) in the subject string as well.  The  order  in  which  repetition
1386           branches  are  tried  is controlled by the greedy or ungreedy nature of
1387           the quantifier.
1388    
1389           If a leaf node is reached, a matching string has  been  found,  and  at
1390           that  point the algorithm stops. Thus, if there is more than one possi-
1391           ble match, this algorithm returns the first one that it finds.  Whether
1392           this  is the shortest, the longest, or some intermediate length depends
1393           on the way the greedy and ungreedy repetition quantifiers are specified
1394           in the pattern.
1395    
1396           Because  it  ends  up  with a single path through the tree, it is rela-
1397           tively straightforward for this algorithm to keep  track  of  the  sub-
1398           strings  that  are  matched  by portions of the pattern in parentheses.
1399           This provides support for capturing parentheses and back references.
1400    
1401    
1402    THE ALTERNATIVE MATCHING ALGORITHM
1403    
1404           This algorithm conducts a breadth-first search of  the  tree.  Starting
1405           from  the  first  matching  point  in the subject, it scans the subject
1406           string from left to right, once, character by character, and as it does
1407           this,  it remembers all the paths through the tree that represent valid
1408           matches. In Friedl's terminology, this is a kind  of  "DFA  algorithm",
1409           though  it is not implemented as a traditional finite state machine (it
1410           keeps multiple states active simultaneously).
1411    
1412           Although the general principle of this matching algorithm  is  that  it
1413           scans  the subject string only once, without backtracking, there is one
1414           exception: when a lookaround assertion is encountered,  the  characters
1415           following  or  preceding  the  current  point  have to be independently
1416           inspected.
1417    
1418           The scan continues until either the end of the subject is  reached,  or
1419           there  are  no more unterminated paths. At this point, terminated paths
1420           represent the different matching possibilities (if there are none,  the
1421           match  has  failed).   Thus,  if there is more than one possible match,
1422           this algorithm finds all of them, and in particular, it finds the long-
1423           est.  The  matches are returned in decreasing order of length. There is
1424           an option to stop the algorithm after the first match (which is  neces-
1425           sarily the shortest) is found.
1426    
1427           Note that all the matches that are found start at the same point in the
1428           subject. If the pattern
1429    
1430             cat(er(pillar)?)?
1431    
1432           is matched against the string "the caterpillar catchment",  the  result
1433           will  be the three strings "caterpillar", "cater", and "cat" that start
1434           at the fifth character of the subject. The algorithm does not automati-
1435           cally move on to find matches that start at later positions.
1436    
1437           There are a number of features of PCRE regular expressions that are not
1438           supported by the alternative matching algorithm. They are as follows:
1439    
1440           1. Because the algorithm finds all  possible  matches,  the  greedy  or
1441           ungreedy  nature  of repetition quantifiers is not relevant. Greedy and
1442           ungreedy quantifiers are treated in exactly the same way. However, pos-
1443           sessive  quantifiers can make a difference when what follows could also
1444           match what is quantified, for example in a pattern like this:
1445    
1446             ^a++\w!
1447    
1448           This pattern matches "aaab!" but not "aaa!", which would be matched  by
1449           a  non-possessive quantifier. Similarly, if an atomic group is present,
1450           it is matched as if it were a standalone pattern at the current  point,
1451           and  the  longest match is then "locked in" for the rest of the overall
1452           pattern.
1453    
1454           2. When dealing with multiple paths through the tree simultaneously, it
1455           is  not  straightforward  to  keep track of captured substrings for the
1456           different matching possibilities, and  PCRE's  implementation  of  this
1457           algorithm does not attempt to do this. This means that no captured sub-
1458           strings are available.
1459    
1460           3. Because no substrings are captured, back references within the  pat-
1461           tern are not supported, and cause errors if encountered.
1462    
1463           4.  For  the same reason, conditional expressions that use a backrefer-
1464           ence as the condition or test for a specific group  recursion  are  not
1465           supported.
1466    
1467           5.  Because  many  paths  through the tree may be active, the \K escape
1468           sequence, which resets the start of the match when encountered (but may
1469           be  on  some  paths  and not on others), is not supported. It causes an
1470           error if encountered.
1471    
1472           6. Callouts are supported, but the value of the  capture_top  field  is
1473           always 1, and the value of the capture_last field is always -1.
1474    
1475           7.  The  \C  escape  sequence, which (in the standard algorithm) always
1476           matches a single data unit, even in UTF-8, UTF-16 or UTF-32  modes,  is
1477           not  supported  in these modes, because the alternative algorithm moves
1478           through the subject string one character (not data unit) at a time, for
1479           all active paths through the tree.
1480    
1481           8.  Except for (*FAIL), the backtracking control verbs such as (*PRUNE)
1482           are not supported. (*FAIL) is supported, and  behaves  like  a  failing
1483           negative assertion.
1484    
1485    
1486    ADVANTAGES OF THE ALTERNATIVE ALGORITHM
1487    
1488           Using  the alternative matching algorithm provides the following advan-
1489           tages:
1490    
1491           1. All possible matches (at a single point in the subject) are automat-
1492           ically  found,  and  in particular, the longest match is found. To find
1493           more than one match using the standard algorithm, you have to do kludgy
1494           things with callouts.
1495    
1496           2.  Because  the  alternative  algorithm  scans the subject string just
1497           once, and never needs to backtrack (except for lookbehinds), it is pos-
1498           sible  to  pass  very  long subject strings to the matching function in
1499           several pieces, checking for partial matching each time. Although it is
1500           possible  to  do multi-segment matching using the standard algorithm by
1501           retaining partially matched substrings, it  is  more  complicated.  The
1502           pcrepartial  documentation  gives  details of partial matching and dis-
1503           cusses multi-segment matching.
1504    
1505    
1506    DISADVANTAGES OF THE ALTERNATIVE ALGORITHM
1507    
1508           The alternative algorithm suffers from a number of disadvantages:
1509    
1510           1. It is substantially slower than  the  standard  algorithm.  This  is
1511           partly  because  it has to search for all possible matches, but is also
1512           because it is less susceptible to optimization.
1513    
1514           2. Capturing parentheses and back references are not supported.
1515    
1516           3. Although atomic groups are supported, their use does not provide the
1517           performance advantage that it does for the standard algorithm.
1518    
1519    
1520    AUTHOR
1521    
1522           Philip Hazel
1523           University Computing Service
1524           Cambridge CB2 3QH, England.
1525    
1526    
1527    REVISION
1528    
1529           Last updated: 08 January 2012
1530           Copyright (c) 1997-2012 University of Cambridge.
1531    ------------------------------------------------------------------------------
1532    
1533    
1534    PCREAPI(3)                                                          PCREAPI(3)
1535    
1536    
1537    NAME
1538           PCRE - Perl-compatible regular expressions
1539    
1540           #include <pcre.h>
1541    
1542    
1543    PCRE NATIVE API BASIC FUNCTIONS
1544    
1545           pcre *pcre_compile(const char *pattern, int options,
1546                const char **errptr, int *erroffset,
1547                const unsigned char *tableptr);
1548    
1549           pcre *pcre_compile2(const char *pattern, int options,
1550                int *errorcodeptr,
1551                const char **errptr, int *erroffset,
1552                const unsigned char *tableptr);
1553    
1554           pcre_extra *pcre_study(const pcre *code, int options,
1555                const char **errptr);
1556    
1557           void pcre_free_study(pcre_extra *extra);
1558    
1559           int pcre_exec(const pcre *code, const pcre_extra *extra,
1560                const char *subject, int length, int startoffset,
1561                int options, int *ovector, int ovecsize);
1562    
1563           int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
1564                const char *subject, int length, int startoffset,
1565                int options, int *ovector, int ovecsize,
1566                int *workspace, int wscount);
1567    
1568    
1569    PCRE NATIVE API STRING EXTRACTION FUNCTIONS
1570    
1571           int pcre_copy_named_substring(const pcre *code,
1572                const char *subject, int *ovector,
1573                int stringcount, const char *stringname,
1574                char *buffer, int buffersize);
1575    
1576           int pcre_copy_substring(const char *subject, int *ovector,
1577                int stringcount, int stringnumber, char *buffer,
1578                int buffersize);
1579    
1580           int pcre_get_named_substring(const pcre *code,
1581                const char *subject, int *ovector,
1582                int stringcount, const char *stringname,
1583                const char **stringptr);
1584    
1585           int pcre_get_stringnumber(const pcre *code,
1586                const char *name);
1587    
1588           int pcre_get_stringtable_entries(const pcre *code,
1589                const char *name, char **first, char **last);
1590    
1591           int pcre_get_substring(const char *subject, int *ovector,
1592                int stringcount, int stringnumber,
1593                const char **stringptr);
1594    
1595           int pcre_get_substring_list(const char *subject,
1596                int *ovector, int stringcount, const char ***listptr);
1597    
1598           void pcre_free_substring(const char *stringptr);
1599    
1600           void pcre_free_substring_list(const char **stringptr);
1601    
1602    
1603    PCRE NATIVE API AUXILIARY FUNCTIONS
1604    
1605           int pcre_jit_exec(const pcre *code, const pcre_extra *extra,
1606                const char *subject, int length, int startoffset,
1607                int options, int *ovector, int ovecsize,
1608                pcre_jit_stack *jstack);
1609    
1610           pcre_jit_stack *pcre_jit_stack_alloc(int startsize, int maxsize);
1611    
1612           void pcre_jit_stack_free(pcre_jit_stack *stack);
1613    
1614           void pcre_assign_jit_stack(pcre_extra *extra,
1615                pcre_jit_callback callback, void *data);
1616    
1617           const unsigned char *pcre_maketables(void);
1618    
1619           int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
1620                int what, void *where);
1621    
1622           int pcre_refcount(pcre *code, int adjust);
1623    
1624           int pcre_config(int what, void *where);
1625    
1626           const char *pcre_version(void);
1627    
1628           int pcre_pattern_to_host_byte_order(pcre *code,
1629                pcre_extra *extra, const unsigned char *tables);
1630    
1631    
1632    PCRE NATIVE API INDIRECTED FUNCTIONS
1633    
1634           void *(*pcre_malloc)(size_t);
1635    
1636           void (*pcre_free)(void *);
1637    
1638           void *(*pcre_stack_malloc)(size_t);
1639    
1640           void (*pcre_stack_free)(void *);
1641    
1642           int (*pcre_callout)(pcre_callout_block *);
1643    
1644    
1645    PCRE 8-BIT, 16-BIT AND 32-BIT LIBRARIES
1646    
1647           From  release  8.30,  PCRE  can  be  compiled as a library for handling
1648           16-bit character strings as  well  as,  or  instead  of,  the  original
1649           library  that  handles 8-bit character strings. From release 8.32, PCRE
1650           can also be  compiled  as  a  library  for  handling  32-bit  character
1651           strings.  To  avoid  too much complication, this document describes the
1652           8-bit versions of the functions, with only occasional references to the
1653           16-bit and 32-bit libraries.
1654    
1655           The  16-bit and 32-bit functions operate in the same way as their 8-bit
1656           counterparts; they just use different data types  for  their  arguments
1657           and  results,  and their names start with pcre16_ or pcre32_ instead of
1658           pcre_. For every option  that  has  UTF8  in  its  name  (for  example,
1659           PCRE_UTF8),  there  are corresponding 16-bit and 32-bit names with UTF8
1660           replaced by UTF16 or UTF32, respectively. This facility is in fact just
1661           cosmetic;  the  16-bit and 32-bit option names define the same bit val-
1662           ues.
1663    
1664           References to bytes and UTF-8 in this document should be read as refer-
1665           ences  to  16-bit  data  quantities  and  UTF-16  when using the 16-bit
1666           library, or 32-bit data quantities and UTF-32  when  using  the  32-bit
1667           library,  unless specified otherwise. More details of the specific dif-
1668           ferences for the 16-bit and 32-bit libraries are given  in  the  pcre16
1669           and pcre32 pages.
1670    
1671    
1672    PCRE API OVERVIEW
1673    
1674           PCRE has its own native API, which is described in this document. There
1675           are also some wrapper functions (for the 8-bit library only) that  cor-
1676           respond  to  the  POSIX  regular  expression  API, but they do not give
1677           access to all the functionality. They are described  in  the  pcreposix
1678           documentation.  Both  of these APIs define a set of C function calls. A
1679           C++ wrapper (again for the 8-bit library only) is also distributed with
1680           PCRE. It is documented in the pcrecpp page.
1681    
1682           The  native  API  C  function prototypes are defined in the header file
1683           pcre.h, and on Unix-like systems the (8-bit) library itself  is  called
1684           libpcre.  It  can  normally be accessed by adding -lpcre to the command
1685           for linking an application that uses PCRE. The header file defines  the
1686           macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release
1687           numbers for the library. Applications can use these to include  support
1688           for different releases of PCRE.
1689    
1690           In a Windows environment, if you want to statically link an application
1691           program against a non-dll pcre.a  file,  you  must  define  PCRE_STATIC
1692           before  including  pcre.h or pcrecpp.h, because otherwise the pcre_mal-
1693           loc()   and   pcre_free()   exported   functions   will   be   declared
1694           __declspec(dllimport), with unwanted results.
1695    
1696           The   functions   pcre_compile(),  pcre_compile2(),  pcre_study(),  and
1697           pcre_exec() are used for compiling and matching regular expressions  in
1698           a  Perl-compatible  manner. A sample program that demonstrates the sim-
1699           plest way of using them is provided in the file  called  pcredemo.c  in
1700           the PCRE source distribution. A listing of this program is given in the
1701           pcredemo documentation, and the pcresample documentation describes  how
1702           to compile and run it.
1703    
1704           Just-in-time  compiler  support is an optional feature of PCRE that can
1705           be built in appropriate hardware environments. It greatly speeds up the
1706           matching  performance  of  many  patterns.  Simple  programs can easily
1707           request that it be used if available, by  setting  an  option  that  is
1708           ignored  when  it is not relevant. More complicated programs might need
1709           to    make    use    of    the    functions     pcre_jit_stack_alloc(),
1710           pcre_jit_stack_free(),  and pcre_assign_jit_stack() in order to control
1711           the JIT code's memory usage.
1712    
1713           From release 8.32 there is also a direct interface for  JIT  execution,
1714           which  gives  improved performance. The JIT-specific functions are dis-
1715           cussed in the pcrejit documentation.
1716    
1717           A second matching function, pcre_dfa_exec(), which is not Perl-compati-
1718           ble,  is  also provided. This uses a different algorithm for the match-
1719           ing. The alternative algorithm finds all possible matches (at  a  given
1720           point  in  the  subject), and scans the subject just once (unless there
1721           are lookbehind assertions). However, this  algorithm  does  not  return
1722           captured  substrings.  A description of the two matching algorithms and
1723           their advantages and disadvantages is given in the  pcrematching  docu-
1724           mentation.
1725    
1726           In  addition  to  the  main compiling and matching functions, there are
1727           convenience functions for extracting captured substrings from a subject
1728           string that is matched by pcre_exec(). They are:
1729    
1730             pcre_copy_substring()
1731             pcre_copy_named_substring()
1732             pcre_get_substring()
1733             pcre_get_named_substring()
1734             pcre_get_substring_list()
1735             pcre_get_stringnumber()
1736             pcre_get_stringtable_entries()
1737    
1738           pcre_free_substring() and pcre_free_substring_list() are also provided,
1739           to free the memory used for extracted strings.
1740    
1741           The function pcre_maketables() is used to  build  a  set  of  character
1742           tables   in   the   current   locale  for  passing  to  pcre_compile(),
1743           pcre_exec(), or pcre_dfa_exec(). This is an optional facility  that  is
1744           provided  for  specialist  use.  Most  commonly,  no special tables are
1745           passed, in which case internal tables that are generated when  PCRE  is
1746           built are used.
1747    
1748           The  function  pcre_fullinfo()  is used to find out information about a
1749           compiled pattern. The function pcre_version() returns a  pointer  to  a
1750           string containing the version of PCRE and its date of release.
1751    
1752           The  function  pcre_refcount()  maintains  a  reference count in a data
1753           block containing a compiled pattern. This is provided for  the  benefit
1754           of object-oriented applications.
1755    
1756           The  global  variables  pcre_malloc and pcre_free initially contain the
1757           entry points of the standard malloc()  and  free()  functions,  respec-
1758           tively. PCRE calls the memory management functions via these variables,
1759           so a calling program can replace them if it  wishes  to  intercept  the
1760           calls. This should be done before calling any PCRE functions.
1761    
1762           The  global  variables  pcre_stack_malloc  and pcre_stack_free are also
1763           indirections to memory management functions.  These  special  functions
1764           are  used  only  when  PCRE is compiled to use the heap for remembering
1765           data, instead of recursive function calls, when running the pcre_exec()
1766           function.  See  the  pcrebuild  documentation  for details of how to do
1767           this. It is a non-standard way of building PCRE, for  use  in  environ-
1768           ments  that  have  limited stacks. Because of the greater use of memory
1769           management, it runs more slowly. Separate  functions  are  provided  so
1770           that  special-purpose  external  code  can  be used for this case. When
1771           used, these functions are always called in a  stack-like  manner  (last
1772           obtained,  first freed), and always for memory blocks of the same size.
1773           There is a discussion about PCRE's stack usage in the  pcrestack  docu-
1774           mentation.
1775    
1776           The global variable pcre_callout initially contains NULL. It can be set
1777           by the caller to a "callout" function, which PCRE  will  then  call  at
1778           specified  points during a matching operation. Details are given in the
1779           pcrecallout documentation.
1780    
1781    
1782    NEWLINES
1783    
1784           PCRE supports five different conventions for indicating line breaks  in
1785           strings:  a  single  CR (carriage return) character, a single LF (line-
1786           feed) character, the two-character sequence CRLF, any of the three pre-
1787           ceding,  or any Unicode newline sequence. The Unicode newline sequences
1788           are the three just mentioned, plus the single characters  VT  (vertical
1789           tab, U+000B), FF (form feed, U+000C), NEL (next line, U+0085), LS (line
1790           separator, U+2028), and PS (paragraph separator, U+2029).
1791    
1792           Each of the first three conventions is used by at least  one  operating
1793           system  as its standard newline sequence. When PCRE is built, a default
1794           can be specified.  The default default is LF, which is the  Unix  stan-
1795           dard.  When  PCRE  is run, the default can be overridden, either when a
1796           pattern is compiled, or when it is matched.
1797    
1798           At compile time, the newline convention can be specified by the options
1799           argument  of  pcre_compile(), or it can be specified by special text at
1800           the start of the pattern itself; this overrides any other settings. See
1801           the pcrepattern page for details of the special character sequences.
1802    
1803           In the PCRE documentation the word "newline" is used to mean "the char-
1804           acter or pair of characters that indicate a line break". The choice  of
1805           newline  convention  affects  the  handling of the dot, circumflex, and
1806           dollar metacharacters, the handling of #-comments in /x mode, and, when
1807           CRLF  is a recognized line ending sequence, the match position advance-
1808           ment for a non-anchored pattern. There is more detail about this in the
1809           section on pcre_exec() options below.
1810    
1811           The  choice of newline convention does not affect the interpretation of
1812           the \n or \r escape sequences, nor does  it  affect  what  \R  matches,
1813           which is controlled in a similar way, but by separate options.
1814    
1815    
1816    MULTITHREADING
1817    
1818           The  PCRE  functions  can be used in multi-threading applications, with
1819           the  proviso  that  the  memory  management  functions  pointed  to  by
1820           pcre_malloc, pcre_free, pcre_stack_malloc, and pcre_stack_free, and the
1821           callout function pointed to by pcre_callout, are shared by all threads.
1822    
1823           The compiled form of a regular expression is not altered during  match-
1824           ing, so the same compiled pattern can safely be used by several threads
1825           at once.
1826    
1827           If the just-in-time optimization feature is being used, it needs  sepa-
1828           rate  memory stack areas for each thread. See the pcrejit documentation
1829           for more details.
1830    
1831    
1832    SAVING PRECOMPILED PATTERNS FOR LATER USE
1833    
1834           The compiled form of a regular expression can be saved and re-used at a
1835           later  time,  possibly by a different program, and even on a host other
1836           than the one on which  it  was  compiled.  Details  are  given  in  the
1837           pcreprecompile  documentation,  which  includes  a  description  of the
1838           pcre_pattern_to_host_byte_order() function. However, compiling a  regu-
1839           lar  expression  with one version of PCRE for use with a different ver-
1840           sion is not guaranteed to work and may cause crashes.
1841    
1842    
1843    CHECKING BUILD-TIME OPTIONS
1844    
1845           int pcre_config(int what, void *where);
1846    
1847           The function pcre_config() makes it possible for a PCRE client to  dis-
1848           cover which optional features have been compiled into the PCRE library.
1849           The pcrebuild documentation has more details about these optional  fea-
1850           tures.
1851    
1852           The  first  argument  for pcre_config() is an integer, specifying which
1853           information is required; the second argument is a pointer to a variable
1854           into  which  the  information  is placed. The returned value is zero on
1855           success, or the negative error code PCRE_ERROR_BADOPTION if  the  value
1856           in  the  first argument is not recognized. The following information is
1857           available:
1858    
1859             PCRE_CONFIG_UTF8
1860    
1861           The output is an integer that is set to one if UTF-8 support is  avail-
1862           able;  otherwise it is set to zero. This value should normally be given
1863           to the 8-bit version of this function, pcre_config(). If it is given to
1864           the   16-bit  or  32-bit  version  of  this  function,  the  result  is
1865           PCRE_ERROR_BADOPTION.
1866    
1867             PCRE_CONFIG_UTF16
1868    
1869           The output is an integer that is set to one if UTF-16 support is avail-
1870           able;  otherwise it is set to zero. This value should normally be given
1871           to the 16-bit version of this function, pcre16_config(). If it is given
1872           to  the  8-bit  or  32-bit  version  of  this  function,  the result is
1873           PCRE_ERROR_BADOPTION.
1874    
1875             PCRE_CONFIG_UTF32
1876    
1877           The output is an integer that is set to one if UTF-32 support is avail-
1878           able;  otherwise it is set to zero. This value should normally be given
1879           to the 32-bit version of this function, pcre32_config(). If it is given
1880           to  the  8-bit  or  16-bit  version  of  this  function,  the result is
1881           PCRE_ERROR_BADOPTION.
1882    
1883             PCRE_CONFIG_UNICODE_PROPERTIES
1884    
1885           The output is an integer that is set to  one  if  support  for  Unicode
1886           character properties is available; otherwise it is set to zero.
1887    
1888             PCRE_CONFIG_JIT
1889    
1890           The output is an integer that is set to one if support for just-in-time
1891           compiling is available; otherwise it is set to zero.
1892    
1893             PCRE_CONFIG_JITTARGET
1894    
1895           The output is a pointer to a zero-terminated "const char *" string.  If
1896           JIT support is available, the string contains the name of the architec-
1897           ture for which the JIT compiler is configured, for example  "x86  32bit
1898           (little  endian  +  unaligned)".  If  JIT support is not available, the
1899           result is NULL.
1900    
1901             PCRE_CONFIG_NEWLINE
1902    
1903           The output is an integer whose value specifies  the  default  character
1904           sequence  that  is recognized as meaning "newline". The values that are
1905           supported in ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338
1906           for  CRLF,  -2 for ANYCRLF, and -1 for ANY. In EBCDIC environments, CR,
1907           ANYCRLF, and ANY yield the same values. However, the value  for  LF  is
1908           normally  21, though some EBCDIC environments use 37. The corresponding
1909           values for CRLF are 3349 and 3365. The default should  normally  corre-
1910           spond to the standard sequence for your operating system.
1911    
1912             PCRE_CONFIG_BSR
1913    
1914           The output is an integer whose value indicates what character sequences
1915           the \R escape sequence matches by default. A value of 0 means  that  \R
1916           matches  any  Unicode  line ending sequence; a value of 1 means that \R
1917           matches only CR, LF, or CRLF. The default can be overridden when a pat-
1918           tern is compiled or matched.
1919    
1920             PCRE_CONFIG_LINK_SIZE
1921    
1922           The  output  is  an  integer that contains the number of bytes used for
1923           internal  linkage  in  compiled  regular  expressions.  For  the  8-bit
1924           library, the value can be 2, 3, or 4. For the 16-bit library, the value
1925           is either 2 or 4 and is  still  a  number  of  bytes.  For  the  32-bit
1926           library, the value is either 2 or 4 and is still a number of bytes. The
1927           default value of 2 is sufficient for all but the most massive patterns,
1928           since  it  allows  the compiled pattern to be up to 64K in size. Larger
1929           values allow larger regular expressions to be compiled, at the  expense
1930           of slower matching.
1931    
1932             PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1933    
1934           The  output  is  an integer that contains the threshold above which the
1935           POSIX interface uses malloc() for output vectors. Further  details  are
1936           given in the pcreposix documentation.
1937    
1938             PCRE_CONFIG_MATCH_LIMIT
1939    
1940           The  output is a long integer that gives the default limit for the num-
1941           ber of internal matching function calls  in  a  pcre_exec()  execution.
1942           Further details are given with pcre_exec() below.
1943    
1944             PCRE_CONFIG_MATCH_LIMIT_RECURSION
1945    
1946           The output is a long integer that gives the default limit for the depth
1947           of  recursion  when  calling  the  internal  matching  function  in   a
1948           pcre_exec()  execution.  Further  details  are  given  with pcre_exec()
1949           below.
1950    
1951             PCRE_CONFIG_STACKRECURSE
1952    
1953           The output is an integer that is set to one if internal recursion  when
1954           running pcre_exec() is implemented by recursive function calls that use
1955           the stack to remember their state. This is the usual way that  PCRE  is
1956           compiled. The output is zero if PCRE was compiled to use blocks of data
1957           on the  heap  instead  of  recursive  function  calls.  In  this  case,
1958           pcre_stack_malloc  and  pcre_stack_free  are  called  to  manage memory
1959           blocks on the heap, thus avoiding the use of the stack.
1960    
1961    
1962    COMPILING A PATTERN
1963    
1964           pcre *pcre_compile(const char *pattern, int options,
1965                const char **errptr, int *erroffset,
1966                const unsigned char *tableptr);
1967    
1968           pcre *pcre_compile2(const char *pattern, int options,
1969                int *errorcodeptr,
1970                const char **errptr, int *erroffset,
1971                const unsigned char *tableptr);
1972    
1973           Either of the functions pcre_compile() or pcre_compile2() can be called
1974           to compile a pattern into an internal form. The only difference between
1975           the two interfaces is that pcre_compile2() has an additional  argument,
1976           errorcodeptr,  via  which  a  numerical  error code can be returned. To
1977           avoid too much repetition, we refer just to pcre_compile()  below,  but
1978           the information applies equally to pcre_compile2().
1979    
1980           The pattern is a C string terminated by a binary zero, and is passed in
1981           the pattern argument. A pointer to a single block  of  memory  that  is
1982           obtained  via  pcre_malloc is returned. This contains the compiled code
1983           and related data. The pcre type is defined for the returned block; this
1984           is a typedef for a structure whose contents are not externally defined.
1985           It is up to the caller to free the memory (via pcre_free) when it is no
1986           longer required.
1987    
1988           Although  the compiled code of a PCRE regex is relocatable, that is, it
1989           does not depend on memory location, the complete pcre data block is not
1990           fully  relocatable, because it may contain a copy of the tableptr argu-
1991           ment, which is an address (see below).
1992    
1993           The options argument contains various bit settings that affect the com-
1994           pilation.  It  should be zero if no options are required. The available
1995           options are described below. Some of them (in  particular,  those  that
1996           are  compatible with Perl, but some others as well) can also be set and
1997           unset from within the pattern (see  the  detailed  description  in  the
1998           pcrepattern  documentation). For those options that can be different in
1999           different parts of the pattern, the contents of  the  options  argument
2000           specifies their settings at the start of compilation and execution. The
2001           PCRE_ANCHORED, PCRE_BSR_xxx, PCRE_NEWLINE_xxx, PCRE_NO_UTF8_CHECK,  and
2002           PCRE_NO_START_OPTIMIZE  options  can  be set at the time of matching as
2003           well as at compile time.
2004    
2005           If errptr is NULL, pcre_compile() returns NULL immediately.  Otherwise,
2006           if  compilation  of  a  pattern fails, pcre_compile() returns NULL, and
2007           sets the variable pointed to by errptr to point to a textual error mes-
2008           sage. This is a static string that is part of the library. You must not
2009           try to free it. Normally, the offset from the start of the  pattern  to
2010           the  byte  that  was  being  processed when the error was discovered is
2011           placed in the variable pointed to by erroffset, which must not be  NULL
2012           (if  it is, an immediate error is given). However, for an invalid UTF-8
2013           string, the offset is that of the first byte of the failing character.
2014    
2015           Some errors are not detected until the whole pattern has been  scanned;
2016           in  these  cases,  the offset passed back is the length of the pattern.
2017           Note that the offset is in bytes, not characters, even in  UTF-8  mode.
2018           It may sometimes point into the middle of a UTF-8 character.
2019    
2020           If  pcre_compile2()  is  used instead of pcre_compile(), and the error-
2021           codeptr argument is not NULL, a non-zero error code number is  returned
2022           via  this argument in the event of an error. This is in addition to the
2023           textual error message. Error codes and messages are listed below.
2024    
2025           If the final argument, tableptr, is NULL, PCRE uses a  default  set  of
2026           character  tables  that  are  built  when  PCRE  is compiled, using the
2027           default C locale. Otherwise, tableptr must be an address  that  is  the
2028           result  of  a  call to pcre_maketables(). This value is stored with the
2029           compiled pattern, and used again by pcre_exec(), unless  another  table
2030           pointer is passed to it. For more discussion, see the section on locale
2031           support below.
2032    
2033           This code fragment shows a typical straightforward  call  to  pcre_com-
2034           pile():
2035    
2036             pcre *re;
2037             const char *error;
2038             int erroffset;
2039             re = pcre_compile(
2040               "^A.*Z",          /* the pattern */
2041               0,                /* default options */
2042               &error,           /* for error message */
2043               &erroffset,       /* for error offset */
2044               NULL);            /* use default character tables */
2045    
2046           The  following  names  for option bits are defined in the pcre.h header
2047           file:
2048    
2049             PCRE_ANCHORED
2050    
2051           If this bit is set, the pattern is forced to be "anchored", that is, it
2052           is  constrained to match only at the first matching point in the string
2053           that is being searched (the "subject string"). This effect can also  be
2054           achieved  by appropriate constructs in the pattern itself, which is the
2055           only way to do it in Perl.
2056    
2057             PCRE_AUTO_CALLOUT
2058    
2059           If this bit is set, pcre_compile() automatically inserts callout items,
2060           all  with  number  255, before each pattern item. For discussion of the
2061           callout facility, see the pcrecallout documentation.
2062    
2063             PCRE_BSR_ANYCRLF
2064             PCRE_BSR_UNICODE
2065    
2066           These options (which are mutually exclusive) control what the \R escape
2067           sequence  matches.  The choice is either to match only CR, LF, or CRLF,
2068           or to match any Unicode newline sequence. The default is specified when
2069           PCRE is built. It can be overridden from within the pattern, or by set-
2070           ting an option when a compiled pattern is matched.
2071    
2072             PCRE_CASELESS
2073    
2074           If this bit is set, letters in the pattern match both upper  and  lower
2075           case  letters.  It  is  equivalent  to  Perl's /i option, and it can be
2076           changed within a pattern by a (?i) option setting. In UTF-8 mode,  PCRE
2077           always  understands the concept of case for characters whose values are
2078           less than 128, so caseless matching is always possible. For  characters
2079           with  higher  values,  the concept of case is supported if PCRE is com-
2080           piled with Unicode property support, but not otherwise. If you want  to
2081           use  caseless  matching  for  characters 128 and above, you must ensure
2082           that PCRE is compiled with Unicode property support  as  well  as  with
2083           UTF-8 support.
2084    
2085             PCRE_DOLLAR_ENDONLY
2086    
2087           If  this bit is set, a dollar metacharacter in the pattern matches only
2088           at the end of the subject string. Without this option,  a  dollar  also
2089           matches  immediately before a newline at the end of the string (but not
2090           before any other newlines). The PCRE_DOLLAR_ENDONLY option  is  ignored
2091           if  PCRE_MULTILINE  is  set.   There is no equivalent to this option in
2092           Perl, and no way to set it within a pattern.
2093    
2094             PCRE_DOTALL
2095    
2096           If this bit is set, a dot metacharacter in the pattern matches a  char-
2097           acter of any value, including one that indicates a newline. However, it
2098           only ever matches one character, even if newlines are  coded  as  CRLF.
2099           Without  this option, a dot does not match when the current position is
2100           at a newline. This option is equivalent to Perl's /s option, and it can
2101           be  changed within a pattern by a (?s) option setting. A negative class
2102           such as [^a] always matches newline characters, independent of the set-
2103           ting of this option.
2104    
2105             PCRE_DUPNAMES
2106    
2107           If  this  bit is set, names used to identify capturing subpatterns need
2108           not be unique. This can be helpful for certain types of pattern when it
2109           is  known  that  only  one instance of the named subpattern can ever be
2110           matched. There are more details of named subpatterns  below;  see  also
2111           the pcrepattern documentation.
2112    
2113             PCRE_EXTENDED
2114    
2115           If  this  bit  is  set,  white space data characters in the pattern are
2116           totally ignored except when escaped or inside a character class.  White
2117           space does not include the VT character (code 11). In addition, charac-
2118           ters between an unescaped # outside a character class and the next new-
2119           line,  inclusive,  are  also  ignored.  This is equivalent to Perl's /x
2120           option, and it can be changed within a pattern by a  (?x)  option  set-
2121           ting.
2122    
2123           Which  characters  are  interpreted  as  newlines  is controlled by the
2124           options passed to pcre_compile() or by a special sequence at the  start
2125           of  the  pattern, as described in the section entitled "Newline conven-
2126           tions" in the pcrepattern documentation. Note that the end of this type
2127           of  comment  is  a  literal  newline  sequence  in  the pattern; escape
2128           sequences that happen to represent a newline do not count.
2129    
2130           This option makes it possible to include  comments  inside  complicated
2131           patterns.   Note,  however,  that this applies only to data characters.
2132           White space  characters  may  never  appear  within  special  character
2133           sequences in a pattern, for example within the sequence (?( that intro-
2134           duces a conditional subpattern.
2135    
2136             PCRE_EXTRA
2137    
2138           This option was invented in order to turn on  additional  functionality
2139           of  PCRE  that  is  incompatible with Perl, but it is currently of very
2140           little use. When set, any backslash in a pattern that is followed by  a
2141           letter  that  has  no  special  meaning causes an error, thus reserving
2142           these combinations for future expansion. By  default,  as  in  Perl,  a
2143           backslash  followed by a letter with no special meaning is treated as a
2144           literal. (Perl can, however, be persuaded to give an error for this, by
2145           running  it with the -w option.) There are at present no other features
2146           controlled by this option. It can also be set by a (?X) option  setting
2147           within a pattern.
2148    
2149             PCRE_FIRSTLINE
2150    
2151           If  this  option  is  set,  an  unanchored pattern is required to match
2152           before or at the first  newline  in  the  subject  string,  though  the
2153           matched text may continue over the newline.
2154    
2155             PCRE_JAVASCRIPT_COMPAT
2156    
2157           If this option is set, PCRE's behaviour is changed in some ways so that
2158           it is compatible with JavaScript rather than Perl. The changes  are  as
2159           follows:
2160    
2161           (1)  A  lone  closing square bracket in a pattern causes a compile-time
2162           error, because this is illegal in JavaScript (by default it is  treated
2163           as a data character). Thus, the pattern AB]CD becomes illegal when this
2164           option is set.
2165    
2166           (2) At run time, a back reference to an unset subpattern group  matches
2167           an  empty  string (by default this causes the current matching alterna-
2168           tive to fail). A pattern such as (\1)(a) succeeds when this  option  is
2169           set  (assuming  it can find an "a" in the subject), whereas it fails by
2170           default, for Perl compatibility.
2171    
2172           (3) \U matches an upper case "U" character; by default \U causes a com-
2173           pile time error (Perl uses \U to upper case subsequent characters).
2174    
2175           (4) \u matches a lower case "u" character unless it is followed by four
2176           hexadecimal digits, in which case the hexadecimal  number  defines  the
2177           code  point  to match. By default, \u causes a compile time error (Perl
2178           uses it to upper case the following character).
2179    
2180           (5) \x matches a lower case "x" character unless it is followed by  two
2181           hexadecimal  digits,  in  which case the hexadecimal number defines the
2182           code point to match. By default, as in Perl, a  hexadecimal  number  is
2183           always expected after \x, but it may have zero, one, or two digits (so,
2184           for example, \xz matches a binary zero character followed by z).
2185    
2186             PCRE_MULTILINE
2187    
2188           By default, PCRE treats the subject string as consisting  of  a  single
2189           line  of characters (even if it actually contains newlines). The "start
2190           of line" metacharacter (^) matches only at the  start  of  the  string,
2191           while  the  "end  of line" metacharacter ($) matches only at the end of
2192           the string, or before a terminating newline (unless PCRE_DOLLAR_ENDONLY
2193           is set). This is the same as Perl.
2194    
2195           When  PCRE_MULTILINE  it  is set, the "start of line" and "end of line"
2196           constructs match immediately following or immediately  before  internal
2197           newlines  in  the  subject string, respectively, as well as at the very
2198           start and end. This is equivalent to Perl's /m option, and  it  can  be
2199           changed within a pattern by a (?m) option setting. If there are no new-
2200           lines in a subject string, or no occurrences of ^ or $  in  a  pattern,
2201           setting PCRE_MULTILINE has no effect.
2202    
2203             PCRE_NEWLINE_CR
2204             PCRE_NEWLINE_LF
2205             PCRE_NEWLINE_CRLF
2206             PCRE_NEWLINE_ANYCRLF
2207             PCRE_NEWLINE_ANY
2208    
2209           These  options  override the default newline definition that was chosen
2210           when PCRE was built. Setting the first or the second specifies  that  a
2211           newline  is  indicated  by a single character (CR or LF, respectively).
2212           Setting PCRE_NEWLINE_CRLF specifies that a newline is indicated by  the
2213           two-character  CRLF  sequence.  Setting  PCRE_NEWLINE_ANYCRLF specifies
2214           that any of the three preceding sequences should be recognized. Setting
2215           PCRE_NEWLINE_ANY  specifies that any Unicode newline sequence should be
2216           recognized.
2217    
2218           In an ASCII/Unicode environment, the Unicode newline sequences are  the
2219           three  just  mentioned,  plus  the  single characters VT (vertical tab,
2220           U+000B), FF (form feed, U+000C), NEL (next line, U+0085), LS (line sep-
2221           arator,  U+2028),  and  PS (paragraph separator, U+2029). For the 8-bit
2222           library, the last two are recognized only in UTF-8 mode.
2223    
2224           When PCRE is compiled to run in an EBCDIC (mainframe) environment,  the
2225           code for CR is 0x0d, the same as ASCII. However, the character code for
2226           LF is normally 0x15, though in some EBCDIC environments 0x25  is  used.
2227           Whichever  of  these  is  not LF is made to correspond to Unicode's NEL
2228           character. EBCDIC codes are all less than 256. For  more  details,  see
2229           the pcrebuild documentation.
2230    
2231           The  newline  setting  in  the  options  word  uses three bits that are
2232           treated as a number, giving eight possibilities. Currently only six are
2233           used  (default  plus the five values above). This means that if you set
2234           more than one newline option, the combination may or may not be  sensi-
2235           ble. For example, PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to
2236           PCRE_NEWLINE_CRLF, but other combinations may yield unused numbers  and
2237           cause an error.
2238    
2239           The  only  time  that a line break in a pattern is specially recognized
2240           when compiling is when PCRE_EXTENDED is set. CR and LF are white  space
2241           characters,  and so are ignored in this mode. Also, an unescaped # out-
2242           side a character class indicates a comment that lasts until  after  the
2243           next  line break sequence. In other circumstances, line break sequences
2244           in patterns are treated as literal data.
2245    
2246           The newline option that is set at compile time becomes the default that
2247           is used for pcre_exec() and pcre_dfa_exec(), but it can be overridden.
2248    
2249             PCRE_NO_AUTO_CAPTURE
2250    
2251           If this option is set, it disables the use of numbered capturing paren-
2252           theses in the pattern. Any opening parenthesis that is not followed  by
2253           ?  behaves as if it were followed by ?: but named parentheses can still
2254           be used for capturing (and they acquire  numbers  in  the  usual  way).
2255           There is no equivalent of this option in Perl.
2256    
2257             NO_START_OPTIMIZE
2258    
2259           This  is an option that acts at matching time; that is, it is really an
2260           option for pcre_exec() or pcre_dfa_exec(). If  it  is  set  at  compile
2261           time,  it is remembered with the compiled pattern and assumed at match-
2262           ing time. For details  see  the  discussion  of  PCRE_NO_START_OPTIMIZE
2263           below.
2264    
2265             PCRE_UCP
2266    
2267           This  option changes the way PCRE processes \B, \b, \D, \d, \S, \s, \W,
2268           \w, and some of the POSIX character classes.  By  default,  only  ASCII
2269           characters  are  recognized, but if PCRE_UCP is set, Unicode properties
2270           are used instead to classify characters. More details are given in  the
2271           section  on generic character types in the pcrepattern page. If you set
2272           PCRE_UCP, matching one of the items it affects takes much  longer.  The
2273           option  is  available only if PCRE has been compiled with Unicode prop-
2274           erty support.
2275    
2276             PCRE_UNGREEDY
2277    
2278           This option inverts the "greediness" of the quantifiers  so  that  they
2279           are  not greedy by default, but become greedy if followed by "?". It is
2280           not compatible with Perl. It can also be set by a (?U)  option  setting
2281           within the pattern.
2282    
2283             PCRE_UTF8
2284    
2285           This  option  causes PCRE to regard both the pattern and the subject as
2286           strings of UTF-8 characters instead of single-byte strings. However, it
2287           is  available  only  when PCRE is built to include UTF support. If not,
2288           the use of this option provokes an error. Details of  how  this  option
2289           changes the behaviour of PCRE are given in the pcreunicode page.
2290    
2291             PCRE_NO_UTF8_CHECK
2292    
2293           When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
2294           automatically checked. There is a  discussion  about  the  validity  of
2295           UTF-8  strings in the pcreunicode page. If an invalid UTF-8 sequence is
2296           found, pcre_compile() returns an error. If you already know  that  your
2297           pattern  is valid, and you want to skip this check for performance rea-
2298           sons, you can set the PCRE_NO_UTF8_CHECK option.  When it is  set,  the
2299           effect of passing an invalid UTF-8 string as a pattern is undefined. It
2300           may cause your program to crash. Note that  this  option  can  also  be
2301           passed  to  pcre_exec()  and  pcre_dfa_exec(), to suppress the validity
2302           checking of subject strings only. If the same string is  being  matched
2303           many  times, the option can be safely set for the second and subsequent
2304           matchings to improve performance.
2305    
2306    
2307    COMPILATION ERROR CODES
2308    
2309           The following table lists the error  codes  than  may  be  returned  by
2310           pcre_compile2(),  along with the error messages that may be returned by
2311           both compiling functions. Note that error  messages  are  always  8-bit
2312           ASCII  strings,  even  in 16-bit or 32-bit mode. As PCRE has developed,
2313           some error codes have fallen out of use. To avoid confusion, they  have
2314           not been re-used.
2315    
2316              0  no error
2317              1  \ at end of pattern
2318              2  \c at end of pattern
2319              3  unrecognized character follows \
2320              4  numbers out of order in {} quantifier
2321              5  number too big in {} quantifier
2322              6  missing terminating ] for character class
2323              7  invalid escape sequence in character class
2324              8  range out of order in character class
2325              9  nothing to repeat
2326             10  [this code is not in use]
2327             11  internal error: unexpected repeat
2328             12  unrecognized character after (? or (?-
2329             13  POSIX named classes are supported only within a class
2330             14  missing )
2331             15  reference to non-existent subpattern
2332             16  erroffset passed as NULL
2333             17  unknown option bit(s) set
2334             18  missing ) after comment
2335             19  [this code is not in use]
2336             20  regular expression is too large
2337             21  failed to get memory
2338             22  unmatched parentheses
2339             23  internal error: code overflow
2340             24  unrecognized character after (?<
2341             25  lookbehind assertion is not fixed length
2342             26  malformed number or name after (?(
2343             27  conditional group contains more than two branches
2344             28  assertion expected after (?(
2345             29  (?R or (?[+-]digits must be followed by )
2346             30  unknown POSIX class name
2347             31  POSIX collating elements are not supported
2348             32  this version of PCRE is compiled without UTF support
2349             33  [this code is not in use]
2350             34  character value in \x{...} sequence is too large
2351             35  invalid condition (?(0)
2352             36  \C not allowed in lookbehind assertion
2353             37  PCRE does not support \L, \l, \N{name}, \U, or \u
2354             38  number after (?C is > 255
2355             39  closing ) for (?C expected
2356             40  recursive call could loop indefinitely
2357             41  unrecognized character after (?P
2358             42  syntax error in subpattern name (missing terminator)
2359             43  two named subpatterns have the same name
2360             44  invalid UTF-8 string (specifically UTF-8)
2361             45  support for \P, \p, and \X has not been compiled
2362             46  malformed \P or \p sequence
2363             47  unknown property name after \P or \p
2364             48  subpattern name is too long (maximum 32 characters)
2365             49  too many named subpatterns (maximum 10000)
2366             50  [this code is not in use]
2367             51  octal value is greater than \377 in 8-bit non-UTF-8 mode
2368             52  internal error: overran compiling workspace
2369             53  internal error: previously-checked referenced subpattern
2370                   not found
2371             54  DEFINE group contains more than one branch
2372             55  repeating a DEFINE group is not allowed
2373             56  inconsistent NEWLINE options
2374             57  \g is not followed by a braced, angle-bracketed, or quoted
2375                   name/number or by a plain number
2376             58  a numbered reference must not be zero
2377             59  an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
2378             60  (*VERB) not recognized
2379             61  number is too big
2380             62  subpattern name expected
2381             63  digit expected after (?+
2382             64  ] is an invalid data character in JavaScript compatibility mode
2383             65  different names for subpatterns of the same number are
2384                   not allowed
2385             66  (*MARK) must have an argument
2386             67  this version of PCRE is not compiled with Unicode property
2387                   support
2388             68  \c must be followed by an ASCII character
2389             69  \k is not followed by a braced, angle-bracketed, or quoted name
2390             70  internal error: unknown opcode in find_fixedlength()
2391             71  \N is not supported in a class
2392             72  too many forward references
2393             73  disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
2394             74  invalid UTF-16 string (specifically UTF-16)
2395             75  name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
2396             76  character value in \u.... sequence is too large
2397             77  invalid UTF-32 string (specifically UTF-32)
2398    
2399           The  numbers  32  and 10000 in errors 48 and 49 are defaults; different
2400           values may be used if the limits were changed when PCRE was built.
2401    
2402    
2403    STUDYING A PATTERN
2404    
2405           pcre_extra *pcre_study(const pcre *code, int options
2406                const char **errptr);
2407    
2408           If a compiled pattern is going to be used several times,  it  is  worth
2409           spending more time analyzing it in order to speed up the time taken for
2410           matching. The function pcre_study() takes a pointer to a compiled  pat-
2411           tern as its first argument. If studying the pattern produces additional
2412           information that will help speed up matching,  pcre_study()  returns  a
2413           pointer  to a pcre_extra block, in which the study_data field points to
2414           the results of the study.
2415    
2416           The  returned  value  from  pcre_study()  can  be  passed  directly  to
2417           pcre_exec()  or  pcre_dfa_exec(). However, a pcre_extra block also con-
2418           tains other fields that can be set by the caller before  the  block  is
2419           passed; these are described below in the section on matching a pattern.
2420    
2421           If  studying  the  pattern  does  not  produce  any useful information,
2422           pcre_study() returns NULL by default.  In  that  circumstance,  if  the
2423           calling program wants to pass any of the other fields to pcre_exec() or
2424           pcre_dfa_exec(), it must set up its own pcre_extra block.  However,  if
2425           pcre_study()  is  called  with  the  PCRE_STUDY_EXTRA_NEEDED option, it
2426           returns a pcre_extra block even if studying did not find any additional
2427           information.  It  may still return NULL, however, if an error occurs in
2428           pcre_study().
2429    
2430           The second argument of pcre_study() contains  option  bits.  There  are
2431           three further options in addition to PCRE_STUDY_EXTRA_NEEDED:
2432    
2433             PCRE_STUDY_JIT_COMPILE
2434             PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
2435             PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
2436    
2437           If  any  of  these are set, and the just-in-time compiler is available,
2438           the pattern is further compiled into machine code  that  executes  much
2439           faster  than  the  pcre_exec()  interpretive  matching function. If the
2440           just-in-time compiler is not available, these options are ignored.  All
2441           undefined bits in the options argument must be zero.
2442    
2443           JIT  compilation  is  a heavyweight optimization. It can take some time
2444           for patterns to be analyzed, and for one-off matches  and  simple  pat-
2445           terns  the benefit of faster execution might be offset by a much slower
2446           study time.  Not all patterns can be optimized by the JIT compiler. For
2447           those  that cannot be handled, matching automatically falls back to the
2448           pcre_exec() interpreter. For more details, see the  pcrejit  documenta-
2449           tion.
2450    
2451           The  third argument for pcre_study() is a pointer for an error message.
2452           If studying succeeds (even if no data is  returned),  the  variable  it
2453           points  to  is  set  to NULL. Otherwise it is set to point to a textual
2454           error message. This is a static string that is part of the library. You
2455           must  not  try  to  free it. You should test the error pointer for NULL
2456           after calling pcre_study(), to be sure that it has run successfully.
2457    
2458           When you are finished with a pattern, you can free the memory used  for
2459           the study data by calling pcre_free_study(). This function was added to
2460           the API for release 8.20. For earlier versions,  the  memory  could  be
2461           freed  with  pcre_free(), just like the pattern itself. This will still
2462           work in cases where JIT optimization is not used, but it  is  advisable
2463           to change to the new function when convenient.
2464    
2465           This  is  a typical way in which pcre_study() is used (except that in a
2466           real application there should be tests for errors):
2467    
2468             int rc;
2469             pcre *re;
2470             pcre_extra *sd;
2471             re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
2472             sd = pcre_study(
2473               re,             /* result of pcre_compile() */
2474               0,              /* no options */
2475               &error);        /* set to NULL or points to a message */
2476             rc = pcre_exec(   /* see below for details of pcre_exec() options */
2477               re, sd, "subject", 7, 0, 0, ovector, 30);
2478             ...
2479             pcre_free_study(sd);
2480             pcre_free(re);
2481    
2482           Studying a pattern does two things: first, a lower bound for the length
2483           of subject string that is needed to match the pattern is computed. This
2484           does not mean that there are any strings of that length that match, but
2485           it  does  guarantee that no shorter strings match. The value is used to
2486           avoid wasting time by trying to match strings that are shorter than the
2487           lower  bound.  You  can find out the value in a calling program via the
2488           pcre_fullinfo() function.
2489    
2490           Studying a pattern is also useful for non-anchored patterns that do not
2491           have  a  single fixed starting character. A bitmap of possible starting
2492           bytes is created. This speeds up finding a position in the  subject  at
2493           which to start matching. (In 16-bit mode, the bitmap is used for 16-bit
2494           values less than 256.  In 32-bit mode, the bitmap is  used  for  32-bit
2495           values less than 256.)
2496    
2497           These  two optimizations apply to both pcre_exec() and pcre_dfa_exec(),
2498           and the information is also used by the JIT  compiler.   The  optimiza-
2499           tions can be disabled by setting the PCRE_NO_START_OPTIMIZE option when
2500           calling pcre_exec() or pcre_dfa_exec(), but if this is done, JIT execu-
2501           tion  is  also disabled. You might want to do this if your pattern con-
2502           tains callouts or (*MARK) and you want to make use of these  facilities
2503           in    cases    where    matching   fails.   See   the   discussion   of
2504           PCRE_NO_START_OPTIMIZE below.
2505    
2506    
2507    LOCALE SUPPORT
2508    
2509           PCRE handles caseless matching, and determines whether  characters  are
2510           letters,  digits, or whatever, by reference to a set of tables, indexed
2511           by character value. When running in UTF-8 mode, this  applies  only  to
2512           characters  with  codes  less than 128. By default, higher-valued codes
2513           never match escapes such as \w or \d, but they can be tested with \p if
2514           PCRE  is  built with Unicode character property support. Alternatively,
2515           the PCRE_UCP option can be set at compile  time;  this  causes  \w  and
2516           friends to use Unicode property support instead of built-in tables. The
2517           use of locales with Unicode is discouraged. If you are handling charac-
2518           ters  with codes greater than 128, you should either use UTF-8 and Uni-
2519           code, or use locales, but not try to mix the two.
2520    
2521           PCRE contains an internal set of tables that are used  when  the  final
2522           argument  of  pcre_compile()  is  NULL.  These  are sufficient for many
2523           applications.  Normally, the internal tables recognize only ASCII char-
2524           acters. However, when PCRE is built, it is possible to cause the inter-
2525           nal tables to be rebuilt in the default "C" locale of the local system,
2526           which may cause them to be different.
2527    
2528           The  internal tables can always be overridden by tables supplied by the
2529           application that calls PCRE. These may be created in a different locale
2530           from  the  default.  As more and more applications change to using Uni-
2531           code, the need for this locale support is expected to die away.
2532    
2533           External tables are built by calling  the  pcre_maketables()  function,
2534           which  has no arguments, in the relevant locale. The result can then be
2535           passed to pcre_compile() or pcre_exec()  as  often  as  necessary.  For
2536           example,  to  build  and use tables that are appropriate for the French
2537           locale (where accented characters with  values  greater  than  128  are
2538           treated as letters), the following code could be used:
2539    
2540             setlocale(LC_CTYPE, "fr_FR");
2541             tables = pcre_maketables();
2542             re = pcre_compile(..., tables);
2543    
2544           The  locale  name "fr_FR" is used on Linux and other Unix-like systems;
2545           if you are using Windows, the name for the French locale is "french".
2546    
2547           When pcre_maketables() runs, the tables are built  in  memory  that  is
2548           obtained  via  pcre_malloc. It is the caller's responsibility to ensure
2549           that the memory containing the tables remains available for as long  as
2550           it is needed.
2551    
2552           The pointer that is passed to pcre_compile() is saved with the compiled
2553           pattern, and the same tables are used via this pointer by  pcre_study()
2554           and normally also by pcre_exec(). Thus, by default, for any single pat-
2555           tern, compilation, studying and matching all happen in the same locale,
2556           but different patterns can be compiled in different locales.
2557    
2558           It  is  possible to pass a table pointer or NULL (indicating the use of
2559           the internal tables) to pcre_exec(). Although  not  intended  for  this
2560           purpose,  this facility could be used to match a pattern in a different
2561           locale from the one in which it was compiled. Passing table pointers at
2562           run time is discussed below in the section on matching a pattern.
2563    
2564    
2565    INFORMATION ABOUT A PATTERN
2566    
2567           int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
2568                int what, void *where);
2569    
2570           The  pcre_fullinfo() function returns information about a compiled pat-
2571           tern. It replaces the pcre_info() function, which was removed from  the
2572           library at version 8.30, after more than 10 years of obsolescence.
2573    
2574           The  first  argument  for  pcre_fullinfo() is a pointer to the compiled
2575           pattern. The second argument is the result of pcre_study(), or NULL  if
2576           the  pattern  was not studied. The third argument specifies which piece
2577           of information is required, and the fourth argument is a pointer  to  a
2578           variable  to  receive  the  data. The yield of the function is zero for
2579           success, or one of the following negative numbers:
2580    
2581             PCRE_ERROR_NULL           the argument code was NULL
2582                                       the argument where was NULL
2583             PCRE_ERROR_BADMAGIC       the "magic number" was not found
2584             PCRE_ERROR_BADENDIANNESS  the pattern was compiled with different
2585                                       endianness
2586             PCRE_ERROR_BADOPTION      the value of what was invalid
2587    
2588           The "magic number" is placed at the start of each compiled  pattern  as
2589           an  simple check against passing an arbitrary memory pointer. The endi-
2590           anness error can occur if a compiled pattern is saved and reloaded on a
2591           different  host.  Here  is a typical call of pcre_fullinfo(), to obtain
2592           the length of the compiled pattern:
2593    
2594             int rc;
2595             size_t length;
2596             rc = pcre_fullinfo(
2597               re,               /* result of pcre_compile() */
2598               sd,               /* result of pcre_study(), or NULL */
2599               PCRE_INFO_SIZE,   /* what is required */
2600               &length);         /* where to put the data */
2601    
2602           The possible values for the third argument are defined in  pcre.h,  and
2603           are as follows:
2604    
2605             PCRE_INFO_BACKREFMAX
2606    
2607           Return  the  number  of  the highest back reference in the pattern. The
2608           fourth argument should point to an int variable. Zero  is  returned  if
2609           there are no back references.
2610    
2611             PCRE_INFO_CAPTURECOUNT
2612    
2613           Return  the  number of capturing subpatterns in the pattern. The fourth
2614           argument should point to an int variable.
2615    
2616             PCRE_INFO_DEFAULT_TABLES
2617    
2618           Return a pointer to the internal default character tables within  PCRE.
2619           The  fourth  argument should point to an unsigned char * variable. This
2620           information call is provided for internal use by the pcre_study() func-
2621           tion.  External  callers  can  cause PCRE to use its internal tables by
2622           passing a NULL table pointer.
2623    
2624             PCRE_INFO_FIRSTBYTE
2625    
2626           Return information about the first data unit of any matched string, for
2627           a  non-anchored  pattern.  (The name of this option refers to the 8-bit
2628           library, where data units are bytes.) The fourth argument should  point
2629           to an int variable.
2630    
2631           If  there  is  a  fixed first value, for example, the letter "c" from a
2632           pattern such as (cat|cow|coyote), its value is returned. In  the  8-bit
2633           library,  the  value is always less than 256. In the 16-bit library the
2634           value can be up to 0xffff. In the 32-bit library the value can be up to
2635           0x10ffff.
2636    
2637           If there is no fixed first value, and if either
2638    
2639           (a)  the pattern was compiled with the PCRE_MULTILINE option, and every
2640           branch starts with "^", or
2641    
2642           (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not
2643           set (if it were set, the pattern would be anchored),
2644    
2645           -1  is  returned, indicating that the pattern matches only at the start
2646           of a subject string or after any newline within the  string.  Otherwise
2647           -2 is returned. For anchored patterns, -2 is returned.
2648    
2649           Since  for  the 32-bit library using the non-UTF-32 mode, this function
2650           is unable to return the full 32-bit range of the character, this  value
2651           is    deprecated;   instead   the   PCRE_INFO_FIRSTCHARACTERFLAGS   and
2652           PCRE_INFO_FIRSTCHARACTER values should be used.
2653    
2654             PCRE_INFO_FIRSTTABLE
2655    
2656           If the pattern was studied, and this resulted in the construction of  a
2657           256-bit  table indicating a fixed set of values for the first data unit
2658           in any matching string, a pointer to the table is  returned.  Otherwise
2659           NULL  is returned. The fourth argument should point to an unsigned char
2660           * variable.
2661    
2662             PCRE_INFO_HASCRORLF
2663    
2664           Return 1 if the pattern contains any explicit  matches  for  CR  or  LF
2665           characters,  otherwise  0.  The  fourth argument should point to an int
2666           variable. An explicit match is either a literal CR or LF character,  or
2667           \r or \n.
2668    
2669             PCRE_INFO_JCHANGED
2670    
2671           Return  1  if  the (?J) or (?-J) option setting is used in the pattern,
2672           otherwise 0. The fourth argument should point to an int variable.  (?J)
2673           and (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
2674    
2675             PCRE_INFO_JIT
2676    
2677           Return  1  if  the pattern was studied with one of the JIT options, and
2678           just-in-time compiling was successful. The fourth argument should point
2679           to  an  int variable. A return value of 0 means that JIT support is not
2680           available in this version of PCRE, or that the pattern was not  studied
2681           with  a JIT option, or that the JIT compiler could not handle this par-
2682           ticular pattern. See the pcrejit documentation for details of what  can
2683           and cannot be handled.
2684    
2685             PCRE_INFO_JITSIZE
2686    
2687           If  the  pattern was successfully studied with a JIT option, return the
2688           size of the JIT compiled code, otherwise return zero. The fourth  argu-
2689           ment should point to a size_t variable.
2690    
2691             PCRE_INFO_LASTLITERAL
2692    
2693           Return  the value of the rightmost literal data unit that must exist in
2694           any matched string, other than at its start, if such a value  has  been
2695           recorded. The fourth argument should point to an int variable. If there
2696           is no such value, -1 is returned. For anchored patterns, a last literal
2697           value  is recorded only if it follows something of variable length. For
2698           example, for the pattern /^a\d+z\d+/ the returned value is "z", but for
2699           /^a\dz\d/ the returned value is -1.
2700    
2701           Since  for  the 32-bit library using the non-UTF-32 mode, this function
2702           is unable to return the full 32-bit range of the character, this  value
2703           is    deprecated;    instead    the   PCRE_INFO_REQUIREDCHARFLAGS   and
2704           PCRE_INFO_REQUIREDCHAR values should be used.
2705    
2706             PCRE_INFO_MAXLOOKBEHIND
2707    
2708           Return the number of characters (NB not bytes) in the  longest  lookbe-
2709           hind  assertion  in the pattern. Note that the simple assertions \b and
2710           \B require a one-character lookbehind. This information is useful  when
2711           doing multi-segment matching using the partial matching facilities.
2712    
2713             PCRE_INFO_MINLENGTH
2714    
2715           If  the  pattern  was studied and a minimum length for matching subject
2716           strings was computed, its value is  returned.  Otherwise  the  returned
2717           value  is  -1. The value is a number of characters, which in UTF-8 mode
2718           may be different from the number of bytes. The fourth  argument  should
2719           point  to an int variable. A non-negative value is a lower bound to the
2720           length of any matching string. There may not be  any  strings  of  that
2721           length  that  do actually match, but every string that does match is at
2722           least that long.
2723    
2724             PCRE_INFO_NAMECOUNT
2725             PCRE_INFO_NAMEENTRYSIZE
2726             PCRE_INFO_NAMETABLE
2727    
2728           PCRE supports the use of named as well as numbered capturing  parenthe-
2729           ses.  The names are just an additional way of identifying the parenthe-
2730           ses, which still acquire numbers. Several convenience functions such as
2731           pcre_get_named_substring()  are  provided  for extracting captured sub-
2732           strings by name. It is also possible to extract the data  directly,  by
2733           first  converting  the  name to a number in order to access the correct
2734           pointers in the output vector (described with pcre_exec() below). To do
2735           the  conversion,  you  need  to  use  the  name-to-number map, which is
2736           described by these three values.
2737    
2738           The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT
2739           gives the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size
2740           of each entry; both of these  return  an  int  value.  The  entry  size
2741           depends  on the length of the longest name. PCRE_INFO_NAMETABLE returns
2742           a pointer to the first entry of the table. This is a pointer to char in
2743           the 8-bit library, where the first two bytes of each entry are the num-
2744           ber of the capturing parenthesis, most significant byte first.  In  the
2745           16-bit  library,  the pointer points to 16-bit data units, the first of
2746           which contains the parenthesis number.   In  the  32-bit  library,  the
2747           pointer  points  to  32-bit data units, the first of which contains the
2748           parenthesis number. The rest of the entry is  the  corresponding  name,
2749           zero terminated.
2750    
2751           The  names are in alphabetical order. Duplicate names may appear if (?|
2752           is used to create multiple groups with the same number, as described in
2753           the  section  on  duplicate subpattern numbers in the pcrepattern page.
2754           Duplicate names for subpatterns with different  numbers  are  permitted
2755           only  if  PCRE_DUPNAMES  is  set. In all cases of duplicate names, they
2756           appear in the table in the order in which they were found in  the  pat-
2757           tern.  In  the  absence  of (?| this is the order of increasing number;
2758           when (?| is used this is not necessarily the case because later subpat-
2759           terns may have lower numbers.
2760    
2761           As  a  simple  example of the name/number table, consider the following
2762           pattern after compilation by the 8-bit library (assume PCRE_EXTENDED is
2763           set, so white space - including newlines - is ignored):
2764    
2765             (?<date> (?<year>(\d\d)?\d\d) -
2766             (?<month>\d\d) - (?<day>\d\d) )
2767    
2768           There  are  four  named subpatterns, so the table has four entries, and
2769           each entry in the table is eight bytes long. The table is  as  follows,
2770           with non-printing bytes shows in hexadecimal, and undefined bytes shown
2771           as ??:
2772    
2773             00 01 d  a  t  e  00 ??
2774             00 05 d  a  y  00 ?? ??
2775             00 04 m  o  n  t  h  00
2776             00 02 y  e  a  r  00 ??
2777    
2778           When writing code to extract data  from  named  subpatterns  using  the
2779           name-to-number  map,  remember that the length of the entries is likely
2780           to be different for each compiled pattern.
2781    
2782             PCRE_INFO_OKPARTIAL
2783    
2784           Return 1  if  the  pattern  can  be  used  for  partial  matching  with
2785           pcre_exec(),  otherwise  0.  The fourth argument should point to an int
2786           variable. From  release  8.00,  this  always  returns  1,  because  the
2787           restrictions  that  previously  applied  to  partial matching have been
2788           lifted. The pcrepartial documentation gives details of  partial  match-
2789           ing.
2790    
2791             PCRE_INFO_OPTIONS
2792    
2793           Return  a  copy of the options with which the pattern was compiled. The
2794           fourth argument should point to an unsigned long  int  variable.  These
2795           option bits are those specified in the call to pcre_compile(), modified
2796           by any top-level option settings at the start of the pattern itself. In
2797           other  words,  they are the options that will be in force when matching
2798           starts. For example, if the pattern /(?im)abc(?-i)d/ is  compiled  with
2799           the  PCRE_EXTENDED option, the result is PCRE_CASELESS, PCRE_MULTILINE,
2800           and PCRE_EXTENDED.
2801    
2802           A pattern is automatically anchored by PCRE if  all  of  its  top-level
2803           alternatives begin with one of the following:
2804    
2805             ^     unless PCRE_MULTILINE is set
2806             \A    always
2807             \G    always
2808             .*    if PCRE_DOTALL is set and there are no back
2809                     references to the subpattern in which .* appears
2810    
2811           For such patterns, the PCRE_ANCHORED bit is set in the options returned
2812           by pcre_fullinfo().
2813    
2814             PCRE_INFO_SIZE
2815    
2816           Return the size of the compiled pattern in bytes (for both  libraries).
2817           The  fourth argument should point to a size_t variable. This value does
2818           not include the  size  of  the  pcre  structure  that  is  returned  by
2819           pcre_compile().  The  value that is passed as the argument to pcre_mal-
2820           loc() when pcre_compile() is getting memory in which to place the  com-
2821           piled  data  is  the value returned by this option plus the size of the
2822           pcre structure. Studying a compiled pattern, with or without JIT,  does
2823           not alter the value returned by this option.
2824    
2825             PCRE_INFO_STUDYSIZE
2826    
2827           Return the size in bytes of the data block pointed to by the study_data
2828           field in a pcre_extra block. If pcre_extra is  NULL,  or  there  is  no
2829           study  data,  zero  is  returned. The fourth argument should point to a
2830           size_t variable. The study_data field is set by pcre_study() to  record
2831           information  that  will  speed  up  matching  (see the section entitled
2832           "Studying a pattern" above). The format of the study_data block is pri-
2833           vate,  but  its length is made available via this option so that it can
2834           be  saved  and  restored  (see  the  pcreprecompile  documentation  for
2835           details).
2836    
2837             PCRE_INFO_FIRSTCHARACTERFLAGS
2838    
2839           Return information about the first data unit of any matched string, for
2840           a non-anchored pattern. The fourth argument  should  point  to  an  int
2841           variable.
2842    
2843           If  there  is  a  fixed first value, for example, the letter "c" from a
2844           pattern such as (cat|cow|coyote), 1  is  returned,  and  the  character
2845           value can be retrieved using PCRE_INFO_FIRSTCHARACTER.
2846    
2847           If there is no fixed first value, and if either
2848    
2849           (a)  the pattern was compiled with the PCRE_MULTILINE option, and every
2850           branch starts with "^", or
2851    
2852           (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not
2853           set (if it were set, the pattern would be anchored),
2854    
2855           2 is returned, indicating that the pattern matches only at the start of
2856           a subject string or after any newline within the string. Otherwise 0 is
2857           returned. For anchored patterns, 0 is returned.
2858    
2859             PCRE_INFO_FIRSTCHARACTER
2860    
2861           Return  the  fixed  first character value, if PCRE_INFO_FIRSTCHARACTER-
2862           FLAGS returned 1; otherwise returns 0. The fourth argument should point
2863           to an uint_t variable.
2864    
2865           In  the 8-bit library, the value is always less than 256. In the 16-bit
2866           library the value can be up to 0xffff. In the 32-bit library in  UTF-32
2867           mode  the  value  can  be up to 0x10ffff, and up to 0xffffffff when not
2868           using UTF-32 mode.
2869    
2870           If there is no fixed first value, and if either
2871    
2872           (a) the pattern was compiled with the PCRE_MULTILINE option, and  every
2873           branch starts with "^", or
2874    
2875           (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not
2876           set (if it were set, the pattern would be anchored),
2877    
2878           -1 is returned, indicating that the pattern matches only at  the  start
2879           of  a  subject string or after any newline within the string. Otherwise
2880           -2 is returned. For anchored patterns, -2 is returned.
2881    
2882             PCRE_INFO_REQUIREDCHARFLAGS
2883    
2884           Returns 1 if there is a rightmost literal data unit that must exist  in
2885           any matched string, other than at its start. The fourth argument should
2886           point to an int variable. If there is no such value, 0 is returned.  If
2887           returning  1,  the  character  value  itself  can  be  retrieved  using
2888           PCRE_INFO_REQUIREDCHAR.
2889    
2890           For anchored patterns, a last literal value is recorded only if it fol-
2891           lows  something  of  variable  length.  For  example,  for  the pattern
2892           /^a\d+z\d+/  the   returned   value   1   (with   "z"   returned   from
2893           PCRE_INFO_REQUIREDCHAR), but for /^a\dz\d/ the returned value is 0.
2894    
2895             PCRE_INFO_REQUIREDCHAR
2896    
2897           Return  the value of the rightmost literal data unit that must exist in
2898           any matched string, other than at its start, if such a value  has  been
2899           recorded.  The fourth argument should point to an uint32_t variable. If
2900           there is no such value, 0 is returned.
2901    
2902    
2903    REFERENCE COUNTS
2904    
2905           int pcre_refcount(pcre *code, int adjust);
2906    
2907           The pcre_refcount() function is used to maintain a reference  count  in
2908           the data block that contains a compiled pattern. It is provided for the
2909           benefit of applications that  operate  in  an  object-oriented  manner,
2910           where different parts of the application may be using the same compiled
2911           pattern, but you want to free the block when they are all done.
2912    
2913           When a pattern is compiled, the reference count field is initialized to
2914           zero.   It is changed only by calling this function, whose action is to
2915           add the adjust value (which may be positive or  negative)  to  it.  The
2916           yield of the function is the new value. However, the value of the count
2917           is constrained to lie between 0 and 65535, inclusive. If the new  value
2918           is outside these limits, it is forced to the appropriate limit value.
2919    
2920           Except  when it is zero, the reference count is not correctly preserved
2921           if a pattern is compiled on one host and then  transferred  to  a  host
2922           whose byte-order is different. (This seems a highly unlikely scenario.)
2923    
2924    
2925    MATCHING A PATTERN: THE TRADITIONAL FUNCTION
2926    
2927           int pcre_exec(const pcre *code, const pcre_extra *extra,
2928                const char *subject, int length, int startoffset,
2929                int options, int *ovector, int ovecsize);
2930    
2931           The  function pcre_exec() is called to match a subject string against a
2932           compiled pattern, which is passed in the code argument. If the  pattern
2933           was  studied,  the  result  of  the study should be passed in the extra
2934           argument. You can call pcre_exec() with the same code and  extra  argu-
2935           ments  as  many  times as you like, in order to match different subject
2936           strings with the same pattern.
2937    
2938           This function is the main matching facility  of  the  library,  and  it
2939           operates  in  a  Perl-like  manner. For specialist use there is also an
2940           alternative matching function, which is described below in the  section
2941           about the pcre_dfa_exec() function.
2942    
2943           In  most applications, the pattern will have been compiled (and option-
2944           ally studied) in the same process that calls pcre_exec().  However,  it
2945           is possible to save compiled patterns and study data, and then use them
2946           later in different processes, possibly even on different hosts.  For  a
2947           discussion about this, see the pcreprecompile documentation.
2948    
2949           Here is an example of a simple call to pcre_exec():
2950    
2951             int rc;
2952             int ovector[30];
2953             rc = pcre_exec(
2954               re,             /* result of pcre_compile() */
2955               NULL,           /* we didn't study the pattern */
2956               "some string",  /* the subject string */
2957               11,             /* the length of the subject string */
2958               0,              /* start at offset 0 in the subject */
2959               0,              /* default options */
2960               ovector,        /* vector of integers for substring information */
2961               30);            /* number of elements (NOT size in bytes) */
2962    
2963       Extra data for pcre_exec()
2964    
2965           If  the  extra argument is not NULL, it must point to a pcre_extra data
2966           block. The pcre_study() function returns such a block (when it  doesn't
2967           return  NULL), but you can also create one for yourself, and pass addi-
2968           tional information in it. The pcre_extra block contains  the  following
2969           fields (not necessarily in this order):
2970    
2971             unsigned long int flags;
2972             void *study_data;
2973             void *executable_jit;
2974             unsigned long int match_limit;
2975             unsigned long int match_limit_recursion;
2976             void *callout_data;
2977             const unsigned char *tables;
2978             unsigned char **mark;
2979    
2980           In  the  16-bit  version  of  this  structure,  the mark field has type
2981           "PCRE_UCHAR16 **".
2982    
2983           In the 32-bit version of  this  structure,  the  mark  field  has  type
2984           "PCRE_UCHAR32 **".
2985    
2986           The  flags  field is used to specify which of the other fields are set.
2987           The flag bits are:
2988    
2989             PCRE_EXTRA_CALLOUT_DATA
2990             PCRE_EXTRA_EXECUTABLE_JIT
2991             PCRE_EXTRA_MARK
2992             PCRE_EXTRA_MATCH_LIMIT
2993             PCRE_EXTRA_MATCH_LIMIT_RECURSION
2994             PCRE_EXTRA_STUDY_DATA
2995             PCRE_EXTRA_TABLES
2996    
2997           Other flag bits should be set to zero. The study_data field  and  some-
2998           times  the executable_jit field are set in the pcre_extra block that is
2999           returned by pcre_study(), together with the appropriate flag bits.  You
3000           should  not set these yourself, but you may add to the block by setting
3001           other fields and their corresponding flag bits.
3002    
3003           The match_limit field provides a means of preventing PCRE from using up
3004           a  vast amount of resources when running patterns that are not going to
3005           match, but which have a very large number  of  possibilities  in  their
3006           search  trees. The classic example is a pattern that uses nested unlim-
3007           ited repeats.
3008    
3009           Internally, pcre_exec() uses a function called match(), which it  calls
3010           repeatedly  (sometimes  recursively).  The  limit set by match_limit is
3011           imposed on the number of times this function is called during a  match,
3012           which  has  the  effect of limiting the amount of backtracking that can
3013           take place. For patterns that are not anchored, the count restarts from
3014           zero for each position in the subject string.
3015    
3016           When pcre_exec() is called with a pattern that was successfully studied
3017           with a JIT option, the way that the matching is  executed  is  entirely
3018           different.  However, there is still the possibility of runaway matching
3019           that goes on for a very long time, and so the match_limit value is also
3020           used in this case (but in a different way) to limit how long the match-
3021           ing can continue.
3022    
3023           The default value for the limit can be set  when  PCRE  is  built;  the
3024           default  default  is 10 million, which handles all but the most extreme
3025           cases. You can override the default  by  suppling  pcre_exec()  with  a
3026           pcre_extra     block    in    which    match_limit    is    set,    and
3027           PCRE_EXTRA_MATCH_LIMIT is set in the  flags  field.  If  the  limit  is
3028           exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3029    
3030           The  match_limit_recursion field is similar to match_limit, but instead
3031           of limiting the total number of times that match() is called, it limits
3032           the  depth  of  recursion. The recursion depth is a smaller number than
3033           the total number of calls, because not all calls to match() are  recur-
3034           sive.  This limit is of use only if it is set smaller than match_limit.
3035    
3036           Limiting  the  recursion  depth limits the amount of machine stack that
3037           can be used, or, when PCRE has been compiled to use memory on the  heap
3038           instead  of the stack, the amount of heap memory that can be used. This
3039           limit is not relevant, and is ignored, when matching is done using  JIT
3040           compiled code.
3041    
3042           The  default  value  for  match_limit_recursion can be set when PCRE is
3043           built; the default default  is  the  same  value  as  the  default  for
3044           match_limit.  You can override the default by suppling pcre_exec() with
3045           a  pcre_extra  block  in  which  match_limit_recursion  is   set,   and
3046           PCRE_EXTRA_MATCH_LIMIT_RECURSION  is  set  in  the  flags field. If the
3047           limit is exceeded, pcre_exec() returns PCRE_ERROR_RECURSIONLIMIT.
3048    
3049           The callout_data field is used in conjunction with the  "callout"  fea-
3050           ture, and is described in the pcrecallout documentation.
3051    
3052           The  tables  field  is  used  to  pass  a  character  tables pointer to
3053           pcre_exec(); this overrides the value that is stored with the  compiled
3054           pattern.  A  non-NULL value is stored with the compiled pattern only if
3055           custom tables were supplied to pcre_compile() via  its  tableptr  argu-
3056           ment.  If NULL is passed to pcre_exec() using this mechanism, it forces
3057           PCRE's internal tables to be used. This facility is  helpful  when  re-
3058           using  patterns  that  have been saved after compiling with an external
3059           set of tables, because the external tables  might  be  at  a  different
3060           address  when  pcre_exec() is called. See the pcreprecompile documenta-
3061           tion for a discussion of saving compiled patterns for later use.
3062    
3063           If PCRE_EXTRA_MARK is set in the flags field, the mark  field  must  be
3064           set  to point to a suitable variable. If the pattern contains any back-
3065           tracking control verbs such as (*MARK:NAME), and the execution ends  up
3066           with  a  name  to  pass back, a pointer to the name string (zero termi-
3067           nated) is placed in the variable pointed to  by  the  mark  field.  The
3068           names  are  within  the  compiled pattern; if you wish to retain such a
3069           name you must copy it before freeing the memory of a compiled  pattern.
3070           If  there  is no name to pass back, the variable pointed to by the mark
3071           field is set to NULL. For details of the  backtracking  control  verbs,
3072           see the section entitled "Backtracking control" in the pcrepattern doc-
3073           umentation.
3074    
3075       Option bits for pcre_exec()
3076    
3077           The unused bits of the options argument for pcre_exec() must  be  zero.
3078           The  only  bits  that  may  be set are PCRE_ANCHORED, PCRE_NEWLINE_xxx,
3079           PCRE_NOTBOL,   PCRE_NOTEOL,    PCRE_NOTEMPTY,    PCRE_NOTEMPTY_ATSTART,
3080           PCRE_NO_START_OPTIMIZE,   PCRE_NO_UTF8_CHECK,   PCRE_PARTIAL_HARD,  and
3081           PCRE_PARTIAL_SOFT.
3082    
3083           If the pattern was successfully studied with one  of  the  just-in-time
3084           (JIT) compile options, the only supported options for JIT execution are
3085           PCRE_NO_UTF8_CHECK,    PCRE_NOTBOL,     PCRE_NOTEOL,     PCRE_NOTEMPTY,
3086           PCRE_NOTEMPTY_ATSTART,  PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
3087           unsupported option is used, JIT execution is disabled  and  the  normal
3088           interpretive code in pcre_exec() is run.
3089    
3090             PCRE_ANCHORED
3091    
3092           The  PCRE_ANCHORED  option  limits pcre_exec() to matching at the first
3093           matching position. If a pattern was  compiled  with  PCRE_ANCHORED,  or
3094           turned  out to be anchored by virtue of its contents, it cannot be made
3095           unachored at matching time.
3096    
3097             PCRE_BSR_ANYCRLF
3098             PCRE_BSR_UNICODE
3099    
3100           These options (which are mutually exclusive) control what the \R escape
3101           sequence  matches.  The choice is either to match only CR, LF, or CRLF,
3102           or to match any Unicode newline sequence. These  options  override  the
3103           choice that was made or defaulted when the pattern was compiled.
3104    
3105             PCRE_NEWLINE_CR
3106             PCRE_NEWLINE_LF
3107             PCRE_NEWLINE_CRLF
3108             PCRE_NEWLINE_ANYCRLF
3109             PCRE_NEWLINE_ANY
3110    
3111           These  options  override  the  newline  definition  that  was chosen or
3112           defaulted when the pattern was compiled. For details, see the  descrip-
3113           tion  of  pcre_compile()  above.  During  matching,  the newline choice
3114           affects the behaviour of the dot, circumflex,  and  dollar  metacharac-
3115           ters.  It may also alter the way the match position is advanced after a
3116           match failure for an unanchored pattern.
3117    
3118           When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF,  or  PCRE_NEWLINE_ANY  is
3119           set,  and a match attempt for an unanchored pattern fails when the cur-
3120           rent position is at a  CRLF  sequence,  and  the  pattern  contains  no
3121           explicit  matches  for  CR  or  LF  characters,  the  match position is
3122           advanced by two characters instead of one, in other words, to after the
3123           CRLF.
3124    
3125           The above rule is a compromise that makes the most common cases work as
3126           expected. For example, if the  pattern  is  .+A  (and  the  PCRE_DOTALL
3127           option is not set), it does not match the string "\r\nA" because, after
3128           failing at the start, it skips both the CR and the LF before  retrying.
3129           However,  the  pattern  [\r\n]A does match that string, because it con-
3130           tains an explicit CR or LF reference, and so advances only by one char-
3131           acter after the first failure.
3132    
3133           An explicit match for CR of LF is either a literal appearance of one of
3134           those characters, or one of the \r or  \n  escape  sequences.  Implicit
3135           matches  such  as [^X] do not count, nor does \s (which includes CR and
3136           LF in the characters that it matches).
3137    
3138           Notwithstanding the above, anomalous effects may still occur when  CRLF
3139           is a valid newline sequence and explicit \r or \n escapes appear in the
3140           pattern.
3141    
3142             PCRE_NOTBOL
3143    
3144           This option specifies that first character of the subject string is not
3145           the  beginning  of  a  line, so the circumflex metacharacter should not
3146           match before it. Setting this without PCRE_MULTILINE (at compile  time)
3147           causes  circumflex  never to match. This option affects only the behav-
3148           iour of the circumflex metacharacter. It does not affect \A.
3149    
3150             PCRE_NOTEOL
3151    
3152           This option specifies that the end of the subject string is not the end
3153           of  a line, so the dollar metacharacter should not match it nor (except
3154           in multiline mode) a newline immediately before it. Setting this  with-
3155           out PCRE_MULTILINE (at compile time) causes dollar never to match. This
3156           option affects only the behaviour of the dollar metacharacter. It  does
3157           not affect \Z or \z.
3158    
3159             PCRE_NOTEMPTY
3160    
3161           An empty string is not considered to be a valid match if this option is
3162           set. If there are alternatives in the pattern, they are tried.  If  all
3163           the  alternatives  match  the empty string, the entire match fails. For
3164           example, if the pattern
3165    
3166             a?b?
3167    
3168           is applied to a string not beginning with "a" or  "b",  it  matches  an
3169           empty  string at the start of the subject. With PCRE_NOTEMPTY set, this
3170           match is not valid, so PCRE searches further into the string for occur-
3171           rences of "a" or "b".
3172    
3173             PCRE_NOTEMPTY_ATSTART
3174    
3175           This  is  like PCRE_NOTEMPTY, except that an empty string match that is
3176           not at the start of  the  subject  is  permitted.  If  the  pattern  is
3177           anchored, such a match can occur only if the pattern contains \K.
3178    
3179           Perl     has    no    direct    equivalent    of    PCRE_NOTEMPTY    or
3180           PCRE_NOTEMPTY_ATSTART, but it does make a special  case  of  a  pattern
3181           match  of  the empty string within its split() function, and when using
3182           the /g modifier. It is  possible  to  emulate  Perl's  behaviour  after
3183           matching a null string by first trying the match again at the same off-
3184           set with PCRE_NOTEMPTY_ATSTART and  PCRE_ANCHORED,  and  then  if  that
3185           fails, by advancing the starting offset (see below) and trying an ordi-
3186           nary match again. There is some code that demonstrates how to  do  this
3187           in  the  pcredemo sample program. In the most general case, you have to
3188           check to see if the newline convention recognizes CRLF  as  a  newline,
3189           and  if so, and the current character is CR followed by LF, advance the
3190           starting offset by two characters instead of one.
3191    
3192             PCRE_NO_START_OPTIMIZE
3193    
3194           There are a number of optimizations that pcre_exec() uses at the  start
3195           of  a  match,  in  order to speed up the process. For example, if it is
3196           known that an unanchored match must start with a specific character, it
3197           searches  the  subject  for that character, and fails immediately if it
3198           cannot find it, without actually running the  main  matching  function.
3199           This means that a special item such as (*COMMIT) at the start of a pat-
3200           tern is not considered until after a suitable starting  point  for  the
3201           match  has been found. When callouts or (*MARK) items are in use, these
3202           "start-up" optimizations can cause them to be skipped if the pattern is
3203           never  actually  used.  The start-up optimizations are in effect a pre-
3204           scan of the subject that takes place before the pattern is run.
3205    
3206           The PCRE_NO_START_OPTIMIZE option disables the start-up  optimizations,
3207           possibly  causing  performance  to  suffer,  but ensuring that in cases
3208           where the result is "no match", the callouts do occur, and  that  items
3209           such as (*COMMIT) and (*MARK) are considered at every possible starting
3210           position in the subject string. If  PCRE_NO_START_OPTIMIZE  is  set  at
3211           compile  time,  it  cannot  be  unset  at  matching  time.  The  use of
3212           PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set, matching
3213           is always done using interpretively.
3214    
3215           Setting  PCRE_NO_START_OPTIMIZE  can  change  the outcome of a matching
3216           operation.  Consider the pattern
3217    
3218             (*COMMIT)ABC
3219    
3220           When this is compiled, PCRE records the fact that a  match  must  start
3221           with  the  character  "A".  Suppose the subject string is "DEFABC". The
3222           start-up optimization scans along the subject, finds "A" and  runs  the
3223           first  match attempt from there. The (*COMMIT) item means that the pat-
3224           tern must match the current starting position, which in this  case,  it
3225           does.  However,  if  the  same match is run with PCRE_NO_START_OPTIMIZE
3226           set, the initial scan along the subject string  does  not  happen.  The
3227           first  match  attempt  is  run  starting  from "D" and when this fails,
3228           (*COMMIT) prevents any further matches  being  tried,  so  the  overall
3229           result  is  "no  match". If the pattern is studied, more start-up opti-
3230           mizations may be used. For example, a minimum length  for  the  subject
3231           may be recorded. Consider the pattern
3232    
3233             (*MARK:A)(X|Y)
3234    
3235           The  minimum  length  for  a  match is one character. If the subject is
3236           "ABC", there will be attempts to  match  "ABC",  "BC",  "C",  and  then
3237           finally  an empty string.  If the pattern is studied, the final attempt
3238           does not take place, because PCRE knows that the subject is too  short,
3239           and  so  the  (*MARK) is never encountered.  In this case, studying the
3240           pattern does not affect the overall match result, which  is  still  "no
3241           match", but it does affect the auxiliary information that is returned.
3242    
3243             PCRE_NO_UTF8_CHECK
3244    
3245           When PCRE_UTF8 is set at compile time, the validity of the subject as a
3246           UTF-8 string is automatically checked when pcre_exec() is  subsequently
3247           called.  The entire string is checked before any other processing takes
3248           place. The value of startoffset is  also  checked  to  ensure  that  it
3249           points  to  the start of a UTF-8 character. There is a discussion about
3250           the validity of UTF-8 strings in the pcreunicode page.  If  an  invalid
3251           sequence   of   bytes   is   found,   pcre_exec()   returns  the  error
3252           PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
3253           truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In
3254           both cases, information about the precise nature of the error may  also
3255           be  returned (see the descriptions of these errors in the section enti-
3256           tled Error return values from pcre_exec() below).  If startoffset  con-
3257           tains a value that does not point to the start of a UTF-8 character (or
3258           to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is returned.
3259    
3260           If you already know that your subject is valid, and you  want  to  skip
3261           these    checks    for   performance   reasons,   you   can   set   the
3262           PCRE_NO_UTF8_CHECK option when calling pcre_exec(). You might  want  to
3263           do  this  for the second and subsequent calls to pcre_exec() if you are
3264           making repeated calls to find all  the  matches  in  a  single  subject
3265           string.  However,  you  should  be  sure  that the value of startoffset
3266           points to the start of a character (or the end of  the  subject).  When
3267           PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid string as a
3268           subject or an invalid value of startoffset is undefined.  Your  program
3269           may crash.
3270    
3271             PCRE_PARTIAL_HARD
3272             PCRE_PARTIAL_SOFT
3273    
3274           These  options turn on the partial matching feature. For backwards com-
3275           patibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A  partial
3276           match  occurs if the end of the subject string is reached successfully,
3277           but there are not enough subject characters to complete the  match.  If
3278           this happens when PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set,
3279           matching continues by testing any remaining alternatives.  Only  if  no
3280           complete  match  can be found is PCRE_ERROR_PARTIAL returned instead of
3281           PCRE_ERROR_NOMATCH. In other words,  PCRE_PARTIAL_SOFT  says  that  the
3282           caller  is  prepared to handle a partial match, but only if no complete
3283           match can be found.
3284    
3285           If PCRE_PARTIAL_HARD is set, it overrides  PCRE_PARTIAL_SOFT.  In  this
3286           case,  if  a  partial  match  is found, pcre_exec() immediately returns
3287           PCRE_ERROR_PARTIAL, without  considering  any  other  alternatives.  In
3288           other  words, when PCRE_PARTIAL_HARD is set, a partial match is consid-
3289           ered to be more important that an alternative complete match.
3290    
3291           In both cases, the portion of the string that was  inspected  when  the
3292           partial match was found is set as the first matching string. There is a
3293           more detailed discussion of partial and  multi-segment  matching,  with
3294           examples, in the pcrepartial documentation.
3295    
3296       The string to be matched by pcre_exec()
3297    
3298           The  subject string is passed to pcre_exec() as a pointer in subject, a
3299           length in bytes in length, and a starting byte offset  in  startoffset.
3300           If  this  is  negative  or  greater  than  the  length  of the subject,
3301           pcre_exec() returns PCRE_ERROR_BADOFFSET. When the starting  offset  is
3302           zero,  the  search  for a match starts at the beginning of the subject,
3303           and this is by far the most common case. In UTF-8 mode, the byte offset
3304           must  point  to  the start of a UTF-8 character (or the end of the sub-
3305           ject). Unlike the pattern string, the subject may contain  binary  zero
3306           bytes.
3307    
3308           A  non-zero  starting offset is useful when searching for another match
3309           in the same subject by calling pcre_exec() again after a previous  suc-
3310           cess.   Setting  startoffset differs from just passing over a shortened
3311           string and setting PCRE_NOTBOL in the case of  a  pattern  that  begins
3312           with any kind of lookbehind. For example, consider the pattern
3313    
3314             \Biss\B
3315    
3316           which  finds  occurrences  of "iss" in the middle of words. (\B matches
3317           only if the current position in the subject is not  a  word  boundary.)
3318           When  applied  to the string "Mississipi" the first call to pcre_exec()
3319           finds the first occurrence. If pcre_exec() is called  again  with  just
3320           the  remainder  of  the  subject,  namely  "issipi", it does not match,
3321           because \B is always false at the start of the subject, which is deemed
3322           to  be  a  word  boundary. However, if pcre_exec() is passed the entire
3323           string again, but with startoffset set to 4, it finds the second occur-
3324           rence  of "iss" because it is able to look behind the starting point to
3325           discover that it is preceded by a letter.
3326    
3327           Finding all the matches in a subject is tricky  when  the  pattern  can
3328           match an empty string. It is possible to emulate Perl's /g behaviour by
3329           first  trying  the  match  again  at  the   same   offset,   with   the
3330           PCRE_NOTEMPTY_ATSTART  and  PCRE_ANCHORED  options,  and  then  if that
3331           fails, advancing the starting  offset  and  trying  an  ordinary  match
3332           again. There is some code that demonstrates how to do this in the pcre-
3333           demo sample program. In the most general case, you have to check to see
3334           if  the newline convention recognizes CRLF as a newline, and if so, and
3335           the current character is CR followed by LF, advance the starting offset
3336           by two characters instead of one.
3337    
3338           If  a  non-zero starting offset is passed when the pattern is anchored,
3339           one attempt to match at the given offset is made. This can only succeed
3340           if  the  pattern  does  not require the match to be at the start of the
3341           subject.
3342    
3343       How pcre_exec() returns captured substrings
3344    
3345           In general, a pattern matches a certain portion of the subject, and  in
3346           addition,  further  substrings  from  the  subject may be picked out by
3347           parts of the pattern. Following the usage  in  Jeffrey  Friedl's  book,
3348           this  is  called "capturing" in what follows, and the phrase "capturing
3349           subpattern" is used for a fragment of a pattern that picks out  a  sub-
3350           string.  PCRE  supports several other kinds of parenthesized subpattern
3351           that do not cause substrings to be captured.
3352    
3353           Captured substrings are returned to the caller via a vector of integers
3354           whose  address is passed in ovector. The number of elements in the vec-
3355           tor is passed in ovecsize, which must be a non-negative  number.  Note:
3356           this argument is NOT the size of ovector in bytes.
3357    
3358           The  first  two-thirds of the vector is used to pass back captured sub-
3359           strings, each substring using a pair of integers. The  remaining  third
3360           of  the  vector is used as workspace by pcre_exec() while matching cap-
3361           turing subpatterns, and is not available for passing back  information.
3362           The  number passed in ovecsize should always be a multiple of three. If
3363           it is not, it is rounded down.
3364    
3365           When a match is successful, information about  captured  substrings  is
3366           returned  in  pairs  of integers, starting at the beginning of ovector,
3367           and continuing up to two-thirds of its length at the  most.  The  first
3368           element  of  each pair is set to the byte offset of the first character
3369           in a substring, and the second is set to the byte offset of  the  first
3370           character  after  the end of a substring. Note: these values are always
3371           byte offsets, even in UTF-8 mode. They are not character counts.
3372    
3373           The first pair of integers, ovector[0]  and  ovector[1],  identify  the
3374           portion  of  the subject string matched by the entire pattern. The next
3375           pair is used for the first capturing subpattern, and so on.  The  value
3376           returned by pcre_exec() is one more than the highest numbered pair that
3377           has been set.  For example, if two substrings have been  captured,  the
3378           returned  value is 3. If there are no capturing subpatterns, the return
3379           value from a successful match is 1, indicating that just the first pair
3380           of offsets has been set.
3381    
3382           If a capturing subpattern is matched repeatedly, it is the last portion
3383           of the string that it matched that is returned.
3384    
3385           If the vector is too small to hold all the captured substring  offsets,
3386           it is used as far as possible (up to two-thirds of its length), and the
3387           function returns a value of zero. If neither the actual string  matched
3388           nor  any captured substrings are of interest, pcre_exec() may be called
3389           with ovector passed as NULL and ovecsize as zero. However, if the  pat-
3390           tern  contains  back  references  and  the ovector is not big enough to
3391           remember the related substrings, PCRE has to get additional memory  for
3392           use  during matching. Thus it is usually advisable to supply an ovector
3393           of reasonable size.
3394    
3395           There are some cases where zero is returned  (indicating  vector  over-
3396           flow)  when  in fact the vector is exactly the right size for the final
3397           match. For example, consider the pattern
3398    
3399             (a)(?:(b)c|bd)
3400    
3401           If a vector of 6 elements (allowing for only 1 captured  substring)  is
3402           given with subject string "abd", pcre_exec() will try to set the second
3403           captured string, thereby recording a vector overflow, before failing to
3404           match  "c"  and  backing  up  to  try  the second alternative. The zero
3405           return, however, does correctly indicate that  the  maximum  number  of
3406           slots (namely 2) have been filled. In similar cases where there is tem-
3407           porary overflow, but the final number of used slots  is  actually  less
3408           than the maximum, a non-zero value is returned.
3409    
3410           The pcre_fullinfo() function can be used to find out how many capturing
3411           subpatterns there are in a compiled  pattern.  The  smallest  size  for
3412           ovector  that  will allow for n captured substrings, in addition to the