1: .TH PCREAPI 3 "12 May 2013" "PCRE 8.33"
2: .SH NAME
3: PCRE - Perl-compatible regular expressions
4: .sp
5: .B #include <pcre.h>
6: .
7: .
8: .SH "PCRE NATIVE API BASIC FUNCTIONS"
9: .rs
10: .sp
11: .SM
12: .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
13: .ti +5n
14: .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
15: .ti +5n
16: .B const unsigned char *\fItableptr\fP);
17: .PP
18: .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
19: .ti +5n
20: .B int *\fIerrorcodeptr\fP,
21: .ti +5n
22: .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
23: .ti +5n
24: .B const unsigned char *\fItableptr\fP);
25: .PP
26: .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
27: .ti +5n
28: .B const char **\fIerrptr\fP);
29: .PP
30: .B void pcre_free_study(pcre_extra *\fIextra\fP);
31: .PP
32: .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
33: .ti +5n
34: .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
35: .ti +5n
36: .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
37: .PP
38: .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
39: .ti +5n
40: .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
41: .ti +5n
42: .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
43: .ti +5n
44: .B int *\fIworkspace\fP, int \fIwscount\fP);
45: .
46: .
47: .SH "PCRE NATIVE API STRING EXTRACTION FUNCTIONS"
48: .rs
49: .sp
50: .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
51: .ti +5n
52: .B const char *\fIsubject\fP, int *\fIovector\fP,
53: .ti +5n
54: .B int \fIstringcount\fP, const char *\fIstringname\fP,
55: .ti +5n
56: .B char *\fIbuffer\fP, int \fIbuffersize\fP);
57: .PP
58: .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
59: .ti +5n
60: .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
61: .ti +5n
62: .B int \fIbuffersize\fP);
63: .PP
64: .B int pcre_get_named_substring(const pcre *\fIcode\fP,
65: .ti +5n
66: .B const char *\fIsubject\fP, int *\fIovector\fP,
67: .ti +5n
68: .B int \fIstringcount\fP, const char *\fIstringname\fP,
69: .ti +5n
70: .B const char **\fIstringptr\fP);
71: .PP
72: .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
73: .ti +5n
74: .B const char *\fIname\fP);
75: .PP
76: .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
77: .ti +5n
78: .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
79: .PP
80: .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
81: .ti +5n
82: .B int \fIstringcount\fP, int \fIstringnumber\fP,
83: .ti +5n
84: .B const char **\fIstringptr\fP);
85: .PP
86: .B int pcre_get_substring_list(const char *\fIsubject\fP,
87: .ti +5n
88: .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
89: .PP
90: .B void pcre_free_substring(const char *\fIstringptr\fP);
91: .PP
92: .B void pcre_free_substring_list(const char **\fIstringptr\fP);
93: .
94: .
95: .SH "PCRE NATIVE API AUXILIARY FUNCTIONS"
96: .rs
97: .sp
98: .B int pcre_jit_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
99: .ti +5n
100: .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
101: .ti +5n
102: .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
103: .ti +5n
104: .B pcre_jit_stack *\fIjstack\fP);
105: .PP
106: .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
107: .PP
108: .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
109: .PP
110: .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
111: .ti +5n
112: .B pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);
113: .PP
114: .B const unsigned char *pcre_maketables(void);
115: .PP
116: .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
117: .ti +5n
118: .B int \fIwhat\fP, void *\fIwhere\fP);
119: .PP
120: .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
121: .PP
122: .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
123: .PP
124: .B const char *pcre_version(void);
125: .PP
126: .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
127: .ti +5n
128: .B pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);
129: .
130: .
131: .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
132: .rs
133: .sp
134: .B void *(*pcre_malloc)(size_t);
135: .PP
136: .B void (*pcre_free)(void *);
137: .PP
138: .B void *(*pcre_stack_malloc)(size_t);
139: .PP
140: .B void (*pcre_stack_free)(void *);
141: .PP
142: .B int (*pcre_callout)(pcre_callout_block *);
143: .
144: .
145: .SH "PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES"
146: .rs
147: .sp
148: As well as support for 8-bit character strings, PCRE also supports 16-bit
149: strings (from release 8.30) and 32-bit strings (from release 8.32), by means of
150: two additional libraries. They can be built as well as, or instead of, the
151: 8-bit library. To avoid too much complication, this document describes the
152: 8-bit versions of the functions, with only occasional references to the 16-bit
153: and 32-bit libraries.
154: .P
155: The 16-bit and 32-bit functions operate in the same way as their 8-bit
156: counterparts; they just use different data types for their arguments and
157: results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
158: \fBpcre_\fP. For every option that has UTF8 in its name (for example,
159: PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
160: by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
161: 16-bit and 32-bit option names define the same bit values.
162: .P
163: References to bytes and UTF-8 in this document should be read as references to
164: 16-bit data units and UTF-16 when using the 16-bit library, or 32-bit data
165: units and UTF-32 when using the 32-bit library, unless specified otherwise.
166: More details of the specific differences for the 16-bit and 32-bit libraries
167: are given in the
168: .\" HREF
169: \fBpcre16\fP
170: .\"
171: and
172: .\" HREF
173: \fBpcre32\fP
174: .\"
175: pages.
176: .
177: .
178: .SH "PCRE API OVERVIEW"
179: .rs
180: .sp
181: PCRE has its own native API, which is described in this document. There are
182: also some wrapper functions (for the 8-bit library only) that correspond to the
183: POSIX regular expression API, but they do not give access to all the
184: functionality. They are described in the
185: .\" HREF
186: \fBpcreposix\fP
187: .\"
188: documentation. Both of these APIs define a set of C function calls. A C++
189: wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
190: documented in the
191: .\" HREF
192: \fBpcrecpp\fP
193: .\"
194: page.
195: .P
196: The native API C function prototypes are defined in the header file
197: \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
198: \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
199: command for linking an application that uses PCRE. The header file defines the
200: macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
201: for the library. Applications can use these to include support for different
202: releases of PCRE.
203: .P
204: In a Windows environment, if you want to statically link an application program
205: against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
206: including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
207: \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
208: \fB__declspec(dllimport)\fP, with unwanted results.
209: .P
210: The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
211: and \fBpcre_exec()\fP are used for compiling and matching regular expressions
212: in a Perl-compatible manner. A sample program that demonstrates the simplest
213: way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
214: source distribution. A listing of this program is given in the
215: .\" HREF
216: \fBpcredemo\fP
217: .\"
218: documentation, and the
219: .\" HREF
220: \fBpcresample\fP
221: .\"
222: documentation describes how to compile and run it.
223: .P
224: Just-in-time compiler support is an optional feature of PCRE that can be built
225: in appropriate hardware environments. It greatly speeds up the matching
226: performance of many patterns. Simple programs can easily request that it be
227: used if available, by setting an option that is ignored when it is not
228: relevant. More complicated programs might need to make use of the functions
229: \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
230: \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
231: .P
232: From release 8.32 there is also a direct interface for JIT execution, which
233: gives improved performance. The JIT-specific functions are discussed in the
234: .\" HREF
235: \fBpcrejit\fP
236: .\"
237: documentation.
238: .P
239: A second matching function, \fBpcre_dfa_exec()\fP, which is not
240: Perl-compatible, is also provided. This uses a different algorithm for the
241: matching. The alternative algorithm finds all possible matches (at a given
242: point in the subject), and scans the subject just once (unless there are
243: lookbehind assertions). However, this algorithm does not return captured
244: substrings. A description of the two matching algorithms and their advantages
245: and disadvantages is given in the
246: .\" HREF
247: \fBpcrematching\fP
248: .\"
249: documentation.
250: .P
251: In addition to the main compiling and matching functions, there are convenience
252: functions for extracting captured substrings from a subject string that is
253: matched by \fBpcre_exec()\fP. They are:
254: .sp
255: \fBpcre_copy_substring()\fP
256: \fBpcre_copy_named_substring()\fP
257: \fBpcre_get_substring()\fP
258: \fBpcre_get_named_substring()\fP
259: \fBpcre_get_substring_list()\fP
260: \fBpcre_get_stringnumber()\fP
261: \fBpcre_get_stringtable_entries()\fP
262: .sp
263: \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
264: provided, to free the memory used for extracted strings.
265: .P
266: The function \fBpcre_maketables()\fP is used to build a set of character tables
267: in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
268: or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
269: specialist use. Most commonly, no special tables are passed, in which case
270: internal tables that are generated when PCRE is built are used.
271: .P
272: The function \fBpcre_fullinfo()\fP is used to find out information about a
273: compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
274: string containing the version of PCRE and its date of release.
275: .P
276: The function \fBpcre_refcount()\fP maintains a reference count in a data block
277: containing a compiled pattern. This is provided for the benefit of
278: object-oriented applications.
279: .P
280: The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
281: the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
282: respectively. PCRE calls the memory management functions via these variables,
283: so a calling program can replace them if it wishes to intercept the calls. This
284: should be done before calling any PCRE functions.
285: .P
286: The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
287: indirections to memory management functions. These special functions are used
288: only when PCRE is compiled to use the heap for remembering data, instead of
289: recursive function calls, when running the \fBpcre_exec()\fP function. See the
290: .\" HREF
291: \fBpcrebuild\fP
292: .\"
293: documentation for details of how to do this. It is a non-standard way of
294: building PCRE, for use in environments that have limited stacks. Because of the
295: greater use of memory management, it runs more slowly. Separate functions are
296: provided so that special-purpose external code can be used for this case. When
297: used, these functions are always called in a stack-like manner (last obtained,
298: first freed), and always for memory blocks of the same size. There is a
299: discussion about PCRE's stack usage in the
300: .\" HREF
301: \fBpcrestack\fP
302: .\"
303: documentation.
304: .P
305: The global variable \fBpcre_callout\fP initially contains NULL. It can be set
306: by the caller to a "callout" function, which PCRE will then call at specified
307: points during a matching operation. Details are given in the
308: .\" HREF
309: \fBpcrecallout\fP
310: .\"
311: documentation.
312: .
313: .
314: .\" HTML <a name="newlines"></a>
315: .SH NEWLINES
316: .rs
317: .sp
318: PCRE supports five different conventions for indicating line breaks in
319: strings: a single CR (carriage return) character, a single LF (linefeed)
320: character, the two-character sequence CRLF, any of the three preceding, or any
321: Unicode newline sequence. The Unicode newline sequences are the three just
322: mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
323: U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
324: (paragraph separator, U+2029).
325: .P
326: Each of the first three conventions is used by at least one operating system as
327: its standard newline sequence. When PCRE is built, a default can be specified.
328: The default default is LF, which is the Unix standard. When PCRE is run, the
329: default can be overridden, either when a pattern is compiled, or when it is
330: matched.
331: .P
332: At compile time, the newline convention can be specified by the \fIoptions\fP
333: argument of \fBpcre_compile()\fP, or it can be specified by special text at the
334: start of the pattern itself; this overrides any other settings. See the
335: .\" HREF
336: \fBpcrepattern\fP
337: .\"
338: page for details of the special character sequences.
339: .P
340: In the PCRE documentation the word "newline" is used to mean "the character or
341: pair of characters that indicate a line break". The choice of newline
342: convention affects the handling of the dot, circumflex, and dollar
343: metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
344: recognized line ending sequence, the match position advancement for a
345: non-anchored pattern. There is more detail about this in the
346: .\" HTML <a href="#execoptions">
347: .\" </a>
348: section on \fBpcre_exec()\fP options
349: .\"
350: below.
351: .P
352: The choice of newline convention does not affect the interpretation of
353: the \en or \er escape sequences, nor does it affect what \eR matches, which is
354: controlled in a similar way, but by separate options.
355: .
356: .
357: .SH MULTITHREADING
358: .rs
359: .sp
360: The PCRE functions can be used in multi-threading applications, with the
361: proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
362: \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
363: callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
364: .P
365: The compiled form of a regular expression is not altered during matching, so
366: the same compiled pattern can safely be used by several threads at once.
367: .P
368: If the just-in-time optimization feature is being used, it needs separate
369: memory stack areas for each thread. See the
370: .\" HREF
371: \fBpcrejit\fP
372: .\"
373: documentation for more details.
374: .
375: .
376: .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
377: .rs
378: .sp
379: The compiled form of a regular expression can be saved and re-used at a later
380: time, possibly by a different program, and even on a host other than the one on
381: which it was compiled. Details are given in the
382: .\" HREF
383: \fBpcreprecompile\fP
384: .\"
385: documentation, which includes a description of the
386: \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
387: expression with one version of PCRE for use with a different version is not
388: guaranteed to work and may cause crashes.
389: .
390: .
391: .SH "CHECKING BUILD-TIME OPTIONS"
392: .rs
393: .sp
394: .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
395: .PP
396: The function \fBpcre_config()\fP makes it possible for a PCRE client to
397: discover which optional features have been compiled into the PCRE library. The
398: .\" HREF
399: \fBpcrebuild\fP
400: .\"
401: documentation has more details about these optional features.
402: .P
403: The first argument for \fBpcre_config()\fP is an integer, specifying which
404: information is required; the second argument is a pointer to a variable into
405: which the information is placed. The returned value is zero on success, or the
406: negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
407: not recognized. The following information is available:
408: .sp
409: PCRE_CONFIG_UTF8
410: .sp
411: The output is an integer that is set to one if UTF-8 support is available;
412: otherwise it is set to zero. This value should normally be given to the 8-bit
413: version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
414: or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
415: .sp
416: PCRE_CONFIG_UTF16
417: .sp
418: The output is an integer that is set to one if UTF-16 support is available;
419: otherwise it is set to zero. This value should normally be given to the 16-bit
420: version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
421: or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
422: .sp
423: PCRE_CONFIG_UTF32
424: .sp
425: The output is an integer that is set to one if UTF-32 support is available;
426: otherwise it is set to zero. This value should normally be given to the 32-bit
427: version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
428: or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
429: .sp
430: PCRE_CONFIG_UNICODE_PROPERTIES
431: .sp
432: The output is an integer that is set to one if support for Unicode character
433: properties is available; otherwise it is set to zero.
434: .sp
435: PCRE_CONFIG_JIT
436: .sp
437: The output is an integer that is set to one if support for just-in-time
438: compiling is available; otherwise it is set to zero.
439: .sp
440: PCRE_CONFIG_JITTARGET
441: .sp
442: The output is a pointer to a zero-terminated "const char *" string. If JIT
443: support is available, the string contains the name of the architecture for
444: which the JIT compiler is configured, for example "x86 32bit (little endian +
445: unaligned)". If JIT support is not available, the result is NULL.
446: .sp
447: PCRE_CONFIG_NEWLINE
448: .sp
449: The output is an integer whose value specifies the default character sequence
450: that is recognized as meaning "newline". The values that are supported in
451: ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
452: ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
453: same values. However, the value for LF is normally 21, though some EBCDIC
454: environments use 37. The corresponding values for CRLF are 3349 and 3365. The
455: default should normally correspond to the standard sequence for your operating
456: system.
457: .sp
458: PCRE_CONFIG_BSR
459: .sp
460: The output is an integer whose value indicates what character sequences the \eR
461: escape sequence matches by default. A value of 0 means that \eR matches any
462: Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
463: or CRLF. The default can be overridden when a pattern is compiled or matched.
464: .sp
465: PCRE_CONFIG_LINK_SIZE
466: .sp
467: The output is an integer that contains the number of bytes used for internal
468: linkage in compiled regular expressions. For the 8-bit library, the value can
469: be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
470: a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
471: still a number of bytes. The default value of 2 is sufficient for all but the
472: most massive patterns, since it allows the compiled pattern to be up to 64K in
473: size. Larger values allow larger regular expressions to be compiled, at the
474: expense of slower matching.
475: .sp
476: PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
477: .sp
478: The output is an integer that contains the threshold above which the POSIX
479: interface uses \fBmalloc()\fP for output vectors. Further details are given in
480: the
481: .\" HREF
482: \fBpcreposix\fP
483: .\"
484: documentation.
485: .sp
486: PCRE_CONFIG_MATCH_LIMIT
487: .sp
488: The output is a long integer that gives the default limit for the number of
489: internal matching function calls in a \fBpcre_exec()\fP execution. Further
490: details are given with \fBpcre_exec()\fP below.
491: .sp
492: PCRE_CONFIG_MATCH_LIMIT_RECURSION
493: .sp
494: The output is a long integer that gives the default limit for the depth of
495: recursion when calling the internal matching function in a \fBpcre_exec()\fP
496: execution. Further details are given with \fBpcre_exec()\fP below.
497: .sp
498: PCRE_CONFIG_STACKRECURSE
499: .sp
500: The output is an integer that is set to one if internal recursion when running
501: \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
502: to remember their state. This is the usual way that PCRE is compiled. The
503: output is zero if PCRE was compiled to use blocks of data on the heap instead
504: of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
505: \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
506: avoiding the use of the stack.
507: .
508: .
509: .SH "COMPILING A PATTERN"
510: .rs
511: .sp
512: .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
513: .ti +5n
514: .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
515: .ti +5n
516: .B const unsigned char *\fItableptr\fP);
517: .sp
518: .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
519: .ti +5n
520: .B int *\fIerrorcodeptr\fP,
521: .ti +5n
522: .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
523: .ti +5n
524: .B const unsigned char *\fItableptr\fP);
525: .P
526: Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
527: called to compile a pattern into an internal form. The only difference between
528: the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
529: \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
530: too much repetition, we refer just to \fBpcre_compile()\fP below, but the
531: information applies equally to \fBpcre_compile2()\fP.
532: .P
533: The pattern is a C string terminated by a binary zero, and is passed in the
534: \fIpattern\fP argument. A pointer to a single block of memory that is obtained
535: via \fBpcre_malloc\fP is returned. This contains the compiled code and related
536: data. The \fBpcre\fP type is defined for the returned block; this is a typedef
537: for a structure whose contents are not externally defined. It is up to the
538: caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
539: .P
540: Although the compiled code of a PCRE regex is relocatable, that is, it does not
541: depend on memory location, the complete \fBpcre\fP data block is not
542: fully relocatable, because it may contain a copy of the \fItableptr\fP
543: argument, which is an address (see below).
544: .P
545: The \fIoptions\fP argument contains various bit settings that affect the
546: compilation. It should be zero if no options are required. The available
547: options are described below. Some of them (in particular, those that are
548: compatible with Perl, but some others as well) can also be set and unset from
549: within the pattern (see the detailed description in the
550: .\" HREF
551: \fBpcrepattern\fP
552: .\"
553: documentation). For those options that can be different in different parts of
554: the pattern, the contents of the \fIoptions\fP argument specifies their
555: settings at the start of compilation and execution. The PCRE_ANCHORED,
556: PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
557: PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
558: compile time.
559: .P
560: If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
561: Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
562: NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
563: error message. This is a static string that is part of the library. You must
564: not try to free it. Normally, the offset from the start of the pattern to the
565: data unit that was being processed when the error was discovered is placed in
566: the variable pointed to by \fIerroffset\fP, which must not be NULL (if it is,
567: an immediate error is given). However, for an invalid UTF-8 or UTF-16 string,
568: the offset is that of the first data unit of the failing character.
569: .P
570: Some errors are not detected until the whole pattern has been scanned; in these
571: cases, the offset passed back is the length of the pattern. Note that the
572: offset is in data units, not characters, even in a UTF mode. It may sometimes
573: point into the middle of a UTF-8 or UTF-16 character.
574: .P
575: If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
576: \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
577: returned via this argument in the event of an error. This is in addition to the
578: textual error message. Error codes and messages are listed below.
579: .P
580: If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
581: character tables that are built when PCRE is compiled, using the default C
582: locale. Otherwise, \fItableptr\fP must be an address that is the result of a
583: call to \fBpcre_maketables()\fP. This value is stored with the compiled
584: pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
585: passed to it. For more discussion, see the section on locale support below.
586: .P
587: This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
588: .sp
589: pcre *re;
590: const char *error;
591: int erroffset;
592: re = pcre_compile(
593: "^A.*Z", /* the pattern */
594: 0, /* default options */
595: &error, /* for error message */
596: &erroffset, /* for error offset */
597: NULL); /* use default character tables */
598: .sp
599: The following names for option bits are defined in the \fBpcre.h\fP header
600: file:
601: .sp
602: PCRE_ANCHORED
603: .sp
604: If this bit is set, the pattern is forced to be "anchored", that is, it is
605: constrained to match only at the first matching point in the string that is
606: being searched (the "subject string"). This effect can also be achieved by
607: appropriate constructs in the pattern itself, which is the only way to do it in
608: Perl.
609: .sp
610: PCRE_AUTO_CALLOUT
611: .sp
612: If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
613: all with number 255, before each pattern item. For discussion of the callout
614: facility, see the
615: .\" HREF
616: \fBpcrecallout\fP
617: .\"
618: documentation.
619: .sp
620: PCRE_BSR_ANYCRLF
621: PCRE_BSR_UNICODE
622: .sp
623: These options (which are mutually exclusive) control what the \eR escape
624: sequence matches. The choice is either to match only CR, LF, or CRLF, or to
625: match any Unicode newline sequence. The default is specified when PCRE is
626: built. It can be overridden from within the pattern, or by setting an option
627: when a compiled pattern is matched.
628: .sp
629: PCRE_CASELESS
630: .sp
631: If this bit is set, letters in the pattern match both upper and lower case
632: letters. It is equivalent to Perl's /i option, and it can be changed within a
633: pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
634: concept of case for characters whose values are less than 128, so caseless
635: matching is always possible. For characters with higher values, the concept of
636: case is supported if PCRE is compiled with Unicode property support, but not
637: otherwise. If you want to use caseless matching for characters 128 and above,
638: you must ensure that PCRE is compiled with Unicode property support as well as
639: with UTF-8 support.
640: .sp
641: PCRE_DOLLAR_ENDONLY
642: .sp
643: If this bit is set, a dollar metacharacter in the pattern matches only at the
644: end of the subject string. Without this option, a dollar also matches
645: immediately before a newline at the end of the string (but not before any other
646: newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
647: There is no equivalent to this option in Perl, and no way to set it within a
648: pattern.
649: .sp
650: PCRE_DOTALL
651: .sp
652: If this bit is set, a dot metacharacter in the pattern matches a character of
653: any value, including one that indicates a newline. However, it only ever
654: matches one character, even if newlines are coded as CRLF. Without this option,
655: a dot does not match when the current position is at a newline. This option is
656: equivalent to Perl's /s option, and it can be changed within a pattern by a
657: (?s) option setting. A negative class such as [^a] always matches newline
658: characters, independent of the setting of this option.
659: .sp
660: PCRE_DUPNAMES
661: .sp
662: If this bit is set, names used to identify capturing subpatterns need not be
663: unique. This can be helpful for certain types of pattern when it is known that
664: only one instance of the named subpattern can ever be matched. There are more
665: details of named subpatterns below; see also the
666: .\" HREF
667: \fBpcrepattern\fP
668: .\"
669: documentation.
670: .sp
671: PCRE_EXTENDED
672: .sp
673: If this bit is set, white space data characters in the pattern are totally
674: ignored except when escaped or inside a character class. White space does not
675: include the VT character (code 11). In addition, characters between an
676: unescaped # outside a character class and the next newline, inclusive, are also
677: ignored. This is equivalent to Perl's /x option, and it can be changed within a
678: pattern by a (?x) option setting.
679: .P
680: Which characters are interpreted as newlines is controlled by the options
681: passed to \fBpcre_compile()\fP or by a special sequence at the start of the
682: pattern, as described in the section entitled
683: .\" HTML <a href="pcrepattern.html#newlines">
684: .\" </a>
685: "Newline conventions"
686: .\"
687: in the \fBpcrepattern\fP documentation. Note that the end of this type of
688: comment is a literal newline sequence in the pattern; escape sequences that
689: happen to represent a newline do not count.
690: .P
691: This option makes it possible to include comments inside complicated patterns.
692: Note, however, that this applies only to data characters. White space characters
693: may never appear within special character sequences in a pattern, for example
694: within the sequence (?( that introduces a conditional subpattern.
695: .sp
696: PCRE_EXTRA
697: .sp
698: This option was invented in order to turn on additional functionality of PCRE
699: that is incompatible with Perl, but it is currently of very little use. When
700: set, any backslash in a pattern that is followed by a letter that has no
701: special meaning causes an error, thus reserving these combinations for future
702: expansion. By default, as in Perl, a backslash followed by a letter with no
703: special meaning is treated as a literal. (Perl can, however, be persuaded to
704: give an error for this, by running it with the -w option.) There are at present
705: no other features controlled by this option. It can also be set by a (?X)
706: option setting within a pattern.
707: .sp
708: PCRE_FIRSTLINE
709: .sp
710: If this option is set, an unanchored pattern is required to match before or at
711: the first newline in the subject string, though the matched text may continue
712: over the newline.
713: .sp
714: PCRE_JAVASCRIPT_COMPAT
715: .sp
716: If this option is set, PCRE's behaviour is changed in some ways so that it is
717: compatible with JavaScript rather than Perl. The changes are as follows:
718: .P
719: (1) A lone closing square bracket in a pattern causes a compile-time error,
720: because this is illegal in JavaScript (by default it is treated as a data
721: character). Thus, the pattern AB]CD becomes illegal when this option is set.
722: .P
723: (2) At run time, a back reference to an unset subpattern group matches an empty
724: string (by default this causes the current matching alternative to fail). A
725: pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
726: an "a" in the subject), whereas it fails by default, for Perl compatibility.
727: .P
728: (3) \eU matches an upper case "U" character; by default \eU causes a compile
729: time error (Perl uses \eU to upper case subsequent characters).
730: .P
731: (4) \eu matches a lower case "u" character unless it is followed by four
732: hexadecimal digits, in which case the hexadecimal number defines the code point
733: to match. By default, \eu causes a compile time error (Perl uses it to upper
734: case the following character).
735: .P
736: (5) \ex matches a lower case "x" character unless it is followed by two
737: hexadecimal digits, in which case the hexadecimal number defines the code point
738: to match. By default, as in Perl, a hexadecimal number is always expected after
739: \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
740: binary zero character followed by z).
741: .sp
742: PCRE_MULTILINE
743: .sp
744: By default, for the purposes of matching "start of line" and "end of line",
745: PCRE treats the subject string as consisting of a single line of characters,
746: even if it actually contains newlines. The "start of line" metacharacter (^)
747: matches only at the start of the string, and the "end of line" metacharacter
748: ($) matches only at the end of the string, or before a terminating newline
749: (except when PCRE_DOLLAR_ENDONLY is set). Note, however, that unless
750: PCRE_DOTALL is set, the "any character" metacharacter (.) does not match at a
751: newline. This behaviour (for ^, $, and dot) is the same as Perl.
752: .P
753: When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
754: match immediately following or immediately before internal newlines in the
755: subject string, respectively, as well as at the very start and end. This is
756: equivalent to Perl's /m option, and it can be changed within a pattern by a
757: (?m) option setting. If there are no newlines in a subject string, or no
758: occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
759: .sp
760: PCRE_NEVER_UTF
761: .sp
762: This option locks out interpretation of the pattern as UTF-8 (or UTF-16 or
763: UTF-32 in the 16-bit and 32-bit libraries). In particular, it prevents the
764: creator of the pattern from switching to UTF interpretation by starting the
765: pattern with (*UTF). This may be useful in applications that process patterns
766: from external sources. The combination of PCRE_UTF8 and PCRE_NEVER_UTF also
767: causes an error.
768: .sp
769: PCRE_NEWLINE_CR
770: PCRE_NEWLINE_LF
771: PCRE_NEWLINE_CRLF
772: PCRE_NEWLINE_ANYCRLF
773: PCRE_NEWLINE_ANY
774: .sp
775: These options override the default newline definition that was chosen when PCRE
776: was built. Setting the first or the second specifies that a newline is
777: indicated by a single character (CR or LF, respectively). Setting
778: PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
779: CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
780: preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
781: that any Unicode newline sequence should be recognized.
782: .P
783: In an ASCII/Unicode environment, the Unicode newline sequences are the three
784: just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
785: feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
786: (paragraph separator, U+2029). For the 8-bit library, the last two are
787: recognized only in UTF-8 mode.
788: .P
789: When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
790: CR is 0x0d, the same as ASCII. However, the character code for LF is normally
791: 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
792: not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
793: less than 256. For more details, see the
794: .\" HREF
795: \fBpcrebuild\fP
796: .\"
797: documentation.
798: .P
799: The newline setting in the options word uses three bits that are treated
800: as a number, giving eight possibilities. Currently only six are used (default
801: plus the five values above). This means that if you set more than one newline
802: option, the combination may or may not be sensible. For example,
803: PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
804: other combinations may yield unused numbers and cause an error.
805: .P
806: The only time that a line break in a pattern is specially recognized when
807: compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
808: and so are ignored in this mode. Also, an unescaped # outside a character class
809: indicates a comment that lasts until after the next line break sequence. In
810: other circumstances, line break sequences in patterns are treated as literal
811: data.
812: .P
813: The newline option that is set at compile time becomes the default that is used
814: for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
815: .sp
816: PCRE_NO_AUTO_CAPTURE
817: .sp
818: If this option is set, it disables the use of numbered capturing parentheses in
819: the pattern. Any opening parenthesis that is not followed by ? behaves as if it
820: were followed by ?: but named parentheses can still be used for capturing (and
821: they acquire numbers in the usual way). There is no equivalent of this option
822: in Perl.
823: .sp
824: PCRE_NO_START_OPTIMIZE
825: .sp
826: This is an option that acts at matching time; that is, it is really an option
827: for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
828: it is remembered with the compiled pattern and assumed at matching time. This
829: is necessary if you want to use JIT execution, because the JIT compiler needs
830: to know whether or not this option is set. For details see the discussion of
831: PCRE_NO_START_OPTIMIZE
832: .\" HTML <a href="#execoptions">
833: .\" </a>
834: below.
835: .\"
836: .sp
837: PCRE_UCP
838: .sp
839: This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
840: \ew, and some of the POSIX character classes. By default, only ASCII characters
841: are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
842: classify characters. More details are given in the section on
843: .\" HTML <a href="pcre.html#genericchartypes">
844: .\" </a>
845: generic character types
846: .\"
847: in the
848: .\" HREF
849: \fBpcrepattern\fP
850: .\"
851: page. If you set PCRE_UCP, matching one of the items it affects takes much
852: longer. The option is available only if PCRE has been compiled with Unicode
853: property support.
854: .sp
855: PCRE_UNGREEDY
856: .sp
857: This option inverts the "greediness" of the quantifiers so that they are not
858: greedy by default, but become greedy if followed by "?". It is not compatible
859: with Perl. It can also be set by a (?U) option setting within the pattern.
860: .sp
861: PCRE_UTF8
862: .sp
863: This option causes PCRE to regard both the pattern and the subject as strings
864: of UTF-8 characters instead of single-byte strings. However, it is available
865: only when PCRE is built to include UTF support. If not, the use of this option
866: provokes an error. Details of how this option changes the behaviour of PCRE are
867: given in the
868: .\" HREF
869: \fBpcreunicode\fP
870: .\"
871: page.
872: .sp
873: PCRE_NO_UTF8_CHECK
874: .sp
875: When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
876: automatically checked. There is a discussion about the
877: .\" HTML <a href="pcreunicode.html#utf8strings">
878: .\" </a>
879: validity of UTF-8 strings
880: .\"
881: in the
882: .\" HREF
883: \fBpcreunicode\fP
884: .\"
885: page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
886: error. If you already know that your pattern is valid, and you want to skip
887: this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
888: When it is set, the effect of passing an invalid UTF-8 string as a pattern is
889: undefined. It may cause your program to crash. Note that this option can also
890: be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
891: validity checking of subject strings only. If the same string is being matched
892: many times, the option can be safely set for the second and subsequent
893: matchings to improve performance.
894: .
895: .
896: .SH "COMPILATION ERROR CODES"
897: .rs
898: .sp
899: The following table lists the error codes than may be returned by
900: \fBpcre_compile2()\fP, along with the error messages that may be returned by
901: both compiling functions. Note that error messages are always 8-bit ASCII
902: strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
903: have fallen out of use. To avoid confusion, they have not been re-used.
904: .sp
905: 0 no error
906: 1 \e at end of pattern
907: 2 \ec at end of pattern
908: 3 unrecognized character follows \e
909: 4 numbers out of order in {} quantifier
910: 5 number too big in {} quantifier
911: 6 missing terminating ] for character class
912: 7 invalid escape sequence in character class
913: 8 range out of order in character class
914: 9 nothing to repeat
915: 10 [this code is not in use]
916: 11 internal error: unexpected repeat
917: 12 unrecognized character after (? or (?-
918: 13 POSIX named classes are supported only within a class
919: 14 missing )
920: 15 reference to non-existent subpattern
921: 16 erroffset passed as NULL
922: 17 unknown option bit(s) set
923: 18 missing ) after comment
924: 19 [this code is not in use]
925: 20 regular expression is too large
926: 21 failed to get memory
927: 22 unmatched parentheses
928: 23 internal error: code overflow
929: 24 unrecognized character after (?<
930: 25 lookbehind assertion is not fixed length
931: 26 malformed number or name after (?(
932: 27 conditional group contains more than two branches
933: 28 assertion expected after (?(
934: 29 (?R or (?[+-]digits must be followed by )
935: 30 unknown POSIX class name
936: 31 POSIX collating elements are not supported
937: 32 this version of PCRE is compiled without UTF support
938: 33 [this code is not in use]
939: 34 character value in \ex{...} sequence is too large
940: 35 invalid condition (?(0)
941: 36 \eC not allowed in lookbehind assertion
942: 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
943: 38 number after (?C is > 255
944: 39 closing ) for (?C expected
945: 40 recursive call could loop indefinitely
946: 41 unrecognized character after (?P
947: 42 syntax error in subpattern name (missing terminator)
948: 43 two named subpatterns have the same name
949: 44 invalid UTF-8 string (specifically UTF-8)
950: 45 support for \eP, \ep, and \eX has not been compiled
951: 46 malformed \eP or \ep sequence
952: 47 unknown property name after \eP or \ep
953: 48 subpattern name is too long (maximum 32 characters)
954: 49 too many named subpatterns (maximum 10000)
955: 50 [this code is not in use]
956: 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
957: 52 internal error: overran compiling workspace
958: 53 internal error: previously-checked referenced subpattern
959: not found
960: 54 DEFINE group contains more than one branch
961: 55 repeating a DEFINE group is not allowed
962: 56 inconsistent NEWLINE options
963: 57 \eg is not followed by a braced, angle-bracketed, or quoted
964: name/number or by a plain number
965: 58 a numbered reference must not be zero
966: 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
967: 60 (*VERB) not recognized or malformed
968: 61 number is too big
969: 62 subpattern name expected
970: 63 digit expected after (?+
971: 64 ] is an invalid data character in JavaScript compatibility mode
972: 65 different names for subpatterns of the same number are
973: not allowed
974: 66 (*MARK) must have an argument
975: 67 this version of PCRE is not compiled with Unicode property
976: support
977: 68 \ec must be followed by an ASCII character
978: 69 \ek is not followed by a braced, angle-bracketed, or quoted name
979: 70 internal error: unknown opcode in find_fixedlength()
980: 71 \eN is not supported in a class
981: 72 too many forward references
982: 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
983: 74 invalid UTF-16 string (specifically UTF-16)
984: 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
985: 76 character value in \eu.... sequence is too large
986: 77 invalid UTF-32 string (specifically UTF-32)
987: .sp
988: The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
989: be used if the limits were changed when PCRE was built.
990: .
991: .
992: .\" HTML <a name="studyingapattern"></a>
993: .SH "STUDYING A PATTERN"
994: .rs
995: .sp
996: .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
997: .ti +5n
998: .B const char **\fIerrptr\fP);
999: .PP
1000: If a compiled pattern is going to be used several times, it is worth spending
1001: more time analyzing it in order to speed up the time taken for matching. The
1002: function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
1003: argument. If studying the pattern produces additional information that will
1004: help speed up matching, \fBpcre_study()\fP returns a pointer to a
1005: \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
1006: results of the study.
1007: .P
1008: The returned value from \fBpcre_study()\fP can be passed directly to
1009: \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
1010: also contains other fields that can be set by the caller before the block is
1011: passed; these are described
1012: .\" HTML <a href="#extradata">
1013: .\" </a>
1014: below
1015: .\"
1016: in the section on matching a pattern.
1017: .P
1018: If studying the pattern does not produce any useful information,
1019: \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1020: calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1021: \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1022: if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1023: returns a \fBpcre_extra\fP block even if studying did not find any additional
1024: information. It may still return NULL, however, if an error occurs in
1025: \fBpcre_study()\fP.
1026: .P
1027: The second argument of \fBpcre_study()\fP contains option bits. There are three
1028: further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1029: .sp
1030: PCRE_STUDY_JIT_COMPILE
1031: PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1032: PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1033: .sp
1034: If any of these are set, and the just-in-time compiler is available, the
1035: pattern is further compiled into machine code that executes much faster than
1036: the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1037: compiler is not available, these options are ignored. All undefined bits in the
1038: \fIoptions\fP argument must be zero.
1039: .P
1040: JIT compilation is a heavyweight optimization. It can take some time for
1041: patterns to be analyzed, and for one-off matches and simple patterns the
1042: benefit of faster execution might be offset by a much slower study time.
1043: Not all patterns can be optimized by the JIT compiler. For those that cannot be
1044: handled, matching automatically falls back to the \fBpcre_exec()\fP
1045: interpreter. For more details, see the
1046: .\" HREF
1047: \fBpcrejit\fP
1048: .\"
1049: documentation.
1050: .P
1051: The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1052: studying succeeds (even if no data is returned), the variable it points to is
1053: set to NULL. Otherwise it is set to point to a textual error message. This is a
1054: static string that is part of the library. You must not try to free it. You
1055: should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1056: sure that it has run successfully.
1057: .P
1058: When you are finished with a pattern, you can free the memory used for the
1059: study data by calling \fBpcre_free_study()\fP. This function was added to the
1060: API for release 8.20. For earlier versions, the memory could be freed with
1061: \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1062: where JIT optimization is not used, but it is advisable to change to the new
1063: function when convenient.
1064: .P
1065: This is a typical way in which \fBpcre_study\fP() is used (except that in a
1066: real application there should be tests for errors):
1067: .sp
1068: int rc;
1069: pcre *re;
1070: pcre_extra *sd;
1071: re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1072: sd = pcre_study(
1073: re, /* result of pcre_compile() */
1074: 0, /* no options */
1075: &error); /* set to NULL or points to a message */
1076: rc = pcre_exec( /* see below for details of pcre_exec() options */
1077: re, sd, "subject", 7, 0, 0, ovector, 30);
1078: ...
1079: pcre_free_study(sd);
1080: pcre_free(re);
1081: .sp
1082: Studying a pattern does two things: first, a lower bound for the length of
1083: subject string that is needed to match the pattern is computed. This does not
1084: mean that there are any strings of that length that match, but it does
1085: guarantee that no shorter strings match. The value is used to avoid wasting
1086: time by trying to match strings that are shorter than the lower bound. You can
1087: find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1088: .P
1089: Studying a pattern is also useful for non-anchored patterns that do not have a
1090: single fixed starting character. A bitmap of possible starting bytes is
1091: created. This speeds up finding a position in the subject at which to start
1092: matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1093: In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1094: .P
1095: These two optimizations apply to both \fBpcre_exec()\fP and
1096: \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1097: The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1098: You might want to do this if your pattern contains callouts or (*MARK) and you
1099: want to make use of these facilities in cases where matching fails.
1100: .P
1101: PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1102: time. However, if PCRE_NO_START_OPTIMIZE is passed to \fBpcre_exec()\fP, (that
1103: is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1104: execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1105: compile time.
1106: .P
1107: There is a longer discussion of PCRE_NO_START_OPTIMIZE
1108: .\" HTML <a href="#execoptions">
1109: .\" </a>
1110: below.
1111: .\"
1112: .
1113: .
1114: .\" HTML <a name="localesupport"></a>
1115: .SH "LOCALE SUPPORT"
1116: .rs
1117: .sp
1118: PCRE handles caseless matching, and determines whether characters are letters,
1119: digits, or whatever, by reference to a set of tables, indexed by character
1120: value. When running in UTF-8 mode, this applies only to characters
1121: with codes less than 128. By default, higher-valued codes never match escapes
1122: such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1123: Unicode character property support. Alternatively, the PCRE_UCP option can be
1124: set at compile time; this causes \ew and friends to use Unicode property
1125: support instead of built-in tables. The use of locales with Unicode is
1126: discouraged. If you are handling characters with codes greater than 128, you
1127: should either use UTF-8 and Unicode, or use locales, but not try to mix the
1128: two.
1129: .P
1130: PCRE contains an internal set of tables that are used when the final argument
1131: of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1132: Normally, the internal tables recognize only ASCII characters. However, when
1133: PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1134: default "C" locale of the local system, which may cause them to be different.
1135: .P
1136: The internal tables can always be overridden by tables supplied by the
1137: application that calls PCRE. These may be created in a different locale from
1138: the default. As more and more applications change to using Unicode, the need
1139: for this locale support is expected to die away.
1140: .P
1141: External tables are built by calling the \fBpcre_maketables()\fP function,
1142: which has no arguments, in the relevant locale. The result can then be passed
1143: to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1144: example, to build and use tables that are appropriate for the French locale
1145: (where accented characters with values greater than 128 are treated as letters),
1146: the following code could be used:
1147: .sp
1148: setlocale(LC_CTYPE, "fr_FR");
1149: tables = pcre_maketables();
1150: re = pcre_compile(..., tables);
1151: .sp
1152: The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1153: are using Windows, the name for the French locale is "french".
1154: .P
1155: When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1156: obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1157: that the memory containing the tables remains available for as long as it is
1158: needed.
1159: .P
1160: The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1161: pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1162: and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1163: pattern, compilation, studying and matching all happen in the same locale, but
1164: different patterns can be compiled in different locales.
1165: .P
1166: It is possible to pass a table pointer or NULL (indicating the use of the
1167: internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1168: this facility could be used to match a pattern in a different locale from the
1169: one in which it was compiled. Passing table pointers at run time is discussed
1170: below in the section on matching a pattern.
1171: .
1172: .
1173: .\" HTML <a name="infoaboutpattern"></a>
1174: .SH "INFORMATION ABOUT A PATTERN"
1175: .rs
1176: .sp
1177: .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1178: .ti +5n
1179: .B int \fIwhat\fP, void *\fIwhere\fP);
1180: .PP
1181: The \fBpcre_fullinfo()\fP function returns information about a compiled
1182: pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1183: library at version 8.30, after more than 10 years of obsolescence.
1184: .P
1185: The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1186: pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1187: the pattern was not studied. The third argument specifies which piece of
1188: information is required, and the fourth argument is a pointer to a variable
1189: to receive the data. The yield of the function is zero for success, or one of
1190: the following negative numbers:
1191: .sp
1192: PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1193: the argument \fIwhere\fP was NULL
1194: PCRE_ERROR_BADMAGIC the "magic number" was not found
1195: PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1196: endianness
1197: PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1198: PCRE_ERROR_UNSET the requested field is not set
1199: .sp
1200: The "magic number" is placed at the start of each compiled pattern as an simple
1201: check against passing an arbitrary memory pointer. The endianness error can
1202: occur if a compiled pattern is saved and reloaded on a different host. Here is
1203: a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1204: pattern:
1205: .sp
1206: int rc;
1207: size_t length;
1208: rc = pcre_fullinfo(
1209: re, /* result of pcre_compile() */
1210: sd, /* result of pcre_study(), or NULL */
1211: PCRE_INFO_SIZE, /* what is required */
1212: &length); /* where to put the data */
1213: .sp
1214: The possible values for the third argument are defined in \fBpcre.h\fP, and are
1215: as follows:
1216: .sp
1217: PCRE_INFO_BACKREFMAX
1218: .sp
1219: Return the number of the highest back reference in the pattern. The fourth
1220: argument should point to an \fBint\fP variable. Zero is returned if there are
1221: no back references.
1222: .sp
1223: PCRE_INFO_CAPTURECOUNT
1224: .sp
1225: Return the number of capturing subpatterns in the pattern. The fourth argument
1226: should point to an \fBint\fP variable.
1227: .sp
1228: PCRE_INFO_DEFAULT_TABLES
1229: .sp
1230: Return a pointer to the internal default character tables within PCRE. The
1231: fourth argument should point to an \fBunsigned char *\fP variable. This
1232: information call is provided for internal use by the \fBpcre_study()\fP
1233: function. External callers can cause PCRE to use its internal tables by passing
1234: a NULL table pointer.
1235: .sp
1236: PCRE_INFO_FIRSTBYTE
1237: .sp
1238: Return information about the first data unit of any matched string, for a
1239: non-anchored pattern. (The name of this option refers to the 8-bit library,
1240: where data units are bytes.) The fourth argument should point to an \fBint\fP
1241: variable.
1242: .P
1243: If there is a fixed first value, for example, the letter "c" from a pattern
1244: such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1245: value is always less than 256. In the 16-bit library the value can be up to
1246: 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1247: .P
1248: If there is no fixed first value, and if either
1249: .sp
1250: (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1251: starts with "^", or
1252: .sp
1253: (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1254: (if it were set, the pattern would be anchored),
1255: .sp
1256: -1 is returned, indicating that the pattern matches only at the start of a
1257: subject string or after any newline within the string. Otherwise -2 is
1258: returned. For anchored patterns, -2 is returned.
1259: .P
1260: Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1261: to return the full 32-bit range of the character, this value is deprecated;
1262: instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1263: should be used.
1264: .sp
1265: PCRE_INFO_FIRSTTABLE
1266: .sp
1267: If the pattern was studied, and this resulted in the construction of a 256-bit
1268: table indicating a fixed set of values for the first data unit in any matching
1269: string, a pointer to the table is returned. Otherwise NULL is returned. The
1270: fourth argument should point to an \fBunsigned char *\fP variable.
1271: .sp
1272: PCRE_INFO_HASCRORLF
1273: .sp
1274: Return 1 if the pattern contains any explicit matches for CR or LF characters,
1275: otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1276: explicit match is either a literal CR or LF character, or \er or \en.
1277: .sp
1278: PCRE_INFO_JCHANGED
1279: .sp
1280: Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1281: 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1282: (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1283: .sp
1284: PCRE_INFO_JIT
1285: .sp
1286: Return 1 if the pattern was studied with one of the JIT options, and
1287: just-in-time compiling was successful. The fourth argument should point to an
1288: \fBint\fP variable. A return value of 0 means that JIT support is not available
1289: in this version of PCRE, or that the pattern was not studied with a JIT option,
1290: or that the JIT compiler could not handle this particular pattern. See the
1291: .\" HREF
1292: \fBpcrejit\fP
1293: .\"
1294: documentation for details of what can and cannot be handled.
1295: .sp
1296: PCRE_INFO_JITSIZE
1297: .sp
1298: If the pattern was successfully studied with a JIT option, return the size of
1299: the JIT compiled code, otherwise return zero. The fourth argument should point
1300: to a \fBsize_t\fP variable.
1301: .sp
1302: PCRE_INFO_LASTLITERAL
1303: .sp
1304: Return the value of the rightmost literal data unit that must exist in any
1305: matched string, other than at its start, if such a value has been recorded. The
1306: fourth argument should point to an \fBint\fP variable. If there is no such
1307: value, -1 is returned. For anchored patterns, a last literal value is recorded
1308: only if it follows something of variable length. For example, for the pattern
1309: /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1310: is -1.
1311: .P
1312: Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1313: to return the full 32-bit range of the character, this value is deprecated;
1314: instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1315: be used.
1316: .sp
1317: PCRE_INFO_MATCHLIMIT
1318: .sp
1319: If the pattern set a match limit by including an item of the form
1320: (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1321: should point to an unsigned 32-bit integer. If no such value has been set, the
1322: call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1323: .sp
1324: PCRE_INFO_MAXLOOKBEHIND
1325: .sp
1326: Return the number of characters (NB not data units) in the longest lookbehind
1327: assertion in the pattern. This information is useful when doing multi-segment
1328: matching using the partial matching facilities. Note that the simple assertions
1329: \eb and \eB require a one-character lookbehind. \eA also registers a
1330: one-character lookbehind, though it does not actually inspect the previous
1331: character. This is to ensure that at least one character from the old segment
1332: is retained when a new segment is processed. Otherwise, if there are no
1333: lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1334: segment.
1335: .sp
1336: PCRE_INFO_MINLENGTH
1337: .sp
1338: If the pattern was studied and a minimum length for matching subject strings
1339: was computed, its value is returned. Otherwise the returned value is -1. The
1340: value is a number of characters, which in UTF mode may be different from the
1341: number of data units. The fourth argument should point to an \fBint\fP
1342: variable. A non-negative value is a lower bound to the length of any matching
1343: string. There may not be any strings of that length that do actually match, but
1344: every string that does match is at least that long.
1345: .sp
1346: PCRE_INFO_NAMECOUNT
1347: PCRE_INFO_NAMEENTRYSIZE
1348: PCRE_INFO_NAMETABLE
1349: .sp
1350: PCRE supports the use of named as well as numbered capturing parentheses. The
1351: names are just an additional way of identifying the parentheses, which still
1352: acquire numbers. Several convenience functions such as
1353: \fBpcre_get_named_substring()\fP are provided for extracting captured
1354: substrings by name. It is also possible to extract the data directly, by first
1355: converting the name to a number in order to access the correct pointers in the
1356: output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1357: you need to use the name-to-number map, which is described by these three
1358: values.
1359: .P
1360: The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1361: the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1362: entry; both of these return an \fBint\fP value. The entry size depends on the
1363: length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1364: entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1365: the first two bytes of each entry are the number of the capturing parenthesis,
1366: most significant byte first. In the 16-bit library, the pointer points to
1367: 16-bit data units, the first of which contains the parenthesis number. In the
1368: 32-bit library, the pointer points to 32-bit data units, the first of which
1369: contains the parenthesis number. The rest of the entry is the corresponding
1370: name, zero terminated.
1371: .P
1372: The names are in alphabetical order. Duplicate names may appear if (?| is used
1373: to create multiple groups with the same number, as described in the
1374: .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1375: .\" </a>
1376: section on duplicate subpattern numbers
1377: .\"
1378: in the
1379: .\" HREF
1380: \fBpcrepattern\fP
1381: .\"
1382: page. Duplicate names for subpatterns with different numbers are permitted only
1383: if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1384: table in the order in which they were found in the pattern. In the absence of
1385: (?| this is the order of increasing number; when (?| is used this is not
1386: necessarily the case because later subpatterns may have lower numbers.
1387: .P
1388: As a simple example of the name/number table, consider the following pattern
1389: after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1390: space - including newlines - is ignored):
1391: .sp
1392: .\" JOIN
1393: (?<date> (?<year>(\ed\ed)?\ed\ed) -
1394: (?<month>\ed\ed) - (?<day>\ed\ed) )
1395: .sp
1396: There are four named subpatterns, so the table has four entries, and each entry
1397: in the table is eight bytes long. The table is as follows, with non-printing
1398: bytes shows in hexadecimal, and undefined bytes shown as ??:
1399: .sp
1400: 00 01 d a t e 00 ??
1401: 00 05 d a y 00 ?? ??
1402: 00 04 m o n t h 00
1403: 00 02 y e a r 00 ??
1404: .sp
1405: When writing code to extract data from named subpatterns using the
1406: name-to-number map, remember that the length of the entries is likely to be
1407: different for each compiled pattern.
1408: .sp
1409: PCRE_INFO_OKPARTIAL
1410: .sp
1411: Return 1 if the pattern can be used for partial matching with
1412: \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1413: \fBint\fP variable. From release 8.00, this always returns 1, because the
1414: restrictions that previously applied to partial matching have been lifted. The
1415: .\" HREF
1416: \fBpcrepartial\fP
1417: .\"
1418: documentation gives details of partial matching.
1419: .sp
1420: PCRE_INFO_OPTIONS
1421: .sp
1422: Return a copy of the options with which the pattern was compiled. The fourth
1423: argument should point to an \fBunsigned long int\fP variable. These option bits
1424: are those specified in the call to \fBpcre_compile()\fP, modified by any
1425: top-level option settings at the start of the pattern itself. In other words,
1426: they are the options that will be in force when matching starts. For example,
1427: if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1428: result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1429: .P
1430: A pattern is automatically anchored by PCRE if all of its top-level
1431: alternatives begin with one of the following:
1432: .sp
1433: ^ unless PCRE_MULTILINE is set
1434: \eA always
1435: \eG always
1436: .\" JOIN
1437: .* if PCRE_DOTALL is set and there are no back
1438: references to the subpattern in which .* appears
1439: .sp
1440: For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1441: \fBpcre_fullinfo()\fP.
1442: .sp
1443: PCRE_INFO_RECURSIONLIMIT
1444: .sp
1445: If the pattern set a recursion limit by including an item of the form
1446: (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1447: argument should point to an unsigned 32-bit integer. If no such value has been
1448: set, the call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1449: .sp
1450: PCRE_INFO_SIZE
1451: .sp
1452: Return the size of the compiled pattern in bytes (for all three libraries). The
1453: fourth argument should point to a \fBsize_t\fP variable. This value does not
1454: include the size of the \fBpcre\fP structure that is returned by
1455: \fBpcre_compile()\fP. The value that is passed as the argument to
1456: \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1457: place the compiled data is the value returned by this option plus the size of
1458: the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1459: does not alter the value returned by this option.
1460: .sp
1461: PCRE_INFO_STUDYSIZE
1462: .sp
1463: Return the size in bytes (for all three libraries) of the data block pointed to
1464: by the \fIstudy_data\fP field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP
1465: is NULL, or there is no study data, zero is returned. The fourth argument
1466: should point to a \fBsize_t\fP variable. The \fIstudy_data\fP field is set by
1467: \fBpcre_study()\fP to record information that will speed up matching (see the
1468: section entitled
1469: .\" HTML <a href="#studyingapattern">
1470: .\" </a>
1471: "Studying a pattern"
1472: .\"
1473: above). The format of the \fIstudy_data\fP block is private, but its length
1474: is made available via this option so that it can be saved and restored (see the
1475: .\" HREF
1476: \fBpcreprecompile\fP
1477: .\"
1478: documentation for details).
1479: .sp
1480: PCRE_INFO_FIRSTCHARACTERFLAGS
1481: .sp
1482: Return information about the first data unit of any matched string, for a
1483: non-anchored pattern. The fourth argument should point to an \fBint\fP
1484: variable.
1485: .P
1486: If there is a fixed first value, for example, the letter "c" from a pattern
1487: such as (cat|cow|coyote), 1 is returned, and the character value can be
1488: retrieved using PCRE_INFO_FIRSTCHARACTER.
1489: .P
1490: If there is no fixed first value, and if either
1491: .sp
1492: (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1493: starts with "^", or
1494: .sp
1495: (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1496: (if it were set, the pattern would be anchored),
1497: .sp
1498: 2 is returned, indicating that the pattern matches only at the start of a
1499: subject string or after any newline within the string. Otherwise 0 is
1500: returned. For anchored patterns, 0 is returned.
1501: .sp
1502: PCRE_INFO_FIRSTCHARACTER
1503: .sp
1504: Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1505: returned 1; otherwise returns 0. The fourth argument should point to an
1506: \fBuint_t\fP variable.
1507: .P
1508: In the 8-bit library, the value is always less than 256. In the 16-bit library
1509: the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1510: can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1511: .P
1512: If there is no fixed first value, and if either
1513: .sp
1514: (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1515: starts with "^", or
1516: .sp
1517: (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1518: (if it were set, the pattern would be anchored),
1519: .sp
1520: -1 is returned, indicating that the pattern matches only at the start of a
1521: subject string or after any newline within the string. Otherwise -2 is
1522: returned. For anchored patterns, -2 is returned.
1523: .sp
1524: PCRE_INFO_REQUIREDCHARFLAGS
1525: .sp
1526: Returns 1 if there is a rightmost literal data unit that must exist in any
1527: matched string, other than at its start. The fourth argument should point to
1528: an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1529: 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1530: .P
1531: For anchored patterns, a last literal value is recorded only if it follows
1532: something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1533: returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1534: /^a\edz\ed/ the returned value is 0.
1535: .sp
1536: PCRE_INFO_REQUIREDCHAR
1537: .sp
1538: Return the value of the rightmost literal data unit that must exist in any
1539: matched string, other than at its start, if such a value has been recorded. The
1540: fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1541: value, 0 is returned.
1542: .
1543: .
1544: .SH "REFERENCE COUNTS"
1545: .rs
1546: .sp
1547: .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1548: .PP
1549: The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1550: data block that contains a compiled pattern. It is provided for the benefit of
1551: applications that operate in an object-oriented manner, where different parts
1552: of the application may be using the same compiled pattern, but you want to free
1553: the block when they are all done.
1554: .P
1555: When a pattern is compiled, the reference count field is initialized to zero.
1556: It is changed only by calling this function, whose action is to add the
1557: \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1558: function is the new value. However, the value of the count is constrained to
1559: lie between 0 and 65535, inclusive. If the new value is outside these limits,
1560: it is forced to the appropriate limit value.
1561: .P
1562: Except when it is zero, the reference count is not correctly preserved if a
1563: pattern is compiled on one host and then transferred to a host whose byte-order
1564: is different. (This seems a highly unlikely scenario.)
1565: .
1566: .
1567: .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1568: .rs
1569: .sp
1570: .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1571: .ti +5n
1572: .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1573: .ti +5n
1574: .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1575: .P
1576: The function \fBpcre_exec()\fP is called to match a subject string against a
1577: compiled pattern, which is passed in the \fIcode\fP argument. If the
1578: pattern was studied, the result of the study should be passed in the
1579: \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1580: and \fIextra\fP arguments as many times as you like, in order to match
1581: different subject strings with the same pattern.
1582: .P
1583: This function is the main matching facility of the library, and it operates in
1584: a Perl-like manner. For specialist use there is also an alternative matching
1585: function, which is described
1586: .\" HTML <a href="#dfamatch">
1587: .\" </a>
1588: below
1589: .\"
1590: in the section about the \fBpcre_dfa_exec()\fP function.
1591: .P
1592: In most applications, the pattern will have been compiled (and optionally
1593: studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1594: possible to save compiled patterns and study data, and then use them later
1595: in different processes, possibly even on different hosts. For a discussion
1596: about this, see the
1597: .\" HREF
1598: \fBpcreprecompile\fP
1599: .\"
1600: documentation.
1601: .P
1602: Here is an example of a simple call to \fBpcre_exec()\fP:
1603: .sp
1604: int rc;
1605: int ovector[30];
1606: rc = pcre_exec(
1607: re, /* result of pcre_compile() */
1608: NULL, /* we didn't study the pattern */
1609: "some string", /* the subject string */
1610: 11, /* the length of the subject string */
1611: 0, /* start at offset 0 in the subject */
1612: 0, /* default options */
1613: ovector, /* vector of integers for substring information */
1614: 30); /* number of elements (NOT size in bytes) */
1615: .
1616: .
1617: .\" HTML <a name="extradata"></a>
1618: .SS "Extra data for \fBpcre_exec()\fR"
1619: .rs
1620: .sp
1621: If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1622: data block. The \fBpcre_study()\fP function returns such a block (when it
1623: doesn't return NULL), but you can also create one for yourself, and pass
1624: additional information in it. The \fBpcre_extra\fP block contains the following
1625: fields (not necessarily in this order):
1626: .sp
1627: unsigned long int \fIflags\fP;
1628: void *\fIstudy_data\fP;
1629: void *\fIexecutable_jit\fP;
1630: unsigned long int \fImatch_limit\fP;
1631: unsigned long int \fImatch_limit_recursion\fP;
1632: void *\fIcallout_data\fP;
1633: const unsigned char *\fItables\fP;
1634: unsigned char **\fImark\fP;
1635: .sp
1636: In the 16-bit version of this structure, the \fImark\fP field has type
1637: "PCRE_UCHAR16 **".
1638: .sp
1639: In the 32-bit version of this structure, the \fImark\fP field has type
1640: "PCRE_UCHAR32 **".
1641: .P
1642: The \fIflags\fP field is used to specify which of the other fields are set. The
1643: flag bits are:
1644: .sp
1645: PCRE_EXTRA_CALLOUT_DATA
1646: PCRE_EXTRA_EXECUTABLE_JIT
1647: PCRE_EXTRA_MARK
1648: PCRE_EXTRA_MATCH_LIMIT
1649: PCRE_EXTRA_MATCH_LIMIT_RECURSION
1650: PCRE_EXTRA_STUDY_DATA
1651: PCRE_EXTRA_TABLES
1652: .sp
1653: Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1654: the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1655: returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1656: should not set these yourself, but you may add to the block by setting other
1657: fields and their corresponding flag bits.
1658: .P
1659: The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1660: vast amount of resources when running patterns that are not going to match,
1661: but which have a very large number of possibilities in their search trees. The
1662: classic example is a pattern that uses nested unlimited repeats.
1663: .P
1664: Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1665: calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1666: imposed on the number of times this function is called during a match, which
1667: has the effect of limiting the amount of backtracking that can take place. For
1668: patterns that are not anchored, the count restarts from zero for each position
1669: in the subject string.
1670: .P
1671: When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1672: with a JIT option, the way that the matching is executed is entirely different.
1673: However, there is still the possibility of runaway matching that goes on for a
1674: very long time, and so the \fImatch_limit\fP value is also used in this case
1675: (but in a different way) to limit how long the matching can continue.
1676: .P
1677: The default value for the limit can be set when PCRE is built; the default
1678: default is 10 million, which handles all but the most extreme cases. You can
1679: override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1680: block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1681: the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1682: PCRE_ERROR_MATCHLIMIT.
1683: .P
1684: A value for the match limit may also be supplied by an item at the start of a
1685: pattern of the form
1686: .sp
1687: (*LIMIT_MATCH=d)
1688: .sp
1689: where d is a decimal number. However, such a setting is ignored unless d is
1690: less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1691: is set, less than the default.
1692: .P
1693: The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1694: instead of limiting the total number of times that \fBmatch()\fP is called, it
1695: limits the depth of recursion. The recursion depth is a smaller number than the
1696: total number of calls, because not all calls to \fBmatch()\fP are recursive.
1697: This limit is of use only if it is set smaller than \fImatch_limit\fP.
1698: .P
1699: Limiting the recursion depth limits the amount of machine stack that can be
1700: used, or, when PCRE has been compiled to use memory on the heap instead of the
1701: stack, the amount of heap memory that can be used. This limit is not relevant,
1702: and is ignored, when matching is done using JIT compiled code.
1703: .P
1704: The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1705: built; the default default is the same value as the default for
1706: \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1707: with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1708: PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1709: is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1710: .P
1711: A value for the recursion limit may also be supplied by an item at the start of
1712: a pattern of the form
1713: .sp
1714: (*LIMIT_RECURSION=d)
1715: .sp
1716: where d is a decimal number. However, such a setting is ignored unless d is
1717: less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1718: is set, less than the default.
1719: .P
1720: The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1721: and is described in the
1722: .\" HREF
1723: \fBpcrecallout\fP
1724: .\"
1725: documentation.
1726: .P
1727: The \fItables\fP field is used to pass a character tables pointer to
1728: \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1729: pattern. A non-NULL value is stored with the compiled pattern only if custom
1730: tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1731: If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1732: internal tables to be used. This facility is helpful when re-using patterns
1733: that have been saved after compiling with an external set of tables, because
1734: the external tables might be at a different address when \fBpcre_exec()\fP is
1735: called. See the
1736: .\" HREF
1737: \fBpcreprecompile\fP
1738: .\"
1739: documentation for a discussion of saving compiled patterns for later use.
1740: .P
1741: If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1742: be set to point to a suitable variable. If the pattern contains any
1743: backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1744: a name to pass back, a pointer to the name string (zero terminated) is placed
1745: in the variable pointed to by the \fImark\fP field. The names are within the
1746: compiled pattern; if you wish to retain such a name you must copy it before
1747: freeing the memory of a compiled pattern. If there is no name to pass back, the
1748: variable pointed to by the \fImark\fP field is set to NULL. For details of the
1749: backtracking control verbs, see the section entitled
1750: .\" HTML <a href="pcrepattern#backtrackcontrol">
1751: .\" </a>
1752: "Backtracking control"
1753: .\"
1754: in the
1755: .\" HREF
1756: \fBpcrepattern\fP
1757: .\"
1758: documentation.
1759: .
1760: .
1761: .\" HTML <a name="execoptions"></a>
1762: .SS "Option bits for \fBpcre_exec()\fP"
1763: .rs
1764: .sp
1765: The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1766: zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1767: PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1768: PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1769: PCRE_PARTIAL_SOFT.
1770: .P
1771: If the pattern was successfully studied with one of the just-in-time (JIT)
1772: compile options, the only supported options for JIT execution are
1773: PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1774: PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1775: unsupported option is used, JIT execution is disabled and the normal
1776: interpretive code in \fBpcre_exec()\fP is run.
1777: .sp
1778: PCRE_ANCHORED
1779: .sp
1780: The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1781: matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1782: to be anchored by virtue of its contents, it cannot be made unachored at
1783: matching time.
1784: .sp
1785: PCRE_BSR_ANYCRLF
1786: PCRE_BSR_UNICODE
1787: .sp
1788: These options (which are mutually exclusive) control what the \eR escape
1789: sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1790: match any Unicode newline sequence. These options override the choice that was
1791: made or defaulted when the pattern was compiled.
1792: .sp
1793: PCRE_NEWLINE_CR
1794: PCRE_NEWLINE_LF
1795: PCRE_NEWLINE_CRLF
1796: PCRE_NEWLINE_ANYCRLF
1797: PCRE_NEWLINE_ANY
1798: .sp
1799: These options override the newline definition that was chosen or defaulted when
1800: the pattern was compiled. For details, see the description of
1801: \fBpcre_compile()\fP above. During matching, the newline choice affects the
1802: behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1803: the way the match position is advanced after a match failure for an unanchored
1804: pattern.
1805: .P
1806: When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1807: match attempt for an unanchored pattern fails when the current position is at a
1808: CRLF sequence, and the pattern contains no explicit matches for CR or LF
1809: characters, the match position is advanced by two characters instead of one, in
1810: other words, to after the CRLF.
1811: .P
1812: The above rule is a compromise that makes the most common cases work as
1813: expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1814: set), it does not match the string "\er\enA" because, after failing at the
1815: start, it skips both the CR and the LF before retrying. However, the pattern
1816: [\er\en]A does match that string, because it contains an explicit CR or LF
1817: reference, and so advances only by one character after the first failure.
1818: .P
1819: An explicit match for CR of LF is either a literal appearance of one of those
1820: characters, or one of the \er or \en escape sequences. Implicit matches such as
1821: [^X] do not count, nor does \es (which includes CR and LF in the characters
1822: that it matches).
1823: .P
1824: Notwithstanding the above, anomalous effects may still occur when CRLF is a
1825: valid newline sequence and explicit \er or \en escapes appear in the pattern.
1826: .sp
1827: PCRE_NOTBOL
1828: .sp
1829: This option specifies that first character of the subject string is not the
1830: beginning of a line, so the circumflex metacharacter should not match before
1831: it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1832: never to match. This option affects only the behaviour of the circumflex
1833: metacharacter. It does not affect \eA.
1834: .sp
1835: PCRE_NOTEOL
1836: .sp
1837: This option specifies that the end of the subject string is not the end of a
1838: line, so the dollar metacharacter should not match it nor (except in multiline
1839: mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1840: compile time) causes dollar never to match. This option affects only the
1841: behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1842: .sp
1843: PCRE_NOTEMPTY
1844: .sp
1845: An empty string is not considered to be a valid match if this option is set. If
1846: there are alternatives in the pattern, they are tried. If all the alternatives
1847: match the empty string, the entire match fails. For example, if the pattern
1848: .sp
1849: a?b?
1850: .sp
1851: is applied to a string not beginning with "a" or "b", it matches an empty
1852: string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1853: valid, so PCRE searches further into the string for occurrences of "a" or "b".
1854: .sp
1855: PCRE_NOTEMPTY_ATSTART
1856: .sp
1857: This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1858: the start of the subject is permitted. If the pattern is anchored, such a match
1859: can occur only if the pattern contains \eK.
1860: .P
1861: Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1862: does make a special case of a pattern match of the empty string within its
1863: \fBsplit()\fP function, and when using the /g modifier. It is possible to
1864: emulate Perl's behaviour after matching a null string by first trying the match
1865: again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1866: if that fails, by advancing the starting offset (see below) and trying an
1867: ordinary match again. There is some code that demonstrates how to do this in
1868: the
1869: .\" HREF
1870: \fBpcredemo\fP
1871: .\"
1872: sample program. In the most general case, you have to check to see if the
1873: newline convention recognizes CRLF as a newline, and if so, and the current
1874: character is CR followed by LF, advance the starting offset by two characters
1875: instead of one.
1876: .sp
1877: PCRE_NO_START_OPTIMIZE
1878: .sp
1879: There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1880: a match, in order to speed up the process. For example, if it is known that an
1881: unanchored match must start with a specific character, it searches the subject
1882: for that character, and fails immediately if it cannot find it, without
1883: actually running the main matching function. This means that a special item
1884: such as (*COMMIT) at the start of a pattern is not considered until after a
1885: suitable starting point for the match has been found. Also, when callouts or
1886: (*MARK) items are in use, these "start-up" optimizations can cause them to be
1887: skipped if the pattern is never actually used. The start-up optimizations are
1888: in effect a pre-scan of the subject that takes place before the pattern is run.
1889: .P
1890: The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1891: causing performance to suffer, but ensuring that in cases where the result is
1892: "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1893: are considered at every possible starting position in the subject string. If
1894: PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1895: time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1896: to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1897: always done using interpretively.
1898: .P
1899: Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1900: Consider the pattern
1901: .sp
1902: (*COMMIT)ABC
1903: .sp
1904: When this is compiled, PCRE records the fact that a match must start with the
1905: character "A". Suppose the subject string is "DEFABC". The start-up
1906: optimization scans along the subject, finds "A" and runs the first match
1907: attempt from there. The (*COMMIT) item means that the pattern must match the
1908: current starting position, which in this case, it does. However, if the same
1909: match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1910: subject string does not happen. The first match attempt is run starting from
1911: "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1912: the overall result is "no match". If the pattern is studied, more start-up
1913: optimizations may be used. For example, a minimum length for the subject may be
1914: recorded. Consider the pattern
1915: .sp
1916: (*MARK:A)(X|Y)
1917: .sp
1918: The minimum length for a match is one character. If the subject is "ABC", there
1919: will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1920: If the pattern is studied, the final attempt does not take place, because PCRE
1921: knows that the subject is too short, and so the (*MARK) is never encountered.
1922: In this case, studying the pattern does not affect the overall match result,
1923: which is still "no match", but it does affect the auxiliary information that is
1924: returned.
1925: .sp
1926: PCRE_NO_UTF8_CHECK
1927: .sp
1928: When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1929: string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1930: The entire string is checked before any other processing takes place. The value
1931: of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1932: UTF-8 character. There is a discussion about the
1933: .\" HTML <a href="pcreunicode.html#utf8strings">
1934: .\" </a>
1935: validity of UTF-8 strings
1936: .\"
1937: in the
1938: .\" HREF
1939: \fBpcreunicode\fP
1940: .\"
1941: page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1942: error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1943: truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1944: cases, information about the precise nature of the error may also be returned
1945: (see the descriptions of these errors in the section entitled \fIError return
1946: values from\fP \fBpcre_exec()\fP
1947: .\" HTML <a href="#errorlist">
1948: .\" </a>
1949: below).
1950: .\"
1951: If \fIstartoffset\fP contains a value that does not point to the start of a
1952: UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1953: returned.
1954: .P
1955: If you already know that your subject is valid, and you want to skip these
1956: checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1957: calling \fBpcre_exec()\fP. You might want to do this for the second and
1958: subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1959: all the matches in a single subject string. However, you should be sure that
1960: the value of \fIstartoffset\fP points to the start of a character (or the end
1961: of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1962: invalid string as a subject or an invalid value of \fIstartoffset\fP is
1963: undefined. Your program may crash.
1964: .sp
1965: PCRE_PARTIAL_HARD
1966: PCRE_PARTIAL_SOFT
1967: .sp
1968: These options turn on the partial matching feature. For backwards
1969: compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1970: occurs if the end of the subject string is reached successfully, but there are
1971: not enough subject characters to complete the match. If this happens when
1972: PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1973: testing any remaining alternatives. Only if no complete match can be found is
1974: PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1975: PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1976: but only if no complete match can be found.
1977: .P
1978: If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1979: partial match is found, \fBpcre_exec()\fP immediately returns
1980: PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1981: when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1982: important that an alternative complete match.
1983: .P
1984: In both cases, the portion of the string that was inspected when the partial
1985: match was found is set as the first matching string. There is a more detailed
1986: discussion of partial and multi-segment matching, with examples, in the
1987: .\" HREF
1988: \fBpcrepartial\fP
1989: .\"
1990: documentation.
1991: .
1992: .
1993: .SS "The string to be matched by \fBpcre_exec()\fP"
1994: .rs
1995: .sp
1996: The subject string is passed to \fBpcre_exec()\fP as a pointer in
1997: \fIsubject\fP, a length in \fIlength\fP, and a starting offset in
1998: \fIstartoffset\fP. The units for \fIlength\fP and \fIstartoffset\fP are bytes
1999: for the 8-bit library, 16-bit data items for the 16-bit library, and 32-bit
2000: data items for the 32-bit library.
2001: .P
2002: If \fIstartoffset\fP is negative or greater than the length of the subject,
2003: \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting offset is
2004: zero, the search for a match starts at the beginning of the subject, and this
2005: is by far the most common case. In UTF-8 or UTF-16 mode, the offset must point
2006: to the start of a character, or the end of the subject (in UTF-32 mode, one
2007: data unit equals one character, so all offsets are valid). Unlike the pattern
2008: string, the subject may contain binary zeroes.
2009: .P
2010: A non-zero starting offset is useful when searching for another match in the
2011: same subject by calling \fBpcre_exec()\fP again after a previous success.
2012: Setting \fIstartoffset\fP differs from just passing over a shortened string and
2013: setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
2014: lookbehind. For example, consider the pattern
2015: .sp
2016: \eBiss\eB
2017: .sp
2018: which finds occurrences of "iss" in the middle of words. (\eB matches only if
2019: the current position in the subject is not a word boundary.) When applied to
2020: the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
2021: occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
2022: subject, namely "issipi", it does not match, because \eB is always false at the
2023: start of the subject, which is deemed to be a word boundary. However, if
2024: \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
2025: set to 4, it finds the second occurrence of "iss" because it is able to look
2026: behind the starting point to discover that it is preceded by a letter.
2027: .P
2028: Finding all the matches in a subject is tricky when the pattern can match an
2029: empty string. It is possible to emulate Perl's /g behaviour by first trying the
2030: match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2031: PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2032: and trying an ordinary match again. There is some code that demonstrates how to
2033: do this in the
2034: .\" HREF
2035: \fBpcredemo\fP
2036: .\"
2037: sample program. In the most general case, you have to check to see if the
2038: newline convention recognizes CRLF as a newline, and if so, and the current
2039: character is CR followed by LF, advance the starting offset by two characters
2040: instead of one.
2041: .P
2042: If a non-zero starting offset is passed when the pattern is anchored, one
2043: attempt to match at the given offset is made. This can only succeed if the
2044: pattern does not require the match to be at the start of the subject.
2045: .
2046: .
2047: .SS "How \fBpcre_exec()\fP returns captured substrings"
2048: .rs
2049: .sp
2050: In general, a pattern matches a certain portion of the subject, and in
2051: addition, further substrings from the subject may be picked out by parts of the
2052: pattern. Following the usage in Jeffrey Friedl's book, this is called
2053: "capturing" in what follows, and the phrase "capturing subpattern" is used for
2054: a fragment of a pattern that picks out a substring. PCRE supports several other
2055: kinds of parenthesized subpattern that do not cause substrings to be captured.
2056: .P
2057: Captured substrings are returned to the caller via a vector of integers whose
2058: address is passed in \fIovector\fP. The number of elements in the vector is
2059: passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2060: argument is NOT the size of \fIovector\fP in bytes.
2061: .P
2062: The first two-thirds of the vector is used to pass back captured substrings,
2063: each substring using a pair of integers. The remaining third of the vector is
2064: used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2065: and is not available for passing back information. The number passed in
2066: \fIovecsize\fP should always be a multiple of three. If it is not, it is
2067: rounded down.
2068: .P
2069: When a match is successful, information about captured substrings is returned
2070: in pairs of integers, starting at the beginning of \fIovector\fP, and
2071: continuing up to two-thirds of its length at the most. The first element of
2072: each pair is set to the offset of the first character in a substring, and the
2073: second is set to the offset of the first character after the end of a
2074: substring. These values are always data unit offsets, even in UTF mode. They
2075: are byte offsets in the 8-bit library, 16-bit data item offsets in the 16-bit
2076: library, and 32-bit data item offsets in the 32-bit library. \fBNote\fP: they
2077: are not character counts.
2078: .P
2079: The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2080: portion of the subject string matched by the entire pattern. The next pair is
2081: used for the first capturing subpattern, and so on. The value returned by
2082: \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2083: For example, if two substrings have been captured, the returned value is 3. If
2084: there are no capturing subpatterns, the return value from a successful match is
2085: 1, indicating that just the first pair of offsets has been set.
2086: .P
2087: If a capturing subpattern is matched repeatedly, it is the last portion of the
2088: string that it matched that is returned.
2089: .P
2090: If the vector is too small to hold all the captured substring offsets, it is
2091: used as far as possible (up to two-thirds of its length), and the function
2092: returns a value of zero. If neither the actual string matched nor any captured
2093: substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2094: passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2095: back references and the \fIovector\fP is not big enough to remember the related
2096: substrings, PCRE has to get additional memory for use during matching. Thus it
2097: is usually advisable to supply an \fIovector\fP of reasonable size.
2098: .P
2099: There are some cases where zero is returned (indicating vector overflow) when
2100: in fact the vector is exactly the right size for the final match. For example,
2101: consider the pattern
2102: .sp
2103: (a)(?:(b)c|bd)
2104: .sp
2105: If a vector of 6 elements (allowing for only 1 captured substring) is given
2106: with subject string "abd", \fBpcre_exec()\fP will try to set the second
2107: captured string, thereby recording a vector overflow, before failing to match
2108: "c" and backing up to try the second alternative. The zero return, however,
2109: does correctly indicate that the maximum number of slots (namely 2) have been
2110: filled. In similar cases where there is temporary overflow, but the final
2111: number of used slots is actually less than the maximum, a non-zero value is
2112: returned.
2113: .P
2114: The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2115: subpatterns there are in a compiled pattern. The smallest size for
2116: \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2117: the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2118: .P
2119: It is possible for capturing subpattern number \fIn+1\fP to match some part of
2120: the subject when subpattern \fIn\fP has not been used at all. For example, if
2121: the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2122: function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2123: happens, both values in the offset pairs corresponding to unused subpatterns
2124: are set to -1.
2125: .P
2126: Offset values that correspond to unused subpatterns at the end of the
2127: expression are also set to -1. For example, if the string "abc" is matched
2128: against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2129: return from the function is 2, because the highest used capturing subpattern
2130: number is 1, and the offsets for for the second and third capturing subpatterns
2131: (assuming the vector is large enough, of course) are set to -1.
2132: .P
2133: \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2134: correspond to capturing parentheses in the pattern are never changed. That is,
2135: if a pattern contains \fIn\fP capturing parentheses, no more than
2136: \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2137: elements (in the first two-thirds) retain whatever values they previously had.
2138: .P
2139: Some convenience functions are provided for extracting the captured substrings
2140: as separate strings. These are described below.
2141: .
2142: .
2143: .\" HTML <a name="errorlist"></a>
2144: .SS "Error return values from \fBpcre_exec()\fP"
2145: .rs
2146: .sp
2147: If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2148: defined in the header file:
2149: .sp
2150: PCRE_ERROR_NOMATCH (-1)
2151: .sp
2152: The subject string did not match the pattern.
2153: .sp
2154: PCRE_ERROR_NULL (-2)
2155: .sp
2156: Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2157: NULL and \fIovecsize\fP was not zero.
2158: .sp
2159: PCRE_ERROR_BADOPTION (-3)
2160: .sp
2161: An unrecognized bit was set in the \fIoptions\fP argument.
2162: .sp
2163: PCRE_ERROR_BADMAGIC (-4)
2164: .sp
2165: PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2166: the case when it is passed a junk pointer and to detect when a pattern that was
2167: compiled in an environment of one endianness is run in an environment with the
2168: other endianness. This is the error that PCRE gives when the magic number is
2169: not present.
2170: .sp
2171: PCRE_ERROR_UNKNOWN_OPCODE (-5)
2172: .sp
2173: While running the pattern match, an unknown item was encountered in the
2174: compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2175: of the compiled pattern.
2176: .sp
2177: PCRE_ERROR_NOMEMORY (-6)
2178: .sp
2179: If a pattern contains back references, but the \fIovector\fP that is passed to
2180: \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2181: gets a block of memory at the start of matching to use for this purpose. If the
2182: call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2183: automatically freed at the end of matching.
2184: .P
2185: This error is also given if \fBpcre_stack_malloc()\fP fails in
2186: \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2187: \fB--disable-stack-for-recursion\fP.
2188: .sp
2189: PCRE_ERROR_NOSUBSTRING (-7)
2190: .sp
2191: This error is used by the \fBpcre_copy_substring()\fP,
2192: \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2193: below). It is never returned by \fBpcre_exec()\fP.
2194: .sp
2195: PCRE_ERROR_MATCHLIMIT (-8)
2196: .sp
2197: The backtracking limit, as specified by the \fImatch_limit\fP field in a
2198: \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2199: above.
2200: .sp
2201: PCRE_ERROR_CALLOUT (-9)
2202: .sp
2203: This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2204: use by callout functions that want to yield a distinctive error code. See the
2205: .\" HREF
2206: \fBpcrecallout\fP
2207: .\"
2208: documentation for details.
2209: .sp
2210: PCRE_ERROR_BADUTF8 (-10)
2211: .sp
2212: A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2213: and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2214: (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2215: UTF-8 character is placed in the first element, and a reason code is placed in
2216: the second element. The reason codes are listed in the
2217: .\" HTML <a href="#badutf8reasons">
2218: .\" </a>
2219: following section.
2220: .\"
2221: For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2222: truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2223: PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2224: .sp
2225: PCRE_ERROR_BADUTF8_OFFSET (-11)
2226: .sp
2227: The UTF-8 byte sequence that was passed as a subject was checked and found to
2228: be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2229: \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2230: end of the subject.
2231: .sp
2232: PCRE_ERROR_PARTIAL (-12)
2233: .sp
2234: The subject string did not match, but it did match partially. See the
2235: .\" HREF
2236: \fBpcrepartial\fP
2237: .\"
2238: documentation for details of partial matching.
2239: .sp
2240: PCRE_ERROR_BADPARTIAL (-13)
2241: .sp
2242: This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2243: option was used with a compiled pattern containing items that were not
2244: supported for partial matching. From release 8.00 onwards, there are no
2245: restrictions on partial matching.
2246: .sp
2247: PCRE_ERROR_INTERNAL (-14)
2248: .sp
2249: An unexpected internal error has occurred. This error could be caused by a bug
2250: in PCRE or by overwriting of the compiled pattern.
2251: .sp
2252: PCRE_ERROR_BADCOUNT (-15)
2253: .sp
2254: This error is given if the value of the \fIovecsize\fP argument is negative.
2255: .sp
2256: PCRE_ERROR_RECURSIONLIMIT (-21)
2257: .sp
2258: The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2259: field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2260: description above.
2261: .sp
2262: PCRE_ERROR_BADNEWLINE (-23)
2263: .sp
2264: An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2265: .sp
2266: PCRE_ERROR_BADOFFSET (-24)
2267: .sp
2268: The value of \fIstartoffset\fP was negative or greater than the length of the
2269: subject, that is, the value in \fIlength\fP.
2270: .sp
2271: PCRE_ERROR_SHORTUTF8 (-25)
2272: .sp
2273: This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2274: ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2275: Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2276: fact sufficient to detect this case, but this special error code for
2277: PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2278: retained for backwards compatibility.
2279: .sp
2280: PCRE_ERROR_RECURSELOOP (-26)
2281: .sp
2282: This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2283: the pattern. Specifically, it means that either the whole pattern or a
2284: subpattern has been called recursively for the second time at the same position
2285: in the subject string. Some simple patterns that might do this are detected and
2286: faulted at compile time, but more complicated cases, in particular mutual
2287: recursions between two different subpatterns, cannot be detected until run
2288: time.
2289: .sp
2290: PCRE_ERROR_JIT_STACKLIMIT (-27)
2291: .sp
2292: This error is returned when a pattern that was successfully studied using a
2293: JIT compile option is being matched, but the memory available for the
2294: just-in-time processing stack is not large enough. See the
2295: .\" HREF
2296: \fBpcrejit\fP
2297: .\"
2298: documentation for more details.
2299: .sp
2300: PCRE_ERROR_BADMODE (-28)
2301: .sp
2302: This error is given if a pattern that was compiled by the 8-bit library is
2303: passed to a 16-bit or 32-bit library function, or vice versa.
2304: .sp
2305: PCRE_ERROR_BADENDIANNESS (-29)
2306: .sp
2307: This error is given if a pattern that was compiled and saved is reloaded on a
2308: host with different endianness. The utility function
2309: \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2310: so that it runs on the new host.
2311: .sp
2312: PCRE_ERROR_JIT_BADOPTION
2313: .sp
2314: This error is returned when a pattern that was successfully studied using a JIT
2315: compile option is being matched, but the matching mode (partial or complete
2316: match) does not correspond to any JIT compilation mode. When the JIT fast path
2317: function is used, this error may be also given for invalid options. See the
2318: .\" HREF
2319: \fBpcrejit\fP
2320: .\"
2321: documentation for more details.
2322: .sp
2323: PCRE_ERROR_BADLENGTH (-32)
2324: .sp
2325: This error is given if \fBpcre_exec()\fP is called with a negative value for
2326: the \fIlength\fP argument.
2327: .P
2328: Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2329: .
2330: .
2331: .\" HTML <a name="badutf8reasons"></a>
2332: .SS "Reason codes for invalid UTF-8 strings"
2333: .rs
2334: .sp
2335: This section applies only to the 8-bit library. The corresponding information
2336: for the 16-bit and 32-bit libraries is given in the
2337: .\" HREF
2338: \fBpcre16\fP
2339: .\"
2340: and
2341: .\" HREF
2342: \fBpcre32\fP
2343: .\"
2344: pages.
2345: .P
2346: When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2347: PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2348: least 2, the offset of the start of the invalid UTF-8 character is placed in
2349: the first output vector element (\fIovector[0]\fP) and a reason code is placed
2350: in the second element (\fIovector[1]\fP). The reason codes are given names in
2351: the \fBpcre.h\fP header file:
2352: .sp
2353: PCRE_UTF8_ERR1
2354: PCRE_UTF8_ERR2
2355: PCRE_UTF8_ERR3
2356: PCRE_UTF8_ERR4
2357: PCRE_UTF8_ERR5
2358: .sp
2359: The string ends with a truncated UTF-8 character; the code specifies how many
2360: bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2361: no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2362: allows for up to 6 bytes, and this is checked first; hence the possibility of
2363: 4 or 5 missing bytes.
2364: .sp
2365: PCRE_UTF8_ERR6
2366: PCRE_UTF8_ERR7
2367: PCRE_UTF8_ERR8
2368: PCRE_UTF8_ERR9
2369: PCRE_UTF8_ERR10
2370: .sp
2371: The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2372: character do not have the binary value 0b10 (that is, either the most
2373: significant bit is 0, or the next bit is 1).
2374: .sp
2375: PCRE_UTF8_ERR11
2376: PCRE_UTF8_ERR12
2377: .sp
2378: A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2379: these code points are excluded by RFC 3629.
2380: .sp
2381: PCRE_UTF8_ERR13
2382: .sp
2383: A 4-byte character has a value greater than 0x10fff; these code points are
2384: excluded by RFC 3629.
2385: .sp
2386: PCRE_UTF8_ERR14
2387: .sp
2388: A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2389: code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2390: from UTF-8.
2391: .sp
2392: PCRE_UTF8_ERR15
2393: PCRE_UTF8_ERR16
2394: PCRE_UTF8_ERR17
2395: PCRE_UTF8_ERR18
2396: PCRE_UTF8_ERR19
2397: .sp
2398: A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2399: value that can be represented by fewer bytes, which is invalid. For example,
2400: the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2401: one byte.
2402: .sp
2403: PCRE_UTF8_ERR20
2404: .sp
2405: The two most significant bits of the first byte of a character have the binary
2406: value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2407: byte can only validly occur as the second or subsequent byte of a multi-byte
2408: character.
2409: .sp
2410: PCRE_UTF8_ERR21
2411: .sp
2412: The first byte of a character has the value 0xfe or 0xff. These values can
2413: never occur in a valid UTF-8 string.
2414: .sp
2415: PCRE_UTF8_ERR22
2416: .sp
2417: This error code was formerly used when the presence of a so-called
2418: "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2419: such characters should not cause a string to be rejected, and so this code is
2420: no longer in use and is never returned.
2421: .
2422: .
2423: .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2424: .rs
2425: .sp
2426: .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2427: .ti +5n
2428: .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2429: .ti +5n
2430: .B int \fIbuffersize\fP);
2431: .PP
2432: .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2433: .ti +5n
2434: .B int \fIstringcount\fP, int \fIstringnumber\fP,
2435: .ti +5n
2436: .B const char **\fIstringptr\fP);
2437: .PP
2438: .B int pcre_get_substring_list(const char *\fIsubject\fP,
2439: .ti +5n
2440: .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2441: .PP
2442: Captured substrings can be accessed directly by using the offsets returned by
2443: \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2444: \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2445: \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2446: as new, separate, zero-terminated strings. These functions identify substrings
2447: by number. The next section describes functions for extracting named
2448: substrings.
2449: .P
2450: A substring that contains a binary zero is correctly extracted and has a
2451: further zero added on the end, but the result is not, of course, a C string.
2452: However, you can process such a string by referring to the length that is
2453: returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2454: Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2455: for handling strings containing binary zeros, because the end of the final
2456: string is not independently indicated.
2457: .P
2458: The first three arguments are the same for all three of these functions:
2459: \fIsubject\fP is the subject string that has just been successfully matched,
2460: \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2461: \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2462: captured by the match, including the substring that matched the entire regular
2463: expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2464: than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2465: space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2466: number of elements in the vector divided by three.
2467: .P
2468: The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2469: extract a single substring, whose number is given as \fIstringnumber\fP. A
2470: value of zero extracts the substring that matched the entire pattern, whereas
2471: higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2472: the string is placed in \fIbuffer\fP, whose length is given by
2473: \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2474: obtained via \fBpcre_malloc\fP, and its address is returned via
2475: \fIstringptr\fP. The yield of the function is the length of the string, not
2476: including the terminating zero, or one of these error codes:
2477: .sp
2478: PCRE_ERROR_NOMEMORY (-6)
2479: .sp
2480: The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2481: memory failed for \fBpcre_get_substring()\fP.
2482: .sp
2483: PCRE_ERROR_NOSUBSTRING (-7)
2484: .sp
2485: There is no substring whose number is \fIstringnumber\fP.
2486: .P
2487: The \fBpcre_get_substring_list()\fP function extracts all available substrings
2488: and builds a list of pointers to them. All this is done in a single block of
2489: memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2490: is returned via \fIlistptr\fP, which is also the start of the list of string
2491: pointers. The end of the list is marked by a NULL pointer. The yield of the
2492: function is zero if all went well, or the error code
2493: .sp
2494: PCRE_ERROR_NOMEMORY (-6)
2495: .sp
2496: if the attempt to get the memory block failed.
2497: .P
2498: When any of these functions encounter a substring that is unset, which can
2499: happen when capturing subpattern number \fIn+1\fP matches some part of the
2500: subject, but subpattern \fIn\fP has not been used at all, they return an empty
2501: string. This can be distinguished from a genuine zero-length substring by
2502: inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2503: substrings.
2504: .P
2505: The two convenience functions \fBpcre_free_substring()\fP and
2506: \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2507: a previous call of \fBpcre_get_substring()\fP or
2508: \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2509: the function pointed to by \fBpcre_free\fP, which of course could be called
2510: directly from a C program. However, PCRE is used in some situations where it is
2511: linked via a special interface to another programming language that cannot use
2512: \fBpcre_free\fP directly; it is for these cases that the functions are
2513: provided.
2514: .
2515: .
2516: .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2517: .rs
2518: .sp
2519: .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2520: .ti +5n
2521: .B const char *\fIname\fP);
2522: .PP
2523: .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2524: .ti +5n
2525: .B const char *\fIsubject\fP, int *\fIovector\fP,
2526: .ti +5n
2527: .B int \fIstringcount\fP, const char *\fIstringname\fP,
2528: .ti +5n
2529: .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2530: .PP
2531: .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2532: .ti +5n
2533: .B const char *\fIsubject\fP, int *\fIovector\fP,
2534: .ti +5n
2535: .B int \fIstringcount\fP, const char *\fIstringname\fP,
2536: .ti +5n
2537: .B const char **\fIstringptr\fP);
2538: .PP
2539: To extract a substring by name, you first have to find associated number.
2540: For example, for this pattern
2541: .sp
2542: (a+)b(?<xxx>\ed+)...
2543: .sp
2544: the number of the subpattern called "xxx" is 2. If the name is known to be
2545: unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2546: calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2547: pattern, and the second is the name. The yield of the function is the
2548: subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2549: that name.
2550: .P
2551: Given the number, you can extract the substring directly, or use one of the
2552: functions described in the previous section. For convenience, there are also
2553: two functions that do the whole job.
2554: .P
2555: Most of the arguments of \fBpcre_copy_named_substring()\fP and
2556: \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2557: functions that extract by number. As these are described in the previous
2558: section, they are not re-described here. There are just two differences:
2559: .P
2560: First, instead of a substring number, a substring name is given. Second, there
2561: is an extra argument, given at the start, which is a pointer to the compiled
2562: pattern. This is needed in order to gain access to the name-to-number
2563: translation table.
2564: .P
2565: These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2566: then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2567: appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2568: the behaviour may not be what you want (see the next section).
2569: .P
2570: \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2571: subpatterns with the same number, as described in the
2572: .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2573: .\" </a>
2574: section on duplicate subpattern numbers
2575: .\"
2576: in the
2577: .\" HREF
2578: \fBpcrepattern\fP
2579: .\"
2580: page, you cannot use names to distinguish the different subpatterns, because
2581: names are not included in the compiled code. The matching process uses only
2582: numbers. For this reason, the use of different names for subpatterns of the
2583: same number causes an error at compile time.
2584: .
2585: .
2586: .SH "DUPLICATE SUBPATTERN NAMES"
2587: .rs
2588: .sp
2589: .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2590: .ti +5n
2591: .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2592: .PP
2593: When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2594: are not required to be unique. (Duplicate names are always allowed for
2595: subpatterns with the same number, created by using the (?| feature. Indeed, if
2596: such subpatterns are named, they are required to use the same names.)
2597: .P
2598: Normally, patterns with duplicate names are such that in any one match, only
2599: one of the named subpatterns participates. An example is shown in the
2600: .\" HREF
2601: \fBpcrepattern\fP
2602: .\"
2603: documentation.
2604: .P
2605: When duplicates are present, \fBpcre_copy_named_substring()\fP and
2606: \fBpcre_get_named_substring()\fP return the first substring corresponding to
2607: the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2608: returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2609: returns one of the numbers that are associated with the name, but it is not
2610: defined which it is.
2611: .P
2612: If you want to get full details of all captured substrings for a given name,
2613: you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2614: argument is the compiled pattern, and the second is the name. The third and
2615: fourth are pointers to variables which are updated by the function. After it
2616: has run, they point to the first and last entries in the name-to-number table
2617: for the given name. The function itself returns the length of each entry, or
2618: PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2619: described above in the section entitled \fIInformation about a pattern\fP
2620: .\" HTML <a href="#infoaboutpattern">
2621: .\" </a>
2622: above.
2623: .\"
2624: Given all the relevant entries for the name, you can extract each of their
2625: numbers, and hence the captured data, if any.
2626: .
2627: .
2628: .SH "FINDING ALL POSSIBLE MATCHES"
2629: .rs
2630: .sp
2631: The traditional matching function uses a similar algorithm to Perl, which stops
2632: when it finds the first match, starting at a given point in the subject. If you
2633: want to find all possible matches, or the longest possible match, consider
2634: using the alternative matching function (see below) instead. If you cannot use
2635: the alternative function, but still need to find all possible matches, you
2636: can kludge it up by making use of the callout facility, which is described in
2637: the
2638: .\" HREF
2639: \fBpcrecallout\fP
2640: .\"
2641: documentation.
2642: .P
2643: What you have to do is to insert a callout right at the end of the pattern.
2644: When your callout function is called, extract and save the current matched
2645: substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2646: other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2647: will yield PCRE_ERROR_NOMATCH.
2648: .
2649: .
2650: .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2651: .rs
2652: .sp
2653: Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2654: stack, which in certain environments can be rather limited in size. Some users
2655: find it helpful to have an estimate of the amount of stack that is used by
2656: \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2657: .\" HREF
2658: \fBpcrestack\fP
2659: .\"
2660: documentation. The estimate that is output by \fBpcretest\fP when called with
2661: the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2662: the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2663: .P
2664: Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2665: the negative error code PCRE_ERROR_NULL, but with this special combination of
2666: arguments, it returns instead a negative number whose absolute value is the
2667: approximate stack frame size in bytes. (A negative number is used so that it is
2668: clear that no match has happened.) The value is approximate because in some
2669: cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2670: additional variables on the stack.
2671: .P
2672: If PCRE has been compiled to use the heap instead of the stack for recursion,
2673: the value returned is the size of each block that is obtained from the heap.
2674: .
2675: .
2676: .\" HTML <a name="dfamatch"></a>
2677: .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2678: .rs
2679: .sp
2680: .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2681: .ti +5n
2682: .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2683: .ti +5n
2684: .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2685: .ti +5n
2686: .B int *\fIworkspace\fP, int \fIwscount\fP);
2687: .P
2688: The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2689: a compiled pattern, using a matching algorithm that scans the subject string
2690: just once, and does not backtrack. This has different characteristics to the
2691: normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2692: patterns are not supported. Nevertheless, there are times when this kind of
2693: matching can be useful. For a discussion of the two matching algorithms, and a
2694: list of features that \fBpcre_dfa_exec()\fP does not support, see the
2695: .\" HREF
2696: \fBpcrematching\fP
2697: .\"
2698: documentation.
2699: .P
2700: The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2701: \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2702: different way, and this is described below. The other common arguments are used
2703: in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2704: here.
2705: .P
2706: The two additional arguments provide workspace for the function. The workspace
2707: vector should contain at least 20 elements. It is used for keeping track of
2708: multiple paths through the pattern tree. More workspace will be needed for
2709: patterns and subjects where there are a lot of potential matches.
2710: .P
2711: Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2712: .sp
2713: int rc;
2714: int ovector[10];
2715: int wspace[20];
2716: rc = pcre_dfa_exec(
2717: re, /* result of pcre_compile() */
2718: NULL, /* we didn't study the pattern */
2719: "some string", /* the subject string */
2720: 11, /* the length of the subject string */
2721: 0, /* start at offset 0 in the subject */
2722: 0, /* default options */
2723: ovector, /* vector of integers for substring information */
2724: 10, /* number of elements (NOT size in bytes) */
2725: wspace, /* working space vector */
2726: 20); /* number of elements (NOT size in bytes) */
2727: .
2728: .SS "Option bits for \fBpcre_dfa_exec()\fP"
2729: .rs
2730: .sp
2731: The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2732: zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2733: PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2734: PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2735: PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2736: All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2737: so their description is not repeated here.
2738: .sp
2739: PCRE_PARTIAL_HARD
2740: PCRE_PARTIAL_SOFT
2741: .sp
2742: These have the same general effect as they do for \fBpcre_exec()\fP, but the
2743: details are slightly different. When PCRE_PARTIAL_HARD is set for
2744: \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2745: is reached and there is still at least one matching possibility that requires
2746: additional characters. This happens even if some complete matches have also
2747: been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2748: is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2749: there have been no complete matches, but there is still at least one matching
2750: possibility. The portion of the string that was inspected when the longest
2751: partial match was found is set as the first matching string in both cases.
2752: There is a more detailed discussion of partial and multi-segment matching, with
2753: examples, in the
2754: .\" HREF
2755: \fBpcrepartial\fP
2756: .\"
2757: documentation.
2758: .sp
2759: PCRE_DFA_SHORTEST
2760: .sp
2761: Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2762: soon as it has found one match. Because of the way the alternative algorithm
2763: works, this is necessarily the shortest possible match at the first possible
2764: matching point in the subject string.
2765: .sp
2766: PCRE_DFA_RESTART
2767: .sp
2768: When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2769: again, with additional subject characters, and have it continue with the same
2770: match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2771: \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2772: before because data about the match so far is left in them after a partial
2773: match. There is more discussion of this facility in the
2774: .\" HREF
2775: \fBpcrepartial\fP
2776: .\"
2777: documentation.
2778: .
2779: .
2780: .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2781: .rs
2782: .sp
2783: When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2784: substring in the subject. Note, however, that all the matches from one run of
2785: the function start at the same point in the subject. The shorter matches are
2786: all initial substrings of the longer matches. For example, if the pattern
2787: .sp
2788: <.*>
2789: .sp
2790: is matched against the string
2791: .sp
2792: This is <something> <something else> <something further> no more
2793: .sp
2794: the three matched strings are
2795: .sp
2796: <something>
2797: <something> <something else>
2798: <something> <something else> <something further>
2799: .sp
2800: On success, the yield of the function is a number greater than zero, which is
2801: the number of matched substrings. The substrings themselves are returned in
2802: \fIovector\fP. Each string uses two elements; the first is the offset to the
2803: start, and the second is the offset to the end. In fact, all the strings have
2804: the same start offset. (Space could have been saved by giving this only once,
2805: but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2806: returns data, even though the meaning of the strings is different.)
2807: .P
2808: The strings are returned in reverse order of length; that is, the longest
2809: matching string is given first. If there were too many matches to fit into
2810: \fIovector\fP, the yield of the function is zero, and the vector is filled with
2811: the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2812: the entire \fIovector\fP for returning matched strings.
2813: .
2814: .
2815: .SS "Error returns from \fBpcre_dfa_exec()\fP"
2816: .rs
2817: .sp
2818: The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2819: Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2820: described
2821: .\" HTML <a href="#errorlist">
2822: .\" </a>
2823: above.
2824: .\"
2825: There are in addition the following errors that are specific to
2826: \fBpcre_dfa_exec()\fP:
2827: .sp
2828: PCRE_ERROR_DFA_UITEM (-16)
2829: .sp
2830: This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2831: that it does not support, for instance, the use of \eC or a back reference.
2832: .sp
2833: PCRE_ERROR_DFA_UCOND (-17)
2834: .sp
2835: This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2836: uses a back reference for the condition, or a test for recursion in a specific
2837: group. These are not supported.
2838: .sp
2839: PCRE_ERROR_DFA_UMLIMIT (-18)
2840: .sp
2841: This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2842: block that contains a setting of the \fImatch_limit\fP or
2843: \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2844: meaningless for DFA matching).
2845: .sp
2846: PCRE_ERROR_DFA_WSSIZE (-19)
2847: .sp
2848: This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2849: \fIworkspace\fP vector.
2850: .sp
2851: PCRE_ERROR_DFA_RECURSE (-20)
2852: .sp
2853: When a recursive subpattern is processed, the matching function calls itself
2854: recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2855: error is given if the output vector is not large enough. This should be
2856: extremely rare, as a vector of size 1000 is used.
2857: .sp
2858: PCRE_ERROR_DFA_BADRESTART (-30)
2859: .sp
2860: When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2861: some plausibility checks are made on the contents of the workspace, which
2862: should contain data about the previous partial match. If any of these checks
2863: fail, this error is given.
2864: .
2865: .
2866: .SH "SEE ALSO"
2867: .rs
2868: .sp
2869: \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2870: \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2871: \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2872: \fBpcrestack\fP(3).
2873: .
2874: .
2875: .SH AUTHOR
2876: .rs
2877: .sp
2878: .nf
2879: Philip Hazel
2880: University Computing Service
2881: Cambridge CB2 3QH, England.
2882: .fi
2883: .
2884: .
2885: .SH REVISION
2886: .rs
2887: .sp
2888: .nf
2889: Last updated: 12 May 2013
2890: Copyright (c) 1997-2013 University of Cambridge.
2891: .fi
FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>
![[BACK]](/icons/cvsweb/back.gif){kind=icon}
Return to pcreapi.3 CVS log ![[TXT]](/icons/cvsweb/text.gif){kind=icon}
![[DIR]](/icons/cvsweb/dir.gif){kind=icon}
Up to [ELWIX - Embedded LightWeight unIX -] / embedaddon / pcre / doc