1: .TH PCREAPI 3 "12 November 2013" "PCRE 8.34"
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: .nf
12: .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
13: .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
14: .B " const unsigned char *\fItableptr\fP);"
15: .sp
16: .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
17: .B " int *\fIerrorcodeptr\fP,"
18: .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
19: .B " const unsigned char *\fItableptr\fP);"
20: .sp
21: .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
22: .B " const char **\fIerrptr\fP);"
23: .sp
24: .B void pcre_free_study(pcre_extra *\fIextra\fP);
25: .sp
26: .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
27: .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
28: .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
29: .sp
30: .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
31: .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
32: .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
33: .B " int *\fIworkspace\fP, int \fIwscount\fP);"
34: .fi
35: .
36: .
37: .SH "PCRE NATIVE API STRING EXTRACTION FUNCTIONS"
38: .rs
39: .sp
40: .nf
41: .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
42: .B " const char *\fIsubject\fP, int *\fIovector\fP,"
43: .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
44: .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
45: .sp
46: .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
47: .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
48: .B " int \fIbuffersize\fP);"
49: .sp
50: .B int pcre_get_named_substring(const pcre *\fIcode\fP,
51: .B " const char *\fIsubject\fP, int *\fIovector\fP,"
52: .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
53: .B " const char **\fIstringptr\fP);"
54: .sp
55: .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
56: .B " const char *\fIname\fP);"
57: .sp
58: .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
59: .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
60: .sp
61: .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
62: .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
63: .B " const char **\fIstringptr\fP);"
64: .sp
65: .B int pcre_get_substring_list(const char *\fIsubject\fP,
66: .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
67: .sp
68: .B void pcre_free_substring(const char *\fIstringptr\fP);
69: .sp
70: .B void pcre_free_substring_list(const char **\fIstringptr\fP);
71: .fi
72: .
73: .
74: .SH "PCRE NATIVE API AUXILIARY FUNCTIONS"
75: .rs
76: .sp
77: .nf
78: .B int pcre_jit_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
79: .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
80: .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
81: .B " pcre_jit_stack *\fIjstack\fP);"
82: .sp
83: .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
84: .sp
85: .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
86: .sp
87: .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
88: .B " pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);"
89: .sp
90: .B const unsigned char *pcre_maketables(void);
91: .sp
92: .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
93: .B " int \fIwhat\fP, void *\fIwhere\fP);"
94: .sp
95: .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
96: .sp
97: .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
98: .sp
99: .B const char *pcre_version(void);
100: .sp
101: .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
102: .B " pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);"
103: .fi
104: .
105: .
106: .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
107: .rs
108: .sp
109: .nf
110: .B void *(*pcre_malloc)(size_t);
111: .sp
112: .B void (*pcre_free)(void *);
113: .sp
114: .B void *(*pcre_stack_malloc)(size_t);
115: .sp
116: .B void (*pcre_stack_free)(void *);
117: .sp
118: .B int (*pcre_callout)(pcre_callout_block *);
119: .fi
120: .
121: .
122: .SH "PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES"
123: .rs
124: .sp
125: As well as support for 8-bit character strings, PCRE also supports 16-bit
126: strings (from release 8.30) and 32-bit strings (from release 8.32), by means of
127: two additional libraries. They can be built as well as, or instead of, the
128: 8-bit library. To avoid too much complication, this document describes the
129: 8-bit versions of the functions, with only occasional references to the 16-bit
130: and 32-bit libraries.
131: .P
132: The 16-bit and 32-bit functions operate in the same way as their 8-bit
133: counterparts; they just use different data types for their arguments and
134: results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
135: \fBpcre_\fP. For every option that has UTF8 in its name (for example,
136: PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
137: by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
138: 16-bit and 32-bit option names define the same bit values.
139: .P
140: References to bytes and UTF-8 in this document should be read as references to
141: 16-bit data units and UTF-16 when using the 16-bit library, or 32-bit data
142: units and UTF-32 when using the 32-bit library, unless specified otherwise.
143: More details of the specific differences for the 16-bit and 32-bit libraries
144: are given in the
145: .\" HREF
146: \fBpcre16\fP
147: .\"
148: and
149: .\" HREF
150: \fBpcre32\fP
151: .\"
152: pages.
153: .
154: .
155: .SH "PCRE API OVERVIEW"
156: .rs
157: .sp
158: PCRE has its own native API, which is described in this document. There are
159: also some wrapper functions (for the 8-bit library only) that correspond to the
160: POSIX regular expression API, but they do not give access to all the
161: functionality. They are described in the
162: .\" HREF
163: \fBpcreposix\fP
164: .\"
165: documentation. Both of these APIs define a set of C function calls. A C++
166: wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
167: documented in the
168: .\" HREF
169: \fBpcrecpp\fP
170: .\"
171: page.
172: .P
173: The native API C function prototypes are defined in the header file
174: \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
175: \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
176: command for linking an application that uses PCRE. The header file defines the
177: macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
178: for the library. Applications can use these to include support for different
179: releases of PCRE.
180: .P
181: In a Windows environment, if you want to statically link an application program
182: against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
183: including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
184: \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
185: \fB__declspec(dllimport)\fP, with unwanted results.
186: .P
187: The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
188: and \fBpcre_exec()\fP are used for compiling and matching regular expressions
189: in a Perl-compatible manner. A sample program that demonstrates the simplest
190: way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
191: source distribution. A listing of this program is given in the
192: .\" HREF
193: \fBpcredemo\fP
194: .\"
195: documentation, and the
196: .\" HREF
197: \fBpcresample\fP
198: .\"
199: documentation describes how to compile and run it.
200: .P
201: Just-in-time compiler support is an optional feature of PCRE that can be built
202: in appropriate hardware environments. It greatly speeds up the matching
203: performance of many patterns. Simple programs can easily request that it be
204: used if available, by setting an option that is ignored when it is not
205: relevant. More complicated programs might need to make use of the functions
206: \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
207: \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
208: .P
209: From release 8.32 there is also a direct interface for JIT execution, which
210: gives improved performance. The JIT-specific functions are discussed in the
211: .\" HREF
212: \fBpcrejit\fP
213: .\"
214: documentation.
215: .P
216: A second matching function, \fBpcre_dfa_exec()\fP, which is not
217: Perl-compatible, is also provided. This uses a different algorithm for the
218: matching. The alternative algorithm finds all possible matches (at a given
219: point in the subject), and scans the subject just once (unless there are
220: lookbehind assertions). However, this algorithm does not return captured
221: substrings. A description of the two matching algorithms and their advantages
222: and disadvantages is given in the
223: .\" HREF
224: \fBpcrematching\fP
225: .\"
226: documentation.
227: .P
228: In addition to the main compiling and matching functions, there are convenience
229: functions for extracting captured substrings from a subject string that is
230: matched by \fBpcre_exec()\fP. They are:
231: .sp
232: \fBpcre_copy_substring()\fP
233: \fBpcre_copy_named_substring()\fP
234: \fBpcre_get_substring()\fP
235: \fBpcre_get_named_substring()\fP
236: \fBpcre_get_substring_list()\fP
237: \fBpcre_get_stringnumber()\fP
238: \fBpcre_get_stringtable_entries()\fP
239: .sp
240: \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
241: provided, to free the memory used for extracted strings.
242: .P
243: The function \fBpcre_maketables()\fP is used to build a set of character tables
244: in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
245: or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
246: specialist use. Most commonly, no special tables are passed, in which case
247: internal tables that are generated when PCRE is built are used.
248: .P
249: The function \fBpcre_fullinfo()\fP is used to find out information about a
250: compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
251: string containing the version of PCRE and its date of release.
252: .P
253: The function \fBpcre_refcount()\fP maintains a reference count in a data block
254: containing a compiled pattern. This is provided for the benefit of
255: object-oriented applications.
256: .P
257: The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
258: the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
259: respectively. PCRE calls the memory management functions via these variables,
260: so a calling program can replace them if it wishes to intercept the calls. This
261: should be done before calling any PCRE functions.
262: .P
263: The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
264: indirections to memory management functions. These special functions are used
265: only when PCRE is compiled to use the heap for remembering data, instead of
266: recursive function calls, when running the \fBpcre_exec()\fP function. See the
267: .\" HREF
268: \fBpcrebuild\fP
269: .\"
270: documentation for details of how to do this. It is a non-standard way of
271: building PCRE, for use in environments that have limited stacks. Because of the
272: greater use of memory management, it runs more slowly. Separate functions are
273: provided so that special-purpose external code can be used for this case. When
274: used, these functions are always called in a stack-like manner (last obtained,
275: first freed), and always for memory blocks of the same size. There is a
276: discussion about PCRE's stack usage in the
277: .\" HREF
278: \fBpcrestack\fP
279: .\"
280: documentation.
281: .P
282: The global variable \fBpcre_callout\fP initially contains NULL. It can be set
283: by the caller to a "callout" function, which PCRE will then call at specified
284: points during a matching operation. Details are given in the
285: .\" HREF
286: \fBpcrecallout\fP
287: .\"
288: documentation.
289: .
290: .
291: .\" HTML <a name="newlines"></a>
292: .SH NEWLINES
293: .rs
294: .sp
295: PCRE supports five different conventions for indicating line breaks in
296: strings: a single CR (carriage return) character, a single LF (linefeed)
297: character, the two-character sequence CRLF, any of the three preceding, or any
298: Unicode newline sequence. The Unicode newline sequences are the three just
299: mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
300: U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
301: (paragraph separator, U+2029).
302: .P
303: Each of the first three conventions is used by at least one operating system as
304: its standard newline sequence. When PCRE is built, a default can be specified.
305: The default default is LF, which is the Unix standard. When PCRE is run, the
306: default can be overridden, either when a pattern is compiled, or when it is
307: matched.
308: .P
309: At compile time, the newline convention can be specified by the \fIoptions\fP
310: argument of \fBpcre_compile()\fP, or it can be specified by special text at the
311: start of the pattern itself; this overrides any other settings. See the
312: .\" HREF
313: \fBpcrepattern\fP
314: .\"
315: page for details of the special character sequences.
316: .P
317: In the PCRE documentation the word "newline" is used to mean "the character or
318: pair of characters that indicate a line break". The choice of newline
319: convention affects the handling of the dot, circumflex, and dollar
320: metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
321: recognized line ending sequence, the match position advancement for a
322: non-anchored pattern. There is more detail about this in the
323: .\" HTML <a href="#execoptions">
324: .\" </a>
325: section on \fBpcre_exec()\fP options
326: .\"
327: below.
328: .P
329: The choice of newline convention does not affect the interpretation of
330: the \en or \er escape sequences, nor does it affect what \eR matches, which is
331: controlled in a similar way, but by separate options.
332: .
333: .
334: .SH MULTITHREADING
335: .rs
336: .sp
337: The PCRE functions can be used in multi-threading applications, with the
338: proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
339: \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
340: callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
341: .P
342: The compiled form of a regular expression is not altered during matching, so
343: the same compiled pattern can safely be used by several threads at once.
344: .P
345: If the just-in-time optimization feature is being used, it needs separate
346: memory stack areas for each thread. See the
347: .\" HREF
348: \fBpcrejit\fP
349: .\"
350: documentation for more details.
351: .
352: .
353: .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
354: .rs
355: .sp
356: The compiled form of a regular expression can be saved and re-used at a later
357: time, possibly by a different program, and even on a host other than the one on
358: which it was compiled. Details are given in the
359: .\" HREF
360: \fBpcreprecompile\fP
361: .\"
362: documentation, which includes a description of the
363: \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
364: expression with one version of PCRE for use with a different version is not
365: guaranteed to work and may cause crashes.
366: .
367: .
368: .SH "CHECKING BUILD-TIME OPTIONS"
369: .rs
370: .sp
371: .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
372: .PP
373: The function \fBpcre_config()\fP makes it possible for a PCRE client to
374: discover which optional features have been compiled into the PCRE library. The
375: .\" HREF
376: \fBpcrebuild\fP
377: .\"
378: documentation has more details about these optional features.
379: .P
380: The first argument for \fBpcre_config()\fP is an integer, specifying which
381: information is required; the second argument is a pointer to a variable into
382: which the information is placed. The returned value is zero on success, or the
383: negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
384: not recognized. The following information is available:
385: .sp
386: PCRE_CONFIG_UTF8
387: .sp
388: The output is an integer that is set to one if UTF-8 support is available;
389: otherwise it is set to zero. This value should normally be given to the 8-bit
390: version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
391: or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
392: .sp
393: PCRE_CONFIG_UTF16
394: .sp
395: The output is an integer that is set to one if UTF-16 support is available;
396: otherwise it is set to zero. This value should normally be given to the 16-bit
397: version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
398: or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
399: .sp
400: PCRE_CONFIG_UTF32
401: .sp
402: The output is an integer that is set to one if UTF-32 support is available;
403: otherwise it is set to zero. This value should normally be given to the 32-bit
404: version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
405: or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
406: .sp
407: PCRE_CONFIG_UNICODE_PROPERTIES
408: .sp
409: The output is an integer that is set to one if support for Unicode character
410: properties is available; otherwise it is set to zero.
411: .sp
412: PCRE_CONFIG_JIT
413: .sp
414: The output is an integer that is set to one if support for just-in-time
415: compiling is available; otherwise it is set to zero.
416: .sp
417: PCRE_CONFIG_JITTARGET
418: .sp
419: The output is a pointer to a zero-terminated "const char *" string. If JIT
420: support is available, the string contains the name of the architecture for
421: which the JIT compiler is configured, for example "x86 32bit (little endian +
422: unaligned)". If JIT support is not available, the result is NULL.
423: .sp
424: PCRE_CONFIG_NEWLINE
425: .sp
426: The output is an integer whose value specifies the default character sequence
427: that is recognized as meaning "newline". The values that are supported in
428: ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
429: ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
430: same values. However, the value for LF is normally 21, though some EBCDIC
431: environments use 37. The corresponding values for CRLF are 3349 and 3365. The
432: default should normally correspond to the standard sequence for your operating
433: system.
434: .sp
435: PCRE_CONFIG_BSR
436: .sp
437: The output is an integer whose value indicates what character sequences the \eR
438: escape sequence matches by default. A value of 0 means that \eR matches any
439: Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
440: or CRLF. The default can be overridden when a pattern is compiled or matched.
441: .sp
442: PCRE_CONFIG_LINK_SIZE
443: .sp
444: The output is an integer that contains the number of bytes used for internal
445: linkage in compiled regular expressions. For the 8-bit library, the value can
446: be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
447: a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
448: still a number of bytes. The default value of 2 is sufficient for all but the
449: most massive patterns, since it allows the compiled pattern to be up to 64K in
450: size. Larger values allow larger regular expressions to be compiled, at the
451: expense of slower matching.
452: .sp
453: PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
454: .sp
455: The output is an integer that contains the threshold above which the POSIX
456: interface uses \fBmalloc()\fP for output vectors. Further details are given in
457: the
458: .\" HREF
459: \fBpcreposix\fP
460: .\"
461: documentation.
462: .sp
463: PCRE_CONFIG_PARENS_LIMIT
464: .sp
465: The output is a long integer that gives the maximum depth of nesting of
466: parentheses (of any kind) in a pattern. This limit is imposed to cap the amount
467: of system stack used when a pattern is compiled. It is specified when PCRE is
468: built; the default is 250.
469: .sp
470: PCRE_CONFIG_MATCH_LIMIT
471: .sp
472: The output is a long integer that gives the default limit for the number of
473: internal matching function calls in a \fBpcre_exec()\fP execution. Further
474: details are given with \fBpcre_exec()\fP below.
475: .sp
476: PCRE_CONFIG_MATCH_LIMIT_RECURSION
477: .sp
478: The output is a long integer that gives the default limit for the depth of
479: recursion when calling the internal matching function in a \fBpcre_exec()\fP
480: execution. Further details are given with \fBpcre_exec()\fP below.
481: .sp
482: PCRE_CONFIG_STACKRECURSE
483: .sp
484: The output is an integer that is set to one if internal recursion when running
485: \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
486: to remember their state. This is the usual way that PCRE is compiled. The
487: output is zero if PCRE was compiled to use blocks of data on the heap instead
488: of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
489: \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
490: avoiding the use of the stack.
491: .
492: .
493: .SH "COMPILING A PATTERN"
494: .rs
495: .sp
496: .nf
497: .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
498: .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
499: .B " const unsigned char *\fItableptr\fP);"
500: .sp
501: .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
502: .B " int *\fIerrorcodeptr\fP,"
503: .B " const char **\fIerrptr\fP, int *\fIerroffset\fP,"
504: .B " const unsigned char *\fItableptr\fP);"
505: .fi
506: .P
507: Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
508: called to compile a pattern into an internal form. The only difference between
509: the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
510: \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
511: too much repetition, we refer just to \fBpcre_compile()\fP below, but the
512: information applies equally to \fBpcre_compile2()\fP.
513: .P
514: The pattern is a C string terminated by a binary zero, and is passed in the
515: \fIpattern\fP argument. A pointer to a single block of memory that is obtained
516: via \fBpcre_malloc\fP is returned. This contains the compiled code and related
517: data. The \fBpcre\fP type is defined for the returned block; this is a typedef
518: for a structure whose contents are not externally defined. It is up to the
519: caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
520: .P
521: Although the compiled code of a PCRE regex is relocatable, that is, it does not
522: depend on memory location, the complete \fBpcre\fP data block is not
523: fully relocatable, because it may contain a copy of the \fItableptr\fP
524: argument, which is an address (see below).
525: .P
526: The \fIoptions\fP argument contains various bit settings that affect the
527: compilation. It should be zero if no options are required. The available
528: options are described below. Some of them (in particular, those that are
529: compatible with Perl, but some others as well) can also be set and unset from
530: within the pattern (see the detailed description in the
531: .\" HREF
532: \fBpcrepattern\fP
533: .\"
534: documentation). For those options that can be different in different parts of
535: the pattern, the contents of the \fIoptions\fP argument specifies their
536: settings at the start of compilation and execution. The PCRE_ANCHORED,
537: PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
538: PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
539: compile time.
540: .P
541: If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
542: Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
543: NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
544: error message. This is a static string that is part of the library. You must
545: not try to free it. Normally, the offset from the start of the pattern to the
546: data unit that was being processed when the error was discovered is placed in
547: the variable pointed to by \fIerroffset\fP, which must not be NULL (if it is,
548: an immediate error is given). However, for an invalid UTF-8 or UTF-16 string,
549: the offset is that of the first data unit of the failing character.
550: .P
551: Some errors are not detected until the whole pattern has been scanned; in these
552: cases, the offset passed back is the length of the pattern. Note that the
553: offset is in data units, not characters, even in a UTF mode. It may sometimes
554: point into the middle of a UTF-8 or UTF-16 character.
555: .P
556: If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
557: \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
558: returned via this argument in the event of an error. This is in addition to the
559: textual error message. Error codes and messages are listed below.
560: .P
561: If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
562: character tables that are built when PCRE is compiled, using the default C
563: locale. Otherwise, \fItableptr\fP must be an address that is the result of a
564: call to \fBpcre_maketables()\fP. This value is stored with the compiled
565: pattern, and used again by \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP when the
566: pattern is matched. For more discussion, see the section on locale support
567: below.
568: .P
569: This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
570: .sp
571: pcre *re;
572: const char *error;
573: int erroffset;
574: re = pcre_compile(
575: "^A.*Z", /* the pattern */
576: 0, /* default options */
577: &error, /* for error message */
578: &erroffset, /* for error offset */
579: NULL); /* use default character tables */
580: .sp
581: The following names for option bits are defined in the \fBpcre.h\fP header
582: file:
583: .sp
584: PCRE_ANCHORED
585: .sp
586: If this bit is set, the pattern is forced to be "anchored", that is, it is
587: constrained to match only at the first matching point in the string that is
588: being searched (the "subject string"). This effect can also be achieved by
589: appropriate constructs in the pattern itself, which is the only way to do it in
590: Perl.
591: .sp
592: PCRE_AUTO_CALLOUT
593: .sp
594: If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
595: all with number 255, before each pattern item. For discussion of the callout
596: facility, see the
597: .\" HREF
598: \fBpcrecallout\fP
599: .\"
600: documentation.
601: .sp
602: PCRE_BSR_ANYCRLF
603: PCRE_BSR_UNICODE
604: .sp
605: These options (which are mutually exclusive) control what the \eR escape
606: sequence matches. The choice is either to match only CR, LF, or CRLF, or to
607: match any Unicode newline sequence. The default is specified when PCRE is
608: built. It can be overridden from within the pattern, or by setting an option
609: when a compiled pattern is matched.
610: .sp
611: PCRE_CASELESS
612: .sp
613: If this bit is set, letters in the pattern match both upper and lower case
614: letters. It is equivalent to Perl's /i option, and it can be changed within a
615: pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
616: concept of case for characters whose values are less than 128, so caseless
617: matching is always possible. For characters with higher values, the concept of
618: case is supported if PCRE is compiled with Unicode property support, but not
619: otherwise. If you want to use caseless matching for characters 128 and above,
620: you must ensure that PCRE is compiled with Unicode property support as well as
621: with UTF-8 support.
622: .sp
623: PCRE_DOLLAR_ENDONLY
624: .sp
625: If this bit is set, a dollar metacharacter in the pattern matches only at the
626: end of the subject string. Without this option, a dollar also matches
627: immediately before a newline at the end of the string (but not before any other
628: newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
629: There is no equivalent to this option in Perl, and no way to set it within a
630: pattern.
631: .sp
632: PCRE_DOTALL
633: .sp
634: If this bit is set, a dot metacharacter in the pattern matches a character of
635: any value, including one that indicates a newline. However, it only ever
636: matches one character, even if newlines are coded as CRLF. Without this option,
637: a dot does not match when the current position is at a newline. This option is
638: equivalent to Perl's /s option, and it can be changed within a pattern by a
639: (?s) option setting. A negative class such as [^a] always matches newline
640: characters, independent of the setting of this option.
641: .sp
642: PCRE_DUPNAMES
643: .sp
644: If this bit is set, names used to identify capturing subpatterns need not be
645: unique. This can be helpful for certain types of pattern when it is known that
646: only one instance of the named subpattern can ever be matched. There are more
647: details of named subpatterns below; see also the
648: .\" HREF
649: \fBpcrepattern\fP
650: .\"
651: documentation.
652: .sp
653: PCRE_EXTENDED
654: .sp
655: If this bit is set, most white space characters in the pattern are totally
656: ignored except when escaped or inside a character class. However, white space
657: is not allowed within sequences such as (?> that introduce various
658: parenthesized subpatterns, nor within a numerical quantifier such as {1,3}.
659: However, ignorable white space is permitted between an item and a following
660: quantifier and between a quantifier and a following + that indicates
661: possessiveness.
662: .P
663: White space did not used to include the VT character (code 11), because Perl
664: did not treat this character as white space. However, Perl changed at release
665: 5.18, so PCRE followed at release 8.34, and VT is now treated as white space.
666: .P
667: PCRE_EXTENDED also causes characters between an unescaped # outside a character
668: class and the next newline, inclusive, to be ignored. PCRE_EXTENDED is
669: equivalent to Perl's /x option, and it can be changed within a pattern by a
670: (?x) option setting.
671: .P
672: Which characters are interpreted as newlines is controlled by the options
673: passed to \fBpcre_compile()\fP or by a special sequence at the start of the
674: pattern, as described in the section entitled
675: .\" HTML <a href="pcrepattern.html#newlines">
676: .\" </a>
677: "Newline conventions"
678: .\"
679: in the \fBpcrepattern\fP documentation. Note that the end of this type of
680: comment is a literal newline sequence in the pattern; escape sequences that
681: happen to represent a newline do not count.
682: .P
683: This option makes it possible to include comments inside complicated patterns.
684: Note, however, that this applies only to data characters. White space characters
685: may never appear within special character sequences in a pattern, for example
686: within the sequence (?( that introduces a conditional subpattern.
687: .sp
688: PCRE_EXTRA
689: .sp
690: This option was invented in order to turn on additional functionality of PCRE
691: that is incompatible with Perl, but it is currently of very little use. When
692: set, any backslash in a pattern that is followed by a letter that has no
693: special meaning causes an error, thus reserving these combinations for future
694: expansion. By default, as in Perl, a backslash followed by a letter with no
695: special meaning is treated as a literal. (Perl can, however, be persuaded to
696: give an error for this, by running it with the -w option.) There are at present
697: no other features controlled by this option. It can also be set by a (?X)
698: option setting within a pattern.
699: .sp
700: PCRE_FIRSTLINE
701: .sp
702: If this option is set, an unanchored pattern is required to match before or at
703: the first newline in the subject string, though the matched text may continue
704: over the newline.
705: .sp
706: PCRE_JAVASCRIPT_COMPAT
707: .sp
708: If this option is set, PCRE's behaviour is changed in some ways so that it is
709: compatible with JavaScript rather than Perl. The changes are as follows:
710: .P
711: (1) A lone closing square bracket in a pattern causes a compile-time error,
712: because this is illegal in JavaScript (by default it is treated as a data
713: character). Thus, the pattern AB]CD becomes illegal when this option is set.
714: .P
715: (2) At run time, a back reference to an unset subpattern group matches an empty
716: string (by default this causes the current matching alternative to fail). A
717: pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
718: an "a" in the subject), whereas it fails by default, for Perl compatibility.
719: .P
720: (3) \eU matches an upper case "U" character; by default \eU causes a compile
721: time error (Perl uses \eU to upper case subsequent characters).
722: .P
723: (4) \eu matches a lower case "u" character unless it is followed by four
724: hexadecimal digits, in which case the hexadecimal number defines the code point
725: to match. By default, \eu causes a compile time error (Perl uses it to upper
726: case the following character).
727: .P
728: (5) \ex matches a lower case "x" character unless it is followed by two
729: hexadecimal digits, in which case the hexadecimal number defines the code point
730: to match. By default, as in Perl, a hexadecimal number is always expected after
731: \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
732: binary zero character followed by z).
733: .sp
734: PCRE_MULTILINE
735: .sp
736: By default, for the purposes of matching "start of line" and "end of line",
737: PCRE treats the subject string as consisting of a single line of characters,
738: even if it actually contains newlines. The "start of line" metacharacter (^)
739: matches only at the start of the string, and the "end of line" metacharacter
740: ($) matches only at the end of the string, or before a terminating newline
741: (except when PCRE_DOLLAR_ENDONLY is set). Note, however, that unless
742: PCRE_DOTALL is set, the "any character" metacharacter (.) does not match at a
743: newline. This behaviour (for ^, $, and dot) is the same as Perl.
744: .P
745: When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
746: match immediately following or immediately before internal newlines in the
747: subject string, respectively, as well as at the very start and end. This is
748: equivalent to Perl's /m option, and it can be changed within a pattern by a
749: (?m) option setting. If there are no newlines in a subject string, or no
750: occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
751: .sp
752: PCRE_NEVER_UTF
753: .sp
754: This option locks out interpretation of the pattern as UTF-8 (or UTF-16 or
755: UTF-32 in the 16-bit and 32-bit libraries). In particular, it prevents the
756: creator of the pattern from switching to UTF interpretation by starting the
757: pattern with (*UTF). This may be useful in applications that process patterns
758: from external sources. The combination of PCRE_UTF8 and PCRE_NEVER_UTF also
759: causes an error.
760: .sp
761: PCRE_NEWLINE_CR
762: PCRE_NEWLINE_LF
763: PCRE_NEWLINE_CRLF
764: PCRE_NEWLINE_ANYCRLF
765: PCRE_NEWLINE_ANY
766: .sp
767: These options override the default newline definition that was chosen when PCRE
768: was built. Setting the first or the second specifies that a newline is
769: indicated by a single character (CR or LF, respectively). Setting
770: PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
771: CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
772: preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
773: that any Unicode newline sequence should be recognized.
774: .P
775: In an ASCII/Unicode environment, the Unicode newline sequences are the three
776: just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
777: feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
778: (paragraph separator, U+2029). For the 8-bit library, the last two are
779: recognized only in UTF-8 mode.
780: .P
781: When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
782: CR is 0x0d, the same as ASCII. However, the character code for LF is normally
783: 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
784: not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
785: less than 256. For more details, see the
786: .\" HREF
787: \fBpcrebuild\fP
788: .\"
789: documentation.
790: .P
791: The newline setting in the options word uses three bits that are treated
792: as a number, giving eight possibilities. Currently only six are used (default
793: plus the five values above). This means that if you set more than one newline
794: option, the combination may or may not be sensible. For example,
795: PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
796: other combinations may yield unused numbers and cause an error.
797: .P
798: The only time that a line break in a pattern is specially recognized when
799: compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
800: and so are ignored in this mode. Also, an unescaped # outside a character class
801: indicates a comment that lasts until after the next line break sequence. In
802: other circumstances, line break sequences in patterns are treated as literal
803: data.
804: .P
805: The newline option that is set at compile time becomes the default that is used
806: for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
807: .sp
808: PCRE_NO_AUTO_CAPTURE
809: .sp
810: If this option is set, it disables the use of numbered capturing parentheses in
811: the pattern. Any opening parenthesis that is not followed by ? behaves as if it
812: were followed by ?: but named parentheses can still be used for capturing (and
813: they acquire numbers in the usual way). There is no equivalent of this option
814: in Perl.
815: .sp
816: PCRE_NO_AUTO_POSSESS
817: .sp
818: If this option is set, it disables "auto-possessification". This is an
819: optimization that, for example, turns a+b into a++b in order to avoid
820: backtracks into a+ that can never be successful. However, if callouts are in
821: use, auto-possessification means that some of them are never taken. You can set
822: this option if you want the matching functions to do a full unoptimized search
823: and run all the callouts, but it is mainly provided for testing purposes.
824: .sp
825: PCRE_NO_START_OPTIMIZE
826: .sp
827: This is an option that acts at matching time; that is, it is really an option
828: for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
829: it is remembered with the compiled pattern and assumed at matching time. This
830: is necessary if you want to use JIT execution, because the JIT compiler needs
831: to know whether or not this option is set. For details see the discussion of
832: PCRE_NO_START_OPTIMIZE
833: .\" HTML <a href="#execoptions">
834: .\" </a>
835: below.
836: .\"
837: .sp
838: PCRE_UCP
839: .sp
840: This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
841: \ew, and some of the POSIX character classes. By default, only ASCII characters
842: are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
843: classify characters. More details are given in the section on
844: .\" HTML <a href="pcre.html#genericchartypes">
845: .\" </a>
846: generic character types
847: .\"
848: in the
849: .\" HREF
850: \fBpcrepattern\fP
851: .\"
852: page. If you set PCRE_UCP, matching one of the items it affects takes much
853: longer. The option is available only if PCRE has been compiled with Unicode
854: property support.
855: .sp
856: PCRE_UNGREEDY
857: .sp
858: This option inverts the "greediness" of the quantifiers so that they are not
859: greedy by default, but become greedy if followed by "?". It is not compatible
860: with Perl. It can also be set by a (?U) option setting within the pattern.
861: .sp
862: PCRE_UTF8
863: .sp
864: This option causes PCRE to regard both the pattern and the subject as strings
865: of UTF-8 characters instead of single-byte strings. However, it is available
866: only when PCRE is built to include UTF support. If not, the use of this option
867: provokes an error. Details of how this option changes the behaviour of PCRE are
868: given in the
869: .\" HREF
870: \fBpcreunicode\fP
871: .\"
872: page.
873: .sp
874: PCRE_NO_UTF8_CHECK
875: .sp
876: When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
877: automatically checked. There is a discussion about the
878: .\" HTML <a href="pcreunicode.html#utf8strings">
879: .\" </a>
880: validity of UTF-8 strings
881: .\"
882: in the
883: .\" HREF
884: \fBpcreunicode\fP
885: .\"
886: page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
887: error. If you already know that your pattern is valid, and you want to skip
888: this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
889: When it is set, the effect of passing an invalid UTF-8 string as a pattern is
890: undefined. It may cause your program to crash or loop. Note that this option
891: can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
892: the validity checking of subject strings only. If the same string is being
893: matched many times, the option can be safely set for the second and subsequent
894: matchings to improve performance.
895: .
896: .
897: .SH "COMPILATION ERROR CODES"
898: .rs
899: .sp
900: The following table lists the error codes than may be returned by
901: \fBpcre_compile2()\fP, along with the error messages that may be returned by
902: both compiling functions. Note that error messages are always 8-bit ASCII
903: strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
904: have fallen out of use. To avoid confusion, they have not been re-used.
905: .sp
906: 0 no error
907: 1 \e at end of pattern
908: 2 \ec at end of pattern
909: 3 unrecognized character follows \e
910: 4 numbers out of order in {} quantifier
911: 5 number too big in {} quantifier
912: 6 missing terminating ] for character class
913: 7 invalid escape sequence in character class
914: 8 range out of order in character class
915: 9 nothing to repeat
916: 10 [this code is not in use]
917: 11 internal error: unexpected repeat
918: 12 unrecognized character after (? or (?-
919: 13 POSIX named classes are supported only within a class
920: 14 missing )
921: 15 reference to non-existent subpattern
922: 16 erroffset passed as NULL
923: 17 unknown option bit(s) set
924: 18 missing ) after comment
925: 19 [this code is not in use]
926: 20 regular expression is too large
927: 21 failed to get memory
928: 22 unmatched parentheses
929: 23 internal error: code overflow
930: 24 unrecognized character after (?<
931: 25 lookbehind assertion is not fixed length
932: 26 malformed number or name after (?(
933: 27 conditional group contains more than two branches
934: 28 assertion expected after (?(
935: 29 (?R or (?[+-]digits must be followed by )
936: 30 unknown POSIX class name
937: 31 POSIX collating elements are not supported
938: 32 this version of PCRE is compiled without UTF support
939: 33 [this code is not in use]
940: 34 character value in \ex{} or \eo{} is too large
941: 35 invalid condition (?(0)
942: 36 \eC not allowed in lookbehind assertion
943: 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
944: 38 number after (?C is > 255
945: 39 closing ) for (?C expected
946: 40 recursive call could loop indefinitely
947: 41 unrecognized character after (?P
948: 42 syntax error in subpattern name (missing terminator)
949: 43 two named subpatterns have the same name
950: 44 invalid UTF-8 string (specifically UTF-8)
951: 45 support for \eP, \ep, and \eX has not been compiled
952: 46 malformed \eP or \ep sequence
953: 47 unknown property name after \eP or \ep
954: 48 subpattern name is too long (maximum 32 characters)
955: 49 too many named subpatterns (maximum 10000)
956: 50 [this code is not in use]
957: 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
958: 52 internal error: overran compiling workspace
959: 53 internal error: previously-checked referenced subpattern
960: not found
961: 54 DEFINE group contains more than one branch
962: 55 repeating a DEFINE group is not allowed
963: 56 inconsistent NEWLINE options
964: 57 \eg is not followed by a braced, angle-bracketed, or quoted
965: name/number or by a plain number
966: 58 a numbered reference must not be zero
967: 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
968: 60 (*VERB) not recognized or malformed
969: 61 number is too big
970: 62 subpattern name expected
971: 63 digit expected after (?+
972: 64 ] is an invalid data character in JavaScript compatibility mode
973: 65 different names for subpatterns of the same number are
974: not allowed
975: 66 (*MARK) must have an argument
976: 67 this version of PCRE is not compiled with Unicode property
977: support
978: 68 \ec must be followed by an ASCII character
979: 69 \ek is not followed by a braced, angle-bracketed, or quoted name
980: 70 internal error: unknown opcode in find_fixedlength()
981: 71 \eN is not supported in a class
982: 72 too many forward references
983: 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
984: 74 invalid UTF-16 string (specifically UTF-16)
985: 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
986: 76 character value in \eu.... sequence is too large
987: 77 invalid UTF-32 string (specifically UTF-32)
988: 78 setting UTF is disabled by the application
989: 79 non-hex character in \ex{} (closing brace missing?)
990: 80 non-octal character in \eo{} (closing brace missing?)
991: 81 missing opening brace after \eo
992: 82 parentheses are too deeply nested
993: 83 invalid range in character class
994: .sp
995: The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
996: be used if the limits were changed when PCRE was built.
997: .
998: .
999: .\" HTML <a name="studyingapattern"></a>
1000: .SH "STUDYING A PATTERN"
1001: .rs
1002: .sp
1003: .nf
1004: .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
1005: .B " const char **\fIerrptr\fP);"
1006: .fi
1007: .PP
1008: If a compiled pattern is going to be used several times, it is worth spending
1009: more time analyzing it in order to speed up the time taken for matching. The
1010: function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
1011: argument. If studying the pattern produces additional information that will
1012: help speed up matching, \fBpcre_study()\fP returns a pointer to a
1013: \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
1014: results of the study.
1015: .P
1016: The returned value from \fBpcre_study()\fP can be passed directly to
1017: \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
1018: also contains other fields that can be set by the caller before the block is
1019: passed; these are described
1020: .\" HTML <a href="#extradata">
1021: .\" </a>
1022: below
1023: .\"
1024: in the section on matching a pattern.
1025: .P
1026: If studying the pattern does not produce any useful information,
1027: \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1028: calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1029: \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1030: if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1031: returns a \fBpcre_extra\fP block even if studying did not find any additional
1032: information. It may still return NULL, however, if an error occurs in
1033: \fBpcre_study()\fP.
1034: .P
1035: The second argument of \fBpcre_study()\fP contains option bits. There are three
1036: further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1037: .sp
1038: PCRE_STUDY_JIT_COMPILE
1039: PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1040: PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1041: .sp
1042: If any of these are set, and the just-in-time compiler is available, the
1043: pattern is further compiled into machine code that executes much faster than
1044: the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1045: compiler is not available, these options are ignored. All undefined bits in the
1046: \fIoptions\fP argument must be zero.
1047: .P
1048: JIT compilation is a heavyweight optimization. It can take some time for
1049: patterns to be analyzed, and for one-off matches and simple patterns the
1050: benefit of faster execution might be offset by a much slower study time.
1051: Not all patterns can be optimized by the JIT compiler. For those that cannot be
1052: handled, matching automatically falls back to the \fBpcre_exec()\fP
1053: interpreter. For more details, see the
1054: .\" HREF
1055: \fBpcrejit\fP
1056: .\"
1057: documentation.
1058: .P
1059: The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1060: studying succeeds (even if no data is returned), the variable it points to is
1061: set to NULL. Otherwise it is set to point to a textual error message. This is a
1062: static string that is part of the library. You must not try to free it. You
1063: should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1064: sure that it has run successfully.
1065: .P
1066: When you are finished with a pattern, you can free the memory used for the
1067: study data by calling \fBpcre_free_study()\fP. This function was added to the
1068: API for release 8.20. For earlier versions, the memory could be freed with
1069: \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1070: where JIT optimization is not used, but it is advisable to change to the new
1071: function when convenient.
1072: .P
1073: This is a typical way in which \fBpcre_study\fP() is used (except that in a
1074: real application there should be tests for errors):
1075: .sp
1076: int rc;
1077: pcre *re;
1078: pcre_extra *sd;
1079: re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1080: sd = pcre_study(
1081: re, /* result of pcre_compile() */
1082: 0, /* no options */
1083: &error); /* set to NULL or points to a message */
1084: rc = pcre_exec( /* see below for details of pcre_exec() options */
1085: re, sd, "subject", 7, 0, 0, ovector, 30);
1086: ...
1087: pcre_free_study(sd);
1088: pcre_free(re);
1089: .sp
1090: Studying a pattern does two things: first, a lower bound for the length of
1091: subject string that is needed to match the pattern is computed. This does not
1092: mean that there are any strings of that length that match, but it does
1093: guarantee that no shorter strings match. The value is used to avoid wasting
1094: time by trying to match strings that are shorter than the lower bound. You can
1095: find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1096: .P
1097: Studying a pattern is also useful for non-anchored patterns that do not have a
1098: single fixed starting character. A bitmap of possible starting bytes is
1099: created. This speeds up finding a position in the subject at which to start
1100: matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1101: In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1102: .P
1103: These two optimizations apply to both \fBpcre_exec()\fP and
1104: \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1105: The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1106: You might want to do this if your pattern contains callouts or (*MARK) and you
1107: want to make use of these facilities in cases where matching fails.
1108: .P
1109: PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1110: time. However, if PCRE_NO_START_OPTIMIZE is passed to \fBpcre_exec()\fP, (that
1111: is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1112: execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1113: compile time.
1114: .P
1115: There is a longer discussion of PCRE_NO_START_OPTIMIZE
1116: .\" HTML <a href="#execoptions">
1117: .\" </a>
1118: below.
1119: .\"
1120: .
1121: .
1122: .\" HTML <a name="localesupport"></a>
1123: .SH "LOCALE SUPPORT"
1124: .rs
1125: .sp
1126: PCRE handles caseless matching, and determines whether characters are letters,
1127: digits, or whatever, by reference to a set of tables, indexed by character
1128: code point. When running in UTF-8 mode, or in the 16- or 32-bit libraries, this
1129: applies only to characters with code points less than 256. By default,
1130: higher-valued code points never match escapes such as \ew or \ed. However, if
1131: PCRE is built with Unicode property support, all characters can be tested with
1132: \ep and \eP, or, alternatively, the PCRE_UCP option can be set when a pattern
1133: is compiled; this causes \ew and friends to use Unicode property support
1134: instead of the built-in tables.
1135: .P
1136: The use of locales with Unicode is discouraged. If you are handling characters
1137: with code points greater than 128, you should either use Unicode support, or
1138: use locales, but not try to mix the two.
1139: .P
1140: PCRE contains an internal set of tables that are used when the final argument
1141: of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1142: Normally, the internal tables recognize only ASCII characters. However, when
1143: PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1144: default "C" locale of the local system, which may cause them to be different.
1145: .P
1146: The internal tables can always be overridden by tables supplied by the
1147: application that calls PCRE. These may be created in a different locale from
1148: the default. As more and more applications change to using Unicode, the need
1149: for this locale support is expected to die away.
1150: .P
1151: External tables are built by calling the \fBpcre_maketables()\fP function,
1152: which has no arguments, in the relevant locale. The result can then be passed
1153: to \fBpcre_compile()\fP as often as necessary. For example, to build and use
1154: tables that are appropriate for the French locale (where accented characters
1155: with values greater than 128 are treated as letters), the following code could
1156: be used:
1157: .sp
1158: setlocale(LC_CTYPE, "fr_FR");
1159: tables = pcre_maketables();
1160: re = pcre_compile(..., tables);
1161: .sp
1162: The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1163: are using Windows, the name for the French locale is "french".
1164: .P
1165: When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1166: obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1167: that the memory containing the tables remains available for as long as it is
1168: needed.
1169: .P
1170: The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1171: pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1172: and also by \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP. Thus, for any single
1173: pattern, compilation, studying and matching all happen in the same locale, but
1174: different patterns can be processed in different locales.
1175: .P
1176: It is possible to pass a table pointer or NULL (indicating the use of the
1177: internal tables) to \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP (see the
1178: discussion below in the section on matching a pattern). This facility is
1179: provided for use with pre-compiled patterns that have been saved and reloaded.
1180: Character tables are not saved with patterns, so if a non-standard table was
1181: used at compile time, it must be provided again when the reloaded pattern is
1182: matched. Attempting to use this facility to match a pattern in a different
1183: locale from the one in which it was compiled is likely to lead to anomalous
1184: (usually incorrect) results.
1185: .
1186: .
1187: .\" HTML <a name="infoaboutpattern"></a>
1188: .SH "INFORMATION ABOUT A PATTERN"
1189: .rs
1190: .sp
1191: .nf
1192: .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1193: .B " int \fIwhat\fP, void *\fIwhere\fP);"
1194: .fi
1195: .PP
1196: The \fBpcre_fullinfo()\fP function returns information about a compiled
1197: pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1198: library at version 8.30, after more than 10 years of obsolescence.
1199: .P
1200: The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1201: pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1202: the pattern was not studied. The third argument specifies which piece of
1203: information is required, and the fourth argument is a pointer to a variable
1204: to receive the data. The yield of the function is zero for success, or one of
1205: the following negative numbers:
1206: .sp
1207: PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1208: the argument \fIwhere\fP was NULL
1209: PCRE_ERROR_BADMAGIC the "magic number" was not found
1210: PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1211: endianness
1212: PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1213: PCRE_ERROR_UNSET the requested field is not set
1214: .sp
1215: The "magic number" is placed at the start of each compiled pattern as an simple
1216: check against passing an arbitrary memory pointer. The endianness error can
1217: occur if a compiled pattern is saved and reloaded on a different host. Here is
1218: a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1219: pattern:
1220: .sp
1221: int rc;
1222: size_t length;
1223: rc = pcre_fullinfo(
1224: re, /* result of pcre_compile() */
1225: sd, /* result of pcre_study(), or NULL */
1226: PCRE_INFO_SIZE, /* what is required */
1227: &length); /* where to put the data */
1228: .sp
1229: The possible values for the third argument are defined in \fBpcre.h\fP, and are
1230: as follows:
1231: .sp
1232: PCRE_INFO_BACKREFMAX
1233: .sp
1234: Return the number of the highest back reference in the pattern. The fourth
1235: argument should point to an \fBint\fP variable. Zero is returned if there are
1236: no back references.
1237: .sp
1238: PCRE_INFO_CAPTURECOUNT
1239: .sp
1240: Return the number of capturing subpatterns in the pattern. The fourth argument
1241: should point to an \fBint\fP variable.
1242: .sp
1243: PCRE_INFO_DEFAULT_TABLES
1244: .sp
1245: Return a pointer to the internal default character tables within PCRE. The
1246: fourth argument should point to an \fBunsigned char *\fP variable. This
1247: information call is provided for internal use by the \fBpcre_study()\fP
1248: function. External callers can cause PCRE to use its internal tables by passing
1249: a NULL table pointer.
1250: .sp
1251: PCRE_INFO_FIRSTBYTE
1252: .sp
1253: Return information about the first data unit of any matched string, for a
1254: non-anchored pattern. (The name of this option refers to the 8-bit library,
1255: where data units are bytes.) The fourth argument should point to an \fBint\fP
1256: variable.
1257: .P
1258: If there is a fixed first value, for example, the letter "c" from a pattern
1259: such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1260: value is always less than 256. In the 16-bit library the value can be up to
1261: 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1262: .P
1263: If there is no fixed first value, and if either
1264: .sp
1265: (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1266: starts with "^", or
1267: .sp
1268: (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1269: (if it were set, the pattern would be anchored),
1270: .sp
1271: -1 is returned, indicating that the pattern matches only at the start of a
1272: subject string or after any newline within the string. Otherwise -2 is
1273: returned. For anchored patterns, -2 is returned.
1274: .P
1275: Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1276: to return the full 32-bit range of the character, this value is deprecated;
1277: instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1278: should be used.
1279: .sp
1280: PCRE_INFO_FIRSTTABLE
1281: .sp
1282: If the pattern was studied, and this resulted in the construction of a 256-bit
1283: table indicating a fixed set of values for the first data unit in any matching
1284: string, a pointer to the table is returned. Otherwise NULL is returned. The
1285: fourth argument should point to an \fBunsigned char *\fP variable.
1286: .sp
1287: PCRE_INFO_HASCRORLF
1288: .sp
1289: Return 1 if the pattern contains any explicit matches for CR or LF characters,
1290: otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1291: explicit match is either a literal CR or LF character, or \er or \en.
1292: .sp
1293: PCRE_INFO_JCHANGED
1294: .sp
1295: Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1296: 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1297: (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1298: .sp
1299: PCRE_INFO_JIT
1300: .sp
1301: Return 1 if the pattern was studied with one of the JIT options, and
1302: just-in-time compiling was successful. The fourth argument should point to an
1303: \fBint\fP variable. A return value of 0 means that JIT support is not available
1304: in this version of PCRE, or that the pattern was not studied with a JIT option,
1305: or that the JIT compiler could not handle this particular pattern. See the
1306: .\" HREF
1307: \fBpcrejit\fP
1308: .\"
1309: documentation for details of what can and cannot be handled.
1310: .sp
1311: PCRE_INFO_JITSIZE
1312: .sp
1313: If the pattern was successfully studied with a JIT option, return the size of
1314: the JIT compiled code, otherwise return zero. The fourth argument should point
1315: to a \fBsize_t\fP variable.
1316: .sp
1317: PCRE_INFO_LASTLITERAL
1318: .sp
1319: Return the value of the rightmost literal data unit that must exist in any
1320: matched string, other than at its start, if such a value has been recorded. The
1321: fourth argument should point to an \fBint\fP variable. If there is no such
1322: value, -1 is returned. For anchored patterns, a last literal value is recorded
1323: only if it follows something of variable length. For example, for the pattern
1324: /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1325: is -1.
1326: .P
1327: Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1328: to return the full 32-bit range of characters, this value is deprecated;
1329: instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1330: be used.
1331: .sp
1332: PCRE_INFO_MATCH_EMPTY
1333: .sp
1334: Return 1 if the pattern can match an empty string, otherwise 0. The fourth
1335: argument should point to an \fBint\fP variable.
1336: .sp
1337: PCRE_INFO_MATCHLIMIT
1338: .sp
1339: If the pattern set a match limit by including an item of the form
1340: (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1341: should point to an unsigned 32-bit integer. If no such value has been set, the
1342: call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1343: .sp
1344: PCRE_INFO_MAXLOOKBEHIND
1345: .sp
1346: Return the number of characters (NB not data units) in the longest lookbehind
1347: assertion in the pattern. This information is useful when doing multi-segment
1348: matching using the partial matching facilities. Note that the simple assertions
1349: \eb and \eB require a one-character lookbehind. \eA also registers a
1350: one-character lookbehind, though it does not actually inspect the previous
1351: character. This is to ensure that at least one character from the old segment
1352: is retained when a new segment is processed. Otherwise, if there are no
1353: lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1354: segment.
1355: .sp
1356: PCRE_INFO_MINLENGTH
1357: .sp
1358: If the pattern was studied and a minimum length for matching subject strings
1359: was computed, its value is returned. Otherwise the returned value is -1. The
1360: value is a number of characters, which in UTF mode may be different from the
1361: number of data units. The fourth argument should point to an \fBint\fP
1362: variable. A non-negative value is a lower bound to the length of any matching
1363: string. There may not be any strings of that length that do actually match, but
1364: every string that does match is at least that long.
1365: .sp
1366: PCRE_INFO_NAMECOUNT
1367: PCRE_INFO_NAMEENTRYSIZE
1368: PCRE_INFO_NAMETABLE
1369: .sp
1370: PCRE supports the use of named as well as numbered capturing parentheses. The
1371: names are just an additional way of identifying the parentheses, which still
1372: acquire numbers. Several convenience functions such as
1373: \fBpcre_get_named_substring()\fP are provided for extracting captured
1374: substrings by name. It is also possible to extract the data directly, by first
1375: converting the name to a number in order to access the correct pointers in the
1376: output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1377: you need to use the name-to-number map, which is described by these three
1378: values.
1379: .P
1380: The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1381: the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1382: entry; both of these return an \fBint\fP value. The entry size depends on the
1383: length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1384: entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1385: the first two bytes of each entry are the number of the capturing parenthesis,
1386: most significant byte first. In the 16-bit library, the pointer points to
1387: 16-bit data units, the first of which contains the parenthesis number. In the
1388: 32-bit library, the pointer points to 32-bit data units, the first of which
1389: contains the parenthesis number. The rest of the entry is the corresponding
1390: name, zero terminated.
1391: .P
1392: The names are in alphabetical order. If (?| is used to create multiple groups
1393: with the same number, as described in the
1394: .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1395: .\" </a>
1396: section on duplicate subpattern numbers
1397: .\"
1398: in the
1399: .\" HREF
1400: \fBpcrepattern\fP
1401: .\"
1402: page, the groups may be given the same name, but there is only one entry in the
1403: table. Different names for groups of the same number are not permitted.
1404: Duplicate names for subpatterns with different numbers are permitted,
1405: but only if PCRE_DUPNAMES is set. They appear in the table in the order in
1406: which they were found in the pattern. In the absence of (?| this is the order
1407: of increasing number; when (?| is used this is not necessarily the case because
1408: later subpatterns may have lower numbers.
1409: .P
1410: As a simple example of the name/number table, consider the following pattern
1411: after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1412: space - including newlines - is ignored):
1413: .sp
1414: .\" JOIN
1415: (?<date> (?<year>(\ed\ed)?\ed\ed) -
1416: (?<month>\ed\ed) - (?<day>\ed\ed) )
1417: .sp
1418: There are four named subpatterns, so the table has four entries, and each entry
1419: in the table is eight bytes long. The table is as follows, with non-printing
1420: bytes shows in hexadecimal, and undefined bytes shown as ??:
1421: .sp
1422: 00 01 d a t e 00 ??
1423: 00 05 d a y 00 ?? ??
1424: 00 04 m o n t h 00
1425: 00 02 y e a r 00 ??
1426: .sp
1427: When writing code to extract data from named subpatterns using the
1428: name-to-number map, remember that the length of the entries is likely to be
1429: different for each compiled pattern.
1430: .sp
1431: PCRE_INFO_OKPARTIAL
1432: .sp
1433: Return 1 if the pattern can be used for partial matching with
1434: \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1435: \fBint\fP variable. From release 8.00, this always returns 1, because the
1436: restrictions that previously applied to partial matching have been lifted. The
1437: .\" HREF
1438: \fBpcrepartial\fP
1439: .\"
1440: documentation gives details of partial matching.
1441: .sp
1442: PCRE_INFO_OPTIONS
1443: .sp
1444: Return a copy of the options with which the pattern was compiled. The fourth
1445: argument should point to an \fBunsigned long int\fP variable. These option bits
1446: are those specified in the call to \fBpcre_compile()\fP, modified by any
1447: top-level option settings at the start of the pattern itself. In other words,
1448: they are the options that will be in force when matching starts. For example,
1449: if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1450: result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1451: .P
1452: A pattern is automatically anchored by PCRE if all of its top-level
1453: alternatives begin with one of the following:
1454: .sp
1455: ^ unless PCRE_MULTILINE is set
1456: \eA always
1457: \eG always
1458: .\" JOIN
1459: .* if PCRE_DOTALL is set and there are no back
1460: references to the subpattern in which .* appears
1461: .sp
1462: For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1463: \fBpcre_fullinfo()\fP.
1464: .sp
1465: PCRE_INFO_RECURSIONLIMIT
1466: .sp
1467: If the pattern set a recursion limit by including an item of the form
1468: (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1469: argument should point to an unsigned 32-bit integer. If no such value has been
1470: set, the call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1471: .sp
1472: PCRE_INFO_SIZE
1473: .sp
1474: Return the size of the compiled pattern in bytes (for all three libraries). The
1475: fourth argument should point to a \fBsize_t\fP variable. This value does not
1476: include the size of the \fBpcre\fP structure that is returned by
1477: \fBpcre_compile()\fP. The value that is passed as the argument to
1478: \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1479: place the compiled data is the value returned by this option plus the size of
1480: the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1481: does not alter the value returned by this option.
1482: .sp
1483: PCRE_INFO_STUDYSIZE
1484: .sp
1485: Return the size in bytes (for all three libraries) of the data block pointed to
1486: by the \fIstudy_data\fP field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP
1487: is NULL, or there is no study data, zero is returned. The fourth argument
1488: should point to a \fBsize_t\fP variable. The \fIstudy_data\fP field is set by
1489: \fBpcre_study()\fP to record information that will speed up matching (see the
1490: section entitled
1491: .\" HTML <a href="#studyingapattern">
1492: .\" </a>
1493: "Studying a pattern"
1494: .\"
1495: above). The format of the \fIstudy_data\fP block is private, but its length
1496: is made available via this option so that it can be saved and restored (see the
1497: .\" HREF
1498: \fBpcreprecompile\fP
1499: .\"
1500: documentation for details).
1501: .sp
1502: PCRE_INFO_FIRSTCHARACTERFLAGS
1503: .sp
1504: Return information about the first data unit of any matched string, for a
1505: non-anchored pattern. The fourth argument should point to an \fBint\fP
1506: variable.
1507: .P
1508: If there is a fixed first value, for example, the letter "c" from a pattern
1509: such as (cat|cow|coyote), 1 is returned, and the character value can be
1510: retrieved using PCRE_INFO_FIRSTCHARACTER.
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: 2 is returned, indicating that the pattern matches only at the start of a
1521: subject string or after any newline within the string. Otherwise 0 is
1522: returned. For anchored patterns, 0 is returned.
1523: .sp
1524: PCRE_INFO_FIRSTCHARACTER
1525: .sp
1526: Return the fixed first character value in the situation where
1527: PCRE_INFO_FIRSTCHARACTERFLAGS returns 1; otherwise return 0. The fourth
1528: argument should point to an \fBuint_t\fP variable.
1529: .P
1530: In the 8-bit library, the value is always less than 256. In the 16-bit library
1531: the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1532: can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1533: .sp
1534: PCRE_INFO_REQUIREDCHARFLAGS
1535: .sp
1536: Returns 1 if there is a rightmost literal data unit that must exist in any
1537: matched string, other than at its start. The fourth argument should point to
1538: an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1539: 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1540: .P
1541: For anchored patterns, a last literal value is recorded only if it follows
1542: something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1543: returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1544: /^a\edz\ed/ the returned value is 0.
1545: .sp
1546: PCRE_INFO_REQUIREDCHAR
1547: .sp
1548: Return the value of the rightmost literal data unit that must exist in any
1549: matched string, other than at its start, if such a value has been recorded. The
1550: fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1551: value, 0 is returned.
1552: .
1553: .
1554: .SH "REFERENCE COUNTS"
1555: .rs
1556: .sp
1557: .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1558: .PP
1559: The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1560: data block that contains a compiled pattern. It is provided for the benefit of
1561: applications that operate in an object-oriented manner, where different parts
1562: of the application may be using the same compiled pattern, but you want to free
1563: the block when they are all done.
1564: .P
1565: When a pattern is compiled, the reference count field is initialized to zero.
1566: It is changed only by calling this function, whose action is to add the
1567: \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1568: function is the new value. However, the value of the count is constrained to
1569: lie between 0 and 65535, inclusive. If the new value is outside these limits,
1570: it is forced to the appropriate limit value.
1571: .P
1572: Except when it is zero, the reference count is not correctly preserved if a
1573: pattern is compiled on one host and then transferred to a host whose byte-order
1574: is different. (This seems a highly unlikely scenario.)
1575: .
1576: .
1577: .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1578: .rs
1579: .sp
1580: .nf
1581: .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1582: .B " const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1583: .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);"
1584: .fi
1585: .P
1586: The function \fBpcre_exec()\fP is called to match a subject string against a
1587: compiled pattern, which is passed in the \fIcode\fP argument. If the
1588: pattern was studied, the result of the study should be passed in the
1589: \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1590: and \fIextra\fP arguments as many times as you like, in order to match
1591: different subject strings with the same pattern.
1592: .P
1593: This function is the main matching facility of the library, and it operates in
1594: a Perl-like manner. For specialist use there is also an alternative matching
1595: function, which is described
1596: .\" HTML <a href="#dfamatch">
1597: .\" </a>
1598: below
1599: .\"
1600: in the section about the \fBpcre_dfa_exec()\fP function.
1601: .P
1602: In most applications, the pattern will have been compiled (and optionally
1603: studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1604: possible to save compiled patterns and study data, and then use them later
1605: in different processes, possibly even on different hosts. For a discussion
1606: about this, see the
1607: .\" HREF
1608: \fBpcreprecompile\fP
1609: .\"
1610: documentation.
1611: .P
1612: Here is an example of a simple call to \fBpcre_exec()\fP:
1613: .sp
1614: int rc;
1615: int ovector[30];
1616: rc = pcre_exec(
1617: re, /* result of pcre_compile() */
1618: NULL, /* we didn't study the pattern */
1619: "some string", /* the subject string */
1620: 11, /* the length of the subject string */
1621: 0, /* start at offset 0 in the subject */
1622: 0, /* default options */
1623: ovector, /* vector of integers for substring information */
1624: 30); /* number of elements (NOT size in bytes) */
1625: .
1626: .
1627: .\" HTML <a name="extradata"></a>
1628: .SS "Extra data for \fBpcre_exec()\fR"
1629: .rs
1630: .sp
1631: If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1632: data block. The \fBpcre_study()\fP function returns such a block (when it
1633: doesn't return NULL), but you can also create one for yourself, and pass
1634: additional information in it. The \fBpcre_extra\fP block contains the following
1635: fields (not necessarily in this order):
1636: .sp
1637: unsigned long int \fIflags\fP;
1638: void *\fIstudy_data\fP;
1639: void *\fIexecutable_jit\fP;
1640: unsigned long int \fImatch_limit\fP;
1641: unsigned long int \fImatch_limit_recursion\fP;
1642: void *\fIcallout_data\fP;
1643: const unsigned char *\fItables\fP;
1644: unsigned char **\fImark\fP;
1645: .sp
1646: In the 16-bit version of this structure, the \fImark\fP field has type
1647: "PCRE_UCHAR16 **".
1648: .sp
1649: In the 32-bit version of this structure, the \fImark\fP field has type
1650: "PCRE_UCHAR32 **".
1651: .P
1652: The \fIflags\fP field is used to specify which of the other fields are set. The
1653: flag bits are:
1654: .sp
1655: PCRE_EXTRA_CALLOUT_DATA
1656: PCRE_EXTRA_EXECUTABLE_JIT
1657: PCRE_EXTRA_MARK
1658: PCRE_EXTRA_MATCH_LIMIT
1659: PCRE_EXTRA_MATCH_LIMIT_RECURSION
1660: PCRE_EXTRA_STUDY_DATA
1661: PCRE_EXTRA_TABLES
1662: .sp
1663: Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1664: the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1665: returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1666: should not set these yourself, but you may add to the block by setting other
1667: fields and their corresponding flag bits.
1668: .P
1669: The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1670: vast amount of resources when running patterns that are not going to match,
1671: but which have a very large number of possibilities in their search trees. The
1672: classic example is a pattern that uses nested unlimited repeats.
1673: .P
1674: Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1675: calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1676: imposed on the number of times this function is called during a match, which
1677: has the effect of limiting the amount of backtracking that can take place. For
1678: patterns that are not anchored, the count restarts from zero for each position
1679: in the subject string.
1680: .P
1681: When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1682: with a JIT option, the way that the matching is executed is entirely different.
1683: However, there is still the possibility of runaway matching that goes on for a
1684: very long time, and so the \fImatch_limit\fP value is also used in this case
1685: (but in a different way) to limit how long the matching can continue.
1686: .P
1687: The default value for the limit can be set when PCRE is built; the default
1688: default is 10 million, which handles all but the most extreme cases. You can
1689: override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1690: block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1691: the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1692: PCRE_ERROR_MATCHLIMIT.
1693: .P
1694: A value for the match limit may also be supplied by an item at the start of a
1695: pattern of the form
1696: .sp
1697: (*LIMIT_MATCH=d)
1698: .sp
1699: where d is a decimal number. However, such a setting is ignored unless d is
1700: less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1701: is set, less than the default.
1702: .P
1703: The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1704: instead of limiting the total number of times that \fBmatch()\fP is called, it
1705: limits the depth of recursion. The recursion depth is a smaller number than the
1706: total number of calls, because not all calls to \fBmatch()\fP are recursive.
1707: This limit is of use only if it is set smaller than \fImatch_limit\fP.
1708: .P
1709: Limiting the recursion depth limits the amount of machine stack that can be
1710: used, or, when PCRE has been compiled to use memory on the heap instead of the
1711: stack, the amount of heap memory that can be used. This limit is not relevant,
1712: and is ignored, when matching is done using JIT compiled code.
1713: .P
1714: The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1715: built; the default default is the same value as the default for
1716: \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1717: with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1718: PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1719: is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1720: .P
1721: A value for the recursion limit may also be supplied by an item at the start of
1722: a pattern of the form
1723: .sp
1724: (*LIMIT_RECURSION=d)
1725: .sp
1726: where d is a decimal number. However, such a setting is ignored unless d is
1727: less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1728: is set, less than the default.
1729: .P
1730: The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1731: and is described in the
1732: .\" HREF
1733: \fBpcrecallout\fP
1734: .\"
1735: documentation.
1736: .P
1737: The \fItables\fP field is provided for use with patterns that have been
1738: pre-compiled using custom character tables, saved to disc or elsewhere, and
1739: then reloaded, because the tables that were used to compile a pattern are not
1740: saved with it. See the
1741: .\" HREF
1742: \fBpcreprecompile\fP
1743: .\"
1744: documentation for a discussion of saving compiled patterns for later use. If
1745: NULL is passed using this mechanism, it forces PCRE's internal tables to be
1746: used.
1747: .P
1748: \fBWarning:\fP The tables that \fBpcre_exec()\fP uses must be the same as those
1749: that were used when the pattern was compiled. If this is not the case, the
1750: behaviour of \fBpcre_exec()\fP is undefined. Therefore, when a pattern is
1751: compiled and matched in the same process, this field should never be set. In
1752: this (the most common) case, the correct table pointer is automatically passed
1753: with the compiled pattern from \fBpcre_compile()\fP to \fBpcre_exec()\fP.
1754: .P
1755: If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1756: be set to point to a suitable variable. If the pattern contains any
1757: backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1758: a name to pass back, a pointer to the name string (zero terminated) is placed
1759: in the variable pointed to by the \fImark\fP field. The names are within the
1760: compiled pattern; if you wish to retain such a name you must copy it before
1761: freeing the memory of a compiled pattern. If there is no name to pass back, the
1762: variable pointed to by the \fImark\fP field is set to NULL. For details of the
1763: backtracking control verbs, see the section entitled
1764: .\" HTML <a href="pcrepattern#backtrackcontrol">
1765: .\" </a>
1766: "Backtracking control"
1767: .\"
1768: in the
1769: .\" HREF
1770: \fBpcrepattern\fP
1771: .\"
1772: documentation.
1773: .
1774: .
1775: .\" HTML <a name="execoptions"></a>
1776: .SS "Option bits for \fBpcre_exec()\fP"
1777: .rs
1778: .sp
1779: The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1780: zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1781: PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1782: PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1783: PCRE_PARTIAL_SOFT.
1784: .P
1785: If the pattern was successfully studied with one of the just-in-time (JIT)
1786: compile options, the only supported options for JIT execution are
1787: PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1788: PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1789: unsupported option is used, JIT execution is disabled and the normal
1790: interpretive code in \fBpcre_exec()\fP is run.
1791: .sp
1792: PCRE_ANCHORED
1793: .sp
1794: The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1795: matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1796: to be anchored by virtue of its contents, it cannot be made unachored at
1797: matching time.
1798: .sp
1799: PCRE_BSR_ANYCRLF
1800: PCRE_BSR_UNICODE
1801: .sp
1802: These options (which are mutually exclusive) control what the \eR escape
1803: sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1804: match any Unicode newline sequence. These options override the choice that was
1805: made or defaulted when the pattern was compiled.
1806: .sp
1807: PCRE_NEWLINE_CR
1808: PCRE_NEWLINE_LF
1809: PCRE_NEWLINE_CRLF
1810: PCRE_NEWLINE_ANYCRLF
1811: PCRE_NEWLINE_ANY
1812: .sp
1813: These options override the newline definition that was chosen or defaulted when
1814: the pattern was compiled. For details, see the description of
1815: \fBpcre_compile()\fP above. During matching, the newline choice affects the
1816: behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1817: the way the match position is advanced after a match failure for an unanchored
1818: pattern.
1819: .P
1820: When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1821: match attempt for an unanchored pattern fails when the current position is at a
1822: CRLF sequence, and the pattern contains no explicit matches for CR or LF
1823: characters, the match position is advanced by two characters instead of one, in
1824: other words, to after the CRLF.
1825: .P
1826: The above rule is a compromise that makes the most common cases work as
1827: expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1828: set), it does not match the string "\er\enA" because, after failing at the
1829: start, it skips both the CR and the LF before retrying. However, the pattern
1830: [\er\en]A does match that string, because it contains an explicit CR or LF
1831: reference, and so advances only by one character after the first failure.
1832: .P
1833: An explicit match for CR of LF is either a literal appearance of one of those
1834: characters, or one of the \er or \en escape sequences. Implicit matches such as
1835: [^X] do not count, nor does \es (which includes CR and LF in the characters
1836: that it matches).
1837: .P
1838: Notwithstanding the above, anomalous effects may still occur when CRLF is a
1839: valid newline sequence and explicit \er or \en escapes appear in the pattern.
1840: .sp
1841: PCRE_NOTBOL
1842: .sp
1843: This option specifies that first character of the subject string is not the
1844: beginning of a line, so the circumflex metacharacter should not match before
1845: it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1846: never to match. This option affects only the behaviour of the circumflex
1847: metacharacter. It does not affect \eA.
1848: .sp
1849: PCRE_NOTEOL
1850: .sp
1851: This option specifies that the end of the subject string is not the end of a
1852: line, so the dollar metacharacter should not match it nor (except in multiline
1853: mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1854: compile time) causes dollar never to match. This option affects only the
1855: behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1856: .sp
1857: PCRE_NOTEMPTY
1858: .sp
1859: An empty string is not considered to be a valid match if this option is set. If
1860: there are alternatives in the pattern, they are tried. If all the alternatives
1861: match the empty string, the entire match fails. For example, if the pattern
1862: .sp
1863: a?b?
1864: .sp
1865: is applied to a string not beginning with "a" or "b", it matches an empty
1866: string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1867: valid, so PCRE searches further into the string for occurrences of "a" or "b".
1868: .sp
1869: PCRE_NOTEMPTY_ATSTART
1870: .sp
1871: This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1872: the start of the subject is permitted. If the pattern is anchored, such a match
1873: can occur only if the pattern contains \eK.
1874: .P
1875: Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1876: does make a special case of a pattern match of the empty string within its
1877: \fBsplit()\fP function, and when using the /g modifier. It is possible to
1878: emulate Perl's behaviour after matching a null string by first trying the match
1879: again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1880: if that fails, by advancing the starting offset (see below) and trying an
1881: ordinary match again. There is some code that demonstrates how to do this in
1882: the
1883: .\" HREF
1884: \fBpcredemo\fP
1885: .\"
1886: sample program. In the most general case, you have to check to see if the
1887: newline convention recognizes CRLF as a newline, and if so, and the current
1888: character is CR followed by LF, advance the starting offset by two characters
1889: instead of one.
1890: .sp
1891: PCRE_NO_START_OPTIMIZE
1892: .sp
1893: There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1894: a match, in order to speed up the process. For example, if it is known that an
1895: unanchored match must start with a specific character, it searches the subject
1896: for that character, and fails immediately if it cannot find it, without
1897: actually running the main matching function. This means that a special item
1898: such as (*COMMIT) at the start of a pattern is not considered until after a
1899: suitable starting point for the match has been found. Also, when callouts or
1900: (*MARK) items are in use, these "start-up" optimizations can cause them to be
1901: skipped if the pattern is never actually used. The start-up optimizations are
1902: in effect a pre-scan of the subject that takes place before the pattern is run.
1903: .P
1904: The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1905: causing performance to suffer, but ensuring that in cases where the result is
1906: "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1907: are considered at every possible starting position in the subject string. If
1908: PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1909: time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1910: to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1911: always done using interpretively.
1912: .P
1913: Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1914: Consider the pattern
1915: .sp
1916: (*COMMIT)ABC
1917: .sp
1918: When this is compiled, PCRE records the fact that a match must start with the
1919: character "A". Suppose the subject string is "DEFABC". The start-up
1920: optimization scans along the subject, finds "A" and runs the first match
1921: attempt from there. The (*COMMIT) item means that the pattern must match the
1922: current starting position, which in this case, it does. However, if the same
1923: match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1924: subject string does not happen. The first match attempt is run starting from
1925: "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1926: the overall result is "no match". If the pattern is studied, more start-up
1927: optimizations may be used. For example, a minimum length for the subject may be
1928: recorded. Consider the pattern
1929: .sp
1930: (*MARK:A)(X|Y)
1931: .sp
1932: The minimum length for a match is one character. If the subject is "ABC", there
1933: will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1934: If the pattern is studied, the final attempt does not take place, because PCRE
1935: knows that the subject is too short, and so the (*MARK) is never encountered.
1936: In this case, studying the pattern does not affect the overall match result,
1937: which is still "no match", but it does affect the auxiliary information that is
1938: returned.
1939: .sp
1940: PCRE_NO_UTF8_CHECK
1941: .sp
1942: When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1943: string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1944: The entire string is checked before any other processing takes place. The value
1945: of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1946: UTF-8 character. There is a discussion about the
1947: .\" HTML <a href="pcreunicode.html#utf8strings">
1948: .\" </a>
1949: validity of UTF-8 strings
1950: .\"
1951: in the
1952: .\" HREF
1953: \fBpcreunicode\fP
1954: .\"
1955: page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1956: error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1957: truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1958: cases, information about the precise nature of the error may also be returned
1959: (see the descriptions of these errors in the section entitled \fIError return
1960: values from\fP \fBpcre_exec()\fP
1961: .\" HTML <a href="#errorlist">
1962: .\" </a>
1963: below).
1964: .\"
1965: If \fIstartoffset\fP contains a value that does not point to the start of a
1966: UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1967: returned.
1968: .P
1969: If you already know that your subject is valid, and you want to skip these
1970: checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1971: calling \fBpcre_exec()\fP. You might want to do this for the second and
1972: subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1973: all the matches in a single subject string. However, you should be sure that
1974: the value of \fIstartoffset\fP points to the start of a character (or the end
1975: of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1976: invalid string as a subject or an invalid value of \fIstartoffset\fP is
1977: undefined. Your program may crash or loop.
1978: .sp
1979: PCRE_PARTIAL_HARD
1980: PCRE_PARTIAL_SOFT
1981: .sp
1982: These options turn on the partial matching feature. For backwards
1983: compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1984: occurs if the end of the subject string is reached successfully, but there are
1985: not enough subject characters to complete the match. If this happens when
1986: PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1987: testing any remaining alternatives. Only if no complete match can be found is
1988: PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1989: PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1990: but only if no complete match can be found.
1991: .P
1992: If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1993: partial match is found, \fBpcre_exec()\fP immediately returns
1994: PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1995: when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1996: important that an alternative complete match.
1997: .P
1998: In both cases, the portion of the string that was inspected when the partial
1999: match was found is set as the first matching string. There is a more detailed
2000: discussion of partial and multi-segment matching, with examples, in the
2001: .\" HREF
2002: \fBpcrepartial\fP
2003: .\"
2004: documentation.
2005: .
2006: .
2007: .SS "The string to be matched by \fBpcre_exec()\fP"
2008: .rs
2009: .sp
2010: The subject string is passed to \fBpcre_exec()\fP as a pointer in
2011: \fIsubject\fP, a length in \fIlength\fP, and a starting offset in
2012: \fIstartoffset\fP. The units for \fIlength\fP and \fIstartoffset\fP are bytes
2013: for the 8-bit library, 16-bit data items for the 16-bit library, and 32-bit
2014: data items for the 32-bit library.
2015: .P
2016: If \fIstartoffset\fP is negative or greater than the length of the subject,
2017: \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting offset is
2018: zero, the search for a match starts at the beginning of the subject, and this
2019: is by far the most common case. In UTF-8 or UTF-16 mode, the offset must point
2020: to the start of a character, or the end of the subject (in UTF-32 mode, one
2021: data unit equals one character, so all offsets are valid). Unlike the pattern
2022: string, the subject may contain binary zeroes.
2023: .P
2024: A non-zero starting offset is useful when searching for another match in the
2025: same subject by calling \fBpcre_exec()\fP again after a previous success.
2026: Setting \fIstartoffset\fP differs from just passing over a shortened string and
2027: setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
2028: lookbehind. For example, consider the pattern
2029: .sp
2030: \eBiss\eB
2031: .sp
2032: which finds occurrences of "iss" in the middle of words. (\eB matches only if
2033: the current position in the subject is not a word boundary.) When applied to
2034: the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
2035: occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
2036: subject, namely "issipi", it does not match, because \eB is always false at the
2037: start of the subject, which is deemed to be a word boundary. However, if
2038: \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
2039: set to 4, it finds the second occurrence of "iss" because it is able to look
2040: behind the starting point to discover that it is preceded by a letter.
2041: .P
2042: Finding all the matches in a subject is tricky when the pattern can match an
2043: empty string. It is possible to emulate Perl's /g behaviour by first trying the
2044: match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2045: PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2046: and trying an ordinary match again. There is some code that demonstrates how to
2047: do this in the
2048: .\" HREF
2049: \fBpcredemo\fP
2050: .\"
2051: sample program. In the most general case, you have to check to see if the
2052: newline convention recognizes CRLF as a newline, and if so, and the current
2053: character is CR followed by LF, advance the starting offset by two characters
2054: instead of one.
2055: .P
2056: If a non-zero starting offset is passed when the pattern is anchored, one
2057: attempt to match at the given offset is made. This can only succeed if the
2058: pattern does not require the match to be at the start of the subject.
2059: .
2060: .
2061: .SS "How \fBpcre_exec()\fP returns captured substrings"
2062: .rs
2063: .sp
2064: In general, a pattern matches a certain portion of the subject, and in
2065: addition, further substrings from the subject may be picked out by parts of the
2066: pattern. Following the usage in Jeffrey Friedl's book, this is called
2067: "capturing" in what follows, and the phrase "capturing subpattern" is used for
2068: a fragment of a pattern that picks out a substring. PCRE supports several other
2069: kinds of parenthesized subpattern that do not cause substrings to be captured.
2070: .P
2071: Captured substrings are returned to the caller via a vector of integers whose
2072: address is passed in \fIovector\fP. The number of elements in the vector is
2073: passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2074: argument is NOT the size of \fIovector\fP in bytes.
2075: .P
2076: The first two-thirds of the vector is used to pass back captured substrings,
2077: each substring using a pair of integers. The remaining third of the vector is
2078: used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2079: and is not available for passing back information. The number passed in
2080: \fIovecsize\fP should always be a multiple of three. If it is not, it is
2081: rounded down.
2082: .P
2083: When a match is successful, information about captured substrings is returned
2084: in pairs of integers, starting at the beginning of \fIovector\fP, and
2085: continuing up to two-thirds of its length at the most. The first element of
2086: each pair is set to the offset of the first character in a substring, and the
2087: second is set to the offset of the first character after the end of a
2088: substring. These values are always data unit offsets, even in UTF mode. They
2089: are byte offsets in the 8-bit library, 16-bit data item offsets in the 16-bit
2090: library, and 32-bit data item offsets in the 32-bit library. \fBNote\fP: they
2091: are not character counts.
2092: .P
2093: The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2094: portion of the subject string matched by the entire pattern. The next pair is
2095: used for the first capturing subpattern, and so on. The value returned by
2096: \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2097: For example, if two substrings have been captured, the returned value is 3. If
2098: there are no capturing subpatterns, the return value from a successful match is
2099: 1, indicating that just the first pair of offsets has been set.
2100: .P
2101: If a capturing subpattern is matched repeatedly, it is the last portion of the
2102: string that it matched that is returned.
2103: .P
2104: If the vector is too small to hold all the captured substring offsets, it is
2105: used as far as possible (up to two-thirds of its length), and the function
2106: returns a value of zero. If neither the actual string matched nor any captured
2107: substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2108: passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2109: back references and the \fIovector\fP is not big enough to remember the related
2110: substrings, PCRE has to get additional memory for use during matching. Thus it
2111: is usually advisable to supply an \fIovector\fP of reasonable size.
2112: .P
2113: There are some cases where zero is returned (indicating vector overflow) when
2114: in fact the vector is exactly the right size for the final match. For example,
2115: consider the pattern
2116: .sp
2117: (a)(?:(b)c|bd)
2118: .sp
2119: If a vector of 6 elements (allowing for only 1 captured substring) is given
2120: with subject string "abd", \fBpcre_exec()\fP will try to set the second
2121: captured string, thereby recording a vector overflow, before failing to match
2122: "c" and backing up to try the second alternative. The zero return, however,
2123: does correctly indicate that the maximum number of slots (namely 2) have been
2124: filled. In similar cases where there is temporary overflow, but the final
2125: number of used slots is actually less than the maximum, a non-zero value is
2126: returned.
2127: .P
2128: The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2129: subpatterns there are in a compiled pattern. The smallest size for
2130: \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2131: the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2132: .P
2133: It is possible for capturing subpattern number \fIn+1\fP to match some part of
2134: the subject when subpattern \fIn\fP has not been used at all. For example, if
2135: the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2136: function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2137: happens, both values in the offset pairs corresponding to unused subpatterns
2138: are set to -1.
2139: .P
2140: Offset values that correspond to unused subpatterns at the end of the
2141: expression are also set to -1. For example, if the string "abc" is matched
2142: against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2143: return from the function is 2, because the highest used capturing subpattern
2144: number is 1, and the offsets for for the second and third capturing subpatterns
2145: (assuming the vector is large enough, of course) are set to -1.
2146: .P
2147: \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2148: correspond to capturing parentheses in the pattern are never changed. That is,
2149: if a pattern contains \fIn\fP capturing parentheses, no more than
2150: \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2151: elements (in the first two-thirds) retain whatever values they previously had.
2152: .P
2153: Some convenience functions are provided for extracting the captured substrings
2154: as separate strings. These are described below.
2155: .
2156: .
2157: .\" HTML <a name="errorlist"></a>
2158: .SS "Error return values from \fBpcre_exec()\fP"
2159: .rs
2160: .sp
2161: If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2162: defined in the header file:
2163: .sp
2164: PCRE_ERROR_NOMATCH (-1)
2165: .sp
2166: The subject string did not match the pattern.
2167: .sp
2168: PCRE_ERROR_NULL (-2)
2169: .sp
2170: Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2171: NULL and \fIovecsize\fP was not zero.
2172: .sp
2173: PCRE_ERROR_BADOPTION (-3)
2174: .sp
2175: An unrecognized bit was set in the \fIoptions\fP argument.
2176: .sp
2177: PCRE_ERROR_BADMAGIC (-4)
2178: .sp
2179: PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2180: the case when it is passed a junk pointer and to detect when a pattern that was
2181: compiled in an environment of one endianness is run in an environment with the
2182: other endianness. This is the error that PCRE gives when the magic number is
2183: not present.
2184: .sp
2185: PCRE_ERROR_UNKNOWN_OPCODE (-5)
2186: .sp
2187: While running the pattern match, an unknown item was encountered in the
2188: compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2189: of the compiled pattern.
2190: .sp
2191: PCRE_ERROR_NOMEMORY (-6)
2192: .sp
2193: If a pattern contains back references, but the \fIovector\fP that is passed to
2194: \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2195: gets a block of memory at the start of matching to use for this purpose. If the
2196: call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2197: automatically freed at the end of matching.
2198: .P
2199: This error is also given if \fBpcre_stack_malloc()\fP fails in
2200: \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2201: \fB--disable-stack-for-recursion\fP.
2202: .sp
2203: PCRE_ERROR_NOSUBSTRING (-7)
2204: .sp
2205: This error is used by the \fBpcre_copy_substring()\fP,
2206: \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2207: below). It is never returned by \fBpcre_exec()\fP.
2208: .sp
2209: PCRE_ERROR_MATCHLIMIT (-8)
2210: .sp
2211: The backtracking limit, as specified by the \fImatch_limit\fP field in a
2212: \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2213: above.
2214: .sp
2215: PCRE_ERROR_CALLOUT (-9)
2216: .sp
2217: This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2218: use by callout functions that want to yield a distinctive error code. See the
2219: .\" HREF
2220: \fBpcrecallout\fP
2221: .\"
2222: documentation for details.
2223: .sp
2224: PCRE_ERROR_BADUTF8 (-10)
2225: .sp
2226: A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2227: and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2228: (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2229: UTF-8 character is placed in the first element, and a reason code is placed in
2230: the second element. The reason codes are listed in the
2231: .\" HTML <a href="#badutf8reasons">
2232: .\" </a>
2233: following section.
2234: .\"
2235: For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2236: truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2237: PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2238: .sp
2239: PCRE_ERROR_BADUTF8_OFFSET (-11)
2240: .sp
2241: The UTF-8 byte sequence that was passed as a subject was checked and found to
2242: be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2243: \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2244: end of the subject.
2245: .sp
2246: PCRE_ERROR_PARTIAL (-12)
2247: .sp
2248: The subject string did not match, but it did match partially. See the
2249: .\" HREF
2250: \fBpcrepartial\fP
2251: .\"
2252: documentation for details of partial matching.
2253: .sp
2254: PCRE_ERROR_BADPARTIAL (-13)
2255: .sp
2256: This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2257: option was used with a compiled pattern containing items that were not
2258: supported for partial matching. From release 8.00 onwards, there are no
2259: restrictions on partial matching.
2260: .sp
2261: PCRE_ERROR_INTERNAL (-14)
2262: .sp
2263: An unexpected internal error has occurred. This error could be caused by a bug
2264: in PCRE or by overwriting of the compiled pattern.
2265: .sp
2266: PCRE_ERROR_BADCOUNT (-15)
2267: .sp
2268: This error is given if the value of the \fIovecsize\fP argument is negative.
2269: .sp
2270: PCRE_ERROR_RECURSIONLIMIT (-21)
2271: .sp
2272: The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2273: field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2274: description above.
2275: .sp
2276: PCRE_ERROR_BADNEWLINE (-23)
2277: .sp
2278: An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2279: .sp
2280: PCRE_ERROR_BADOFFSET (-24)
2281: .sp
2282: The value of \fIstartoffset\fP was negative or greater than the length of the
2283: subject, that is, the value in \fIlength\fP.
2284: .sp
2285: PCRE_ERROR_SHORTUTF8 (-25)
2286: .sp
2287: This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2288: ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2289: Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2290: fact sufficient to detect this case, but this special error code for
2291: PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2292: retained for backwards compatibility.
2293: .sp
2294: PCRE_ERROR_RECURSELOOP (-26)
2295: .sp
2296: This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2297: the pattern. Specifically, it means that either the whole pattern or a
2298: subpattern has been called recursively for the second time at the same position
2299: in the subject string. Some simple patterns that might do this are detected and
2300: faulted at compile time, but more complicated cases, in particular mutual
2301: recursions between two different subpatterns, cannot be detected until run
2302: time.
2303: .sp
2304: PCRE_ERROR_JIT_STACKLIMIT (-27)
2305: .sp
2306: This error is returned when a pattern that was successfully studied using a
2307: JIT compile option is being matched, but the memory available for the
2308: just-in-time processing stack is not large enough. See the
2309: .\" HREF
2310: \fBpcrejit\fP
2311: .\"
2312: documentation for more details.
2313: .sp
2314: PCRE_ERROR_BADMODE (-28)
2315: .sp
2316: This error is given if a pattern that was compiled by the 8-bit library is
2317: passed to a 16-bit or 32-bit library function, or vice versa.
2318: .sp
2319: PCRE_ERROR_BADENDIANNESS (-29)
2320: .sp
2321: This error is given if a pattern that was compiled and saved is reloaded on a
2322: host with different endianness. The utility function
2323: \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2324: so that it runs on the new host.
2325: .sp
2326: PCRE_ERROR_JIT_BADOPTION
2327: .sp
2328: This error is returned when a pattern that was successfully studied using a JIT
2329: compile option is being matched, but the matching mode (partial or complete
2330: match) does not correspond to any JIT compilation mode. When the JIT fast path
2331: function is used, this error may be also given for invalid options. See the
2332: .\" HREF
2333: \fBpcrejit\fP
2334: .\"
2335: documentation for more details.
2336: .sp
2337: PCRE_ERROR_BADLENGTH (-32)
2338: .sp
2339: This error is given if \fBpcre_exec()\fP is called with a negative value for
2340: the \fIlength\fP argument.
2341: .P
2342: Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2343: .
2344: .
2345: .\" HTML <a name="badutf8reasons"></a>
2346: .SS "Reason codes for invalid UTF-8 strings"
2347: .rs
2348: .sp
2349: This section applies only to the 8-bit library. The corresponding information
2350: for the 16-bit and 32-bit libraries is given in the
2351: .\" HREF
2352: \fBpcre16\fP
2353: .\"
2354: and
2355: .\" HREF
2356: \fBpcre32\fP
2357: .\"
2358: pages.
2359: .P
2360: When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2361: PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2362: least 2, the offset of the start of the invalid UTF-8 character is placed in
2363: the first output vector element (\fIovector[0]\fP) and a reason code is placed
2364: in the second element (\fIovector[1]\fP). The reason codes are given names in
2365: the \fBpcre.h\fP header file:
2366: .sp
2367: PCRE_UTF8_ERR1
2368: PCRE_UTF8_ERR2
2369: PCRE_UTF8_ERR3
2370: PCRE_UTF8_ERR4
2371: PCRE_UTF8_ERR5
2372: .sp
2373: The string ends with a truncated UTF-8 character; the code specifies how many
2374: bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2375: no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2376: allows for up to 6 bytes, and this is checked first; hence the possibility of
2377: 4 or 5 missing bytes.
2378: .sp
2379: PCRE_UTF8_ERR6
2380: PCRE_UTF8_ERR7
2381: PCRE_UTF8_ERR8
2382: PCRE_UTF8_ERR9
2383: PCRE_UTF8_ERR10
2384: .sp
2385: The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2386: character do not have the binary value 0b10 (that is, either the most
2387: significant bit is 0, or the next bit is 1).
2388: .sp
2389: PCRE_UTF8_ERR11
2390: PCRE_UTF8_ERR12
2391: .sp
2392: A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2393: these code points are excluded by RFC 3629.
2394: .sp
2395: PCRE_UTF8_ERR13
2396: .sp
2397: A 4-byte character has a value greater than 0x10fff; these code points are
2398: excluded by RFC 3629.
2399: .sp
2400: PCRE_UTF8_ERR14
2401: .sp
2402: A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2403: code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2404: from UTF-8.
2405: .sp
2406: PCRE_UTF8_ERR15
2407: PCRE_UTF8_ERR16
2408: PCRE_UTF8_ERR17
2409: PCRE_UTF8_ERR18
2410: PCRE_UTF8_ERR19
2411: .sp
2412: A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2413: value that can be represented by fewer bytes, which is invalid. For example,
2414: the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2415: one byte.
2416: .sp
2417: PCRE_UTF8_ERR20
2418: .sp
2419: The two most significant bits of the first byte of a character have the binary
2420: value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2421: byte can only validly occur as the second or subsequent byte of a multi-byte
2422: character.
2423: .sp
2424: PCRE_UTF8_ERR21
2425: .sp
2426: The first byte of a character has the value 0xfe or 0xff. These values can
2427: never occur in a valid UTF-8 string.
2428: .sp
2429: PCRE_UTF8_ERR22
2430: .sp
2431: This error code was formerly used when the presence of a so-called
2432: "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2433: such characters should not cause a string to be rejected, and so this code is
2434: no longer in use and is never returned.
2435: .
2436: .
2437: .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2438: .rs
2439: .sp
2440: .nf
2441: .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2442: .B " int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,"
2443: .B " int \fIbuffersize\fP);"
2444: .sp
2445: .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2446: .B " int \fIstringcount\fP, int \fIstringnumber\fP,"
2447: .B " const char **\fIstringptr\fP);"
2448: .sp
2449: .B int pcre_get_substring_list(const char *\fIsubject\fP,
2450: .B " int *\fIovector\fP, int \fIstringcount\fP, const char ***\fIlistptr\fP);"
2451: .fi
2452: .PP
2453: Captured substrings can be accessed directly by using the offsets returned by
2454: \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2455: \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2456: \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2457: as new, separate, zero-terminated strings. These functions identify substrings
2458: by number. The next section describes functions for extracting named
2459: substrings.
2460: .P
2461: A substring that contains a binary zero is correctly extracted and has a
2462: further zero added on the end, but the result is not, of course, a C string.
2463: However, you can process such a string by referring to the length that is
2464: returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2465: Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2466: for handling strings containing binary zeros, because the end of the final
2467: string is not independently indicated.
2468: .P
2469: The first three arguments are the same for all three of these functions:
2470: \fIsubject\fP is the subject string that has just been successfully matched,
2471: \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2472: \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2473: captured by the match, including the substring that matched the entire regular
2474: expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2475: than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2476: space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2477: number of elements in the vector divided by three.
2478: .P
2479: The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2480: extract a single substring, whose number is given as \fIstringnumber\fP. A
2481: value of zero extracts the substring that matched the entire pattern, whereas
2482: higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2483: the string is placed in \fIbuffer\fP, whose length is given by
2484: \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2485: obtained via \fBpcre_malloc\fP, and its address is returned via
2486: \fIstringptr\fP. The yield of the function is the length of the string, not
2487: including the terminating zero, or one of these error codes:
2488: .sp
2489: PCRE_ERROR_NOMEMORY (-6)
2490: .sp
2491: The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2492: memory failed for \fBpcre_get_substring()\fP.
2493: .sp
2494: PCRE_ERROR_NOSUBSTRING (-7)
2495: .sp
2496: There is no substring whose number is \fIstringnumber\fP.
2497: .P
2498: The \fBpcre_get_substring_list()\fP function extracts all available substrings
2499: and builds a list of pointers to them. All this is done in a single block of
2500: memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2501: is returned via \fIlistptr\fP, which is also the start of the list of string
2502: pointers. The end of the list is marked by a NULL pointer. The yield of the
2503: function is zero if all went well, or the error code
2504: .sp
2505: PCRE_ERROR_NOMEMORY (-6)
2506: .sp
2507: if the attempt to get the memory block failed.
2508: .P
2509: When any of these functions encounter a substring that is unset, which can
2510: happen when capturing subpattern number \fIn+1\fP matches some part of the
2511: subject, but subpattern \fIn\fP has not been used at all, they return an empty
2512: string. This can be distinguished from a genuine zero-length substring by
2513: inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2514: substrings.
2515: .P
2516: The two convenience functions \fBpcre_free_substring()\fP and
2517: \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2518: a previous call of \fBpcre_get_substring()\fP or
2519: \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2520: the function pointed to by \fBpcre_free\fP, which of course could be called
2521: directly from a C program. However, PCRE is used in some situations where it is
2522: linked via a special interface to another programming language that cannot use
2523: \fBpcre_free\fP directly; it is for these cases that the functions are
2524: provided.
2525: .
2526: .
2527: .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2528: .rs
2529: .sp
2530: .nf
2531: .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2532: .B " const char *\fIname\fP);"
2533: .sp
2534: .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2535: .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2536: .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2537: .B " char *\fIbuffer\fP, int \fIbuffersize\fP);"
2538: .sp
2539: .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2540: .B " const char *\fIsubject\fP, int *\fIovector\fP,"
2541: .B " int \fIstringcount\fP, const char *\fIstringname\fP,"
2542: .B " const char **\fIstringptr\fP);"
2543: .fi
2544: .PP
2545: To extract a substring by name, you first have to find associated number.
2546: For example, for this pattern
2547: .sp
2548: (a+)b(?<xxx>\ed+)...
2549: .sp
2550: the number of the subpattern called "xxx" is 2. If the name is known to be
2551: unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2552: calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2553: pattern, and the second is the name. The yield of the function is the
2554: subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2555: that name.
2556: .P
2557: Given the number, you can extract the substring directly, or use one of the
2558: functions described in the previous section. For convenience, there are also
2559: two functions that do the whole job.
2560: .P
2561: Most of the arguments of \fBpcre_copy_named_substring()\fP and
2562: \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2563: functions that extract by number. As these are described in the previous
2564: section, they are not re-described here. There are just two differences:
2565: .P
2566: First, instead of a substring number, a substring name is given. Second, there
2567: is an extra argument, given at the start, which is a pointer to the compiled
2568: pattern. This is needed in order to gain access to the name-to-number
2569: translation table.
2570: .P
2571: These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2572: then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2573: appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2574: the behaviour may not be what you want (see the next section).
2575: .P
2576: \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2577: subpatterns with the same number, as described in the
2578: .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2579: .\" </a>
2580: section on duplicate subpattern numbers
2581: .\"
2582: in the
2583: .\" HREF
2584: \fBpcrepattern\fP
2585: .\"
2586: page, you cannot use names to distinguish the different subpatterns, because
2587: names are not included in the compiled code. The matching process uses only
2588: numbers. For this reason, the use of different names for subpatterns of the
2589: same number causes an error at compile time.
2590: .
2591: .
2592: .SH "DUPLICATE SUBPATTERN NAMES"
2593: .rs
2594: .sp
2595: .nf
2596: .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2597: .B " const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);"
2598: .fi
2599: .PP
2600: When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2601: are not required to be unique. (Duplicate names are always allowed for
2602: subpatterns with the same number, created by using the (?| feature. Indeed, if
2603: such subpatterns are named, they are required to use the same names.)
2604: .P
2605: Normally, patterns with duplicate names are such that in any one match, only
2606: one of the named subpatterns participates. An example is shown in the
2607: .\" HREF
2608: \fBpcrepattern\fP
2609: .\"
2610: documentation.
2611: .P
2612: When duplicates are present, \fBpcre_copy_named_substring()\fP and
2613: \fBpcre_get_named_substring()\fP return the first substring corresponding to
2614: the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2615: returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2616: returns one of the numbers that are associated with the name, but it is not
2617: defined which it is.
2618: .P
2619: If you want to get full details of all captured substrings for a given name,
2620: you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2621: argument is the compiled pattern, and the second is the name. The third and
2622: fourth are pointers to variables which are updated by the function. After it
2623: has run, they point to the first and last entries in the name-to-number table
2624: for the given name. The function itself returns the length of each entry, or
2625: PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2626: described above in the section entitled \fIInformation about a pattern\fP
2627: .\" HTML <a href="#infoaboutpattern">
2628: .\" </a>
2629: above.
2630: .\"
2631: Given all the relevant entries for the name, you can extract each of their
2632: numbers, and hence the captured data, if any.
2633: .
2634: .
2635: .SH "FINDING ALL POSSIBLE MATCHES"
2636: .rs
2637: .sp
2638: The traditional matching function uses a similar algorithm to Perl, which stops
2639: when it finds the first match, starting at a given point in the subject. If you
2640: want to find all possible matches, or the longest possible match, consider
2641: using the alternative matching function (see below) instead. If you cannot use
2642: the alternative function, but still need to find all possible matches, you
2643: can kludge it up by making use of the callout facility, which is described in
2644: the
2645: .\" HREF
2646: \fBpcrecallout\fP
2647: .\"
2648: documentation.
2649: .P
2650: What you have to do is to insert a callout right at the end of the pattern.
2651: When your callout function is called, extract and save the current matched
2652: substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2653: other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2654: will yield PCRE_ERROR_NOMATCH.
2655: .
2656: .
2657: .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2658: .rs
2659: .sp
2660: Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2661: stack, which in certain environments can be rather limited in size. Some users
2662: find it helpful to have an estimate of the amount of stack that is used by
2663: \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2664: .\" HREF
2665: \fBpcrestack\fP
2666: .\"
2667: documentation. The estimate that is output by \fBpcretest\fP when called with
2668: the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2669: the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2670: .P
2671: Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2672: the negative error code PCRE_ERROR_NULL, but with this special combination of
2673: arguments, it returns instead a negative number whose absolute value is the
2674: approximate stack frame size in bytes. (A negative number is used so that it is
2675: clear that no match has happened.) The value is approximate because in some
2676: cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2677: additional variables on the stack.
2678: .P
2679: If PCRE has been compiled to use the heap instead of the stack for recursion,
2680: the value returned is the size of each block that is obtained from the heap.
2681: .
2682: .
2683: .\" HTML <a name="dfamatch"></a>
2684: .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2685: .rs
2686: .sp
2687: .nf
2688: .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2689: .B " const char *\fIsubject\fP, int \fIlength\fP, int \fIstartoffset\fP,"
2690: .B " int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,"
2691: .B " int *\fIworkspace\fP, int \fIwscount\fP);"
2692: .fi
2693: .P
2694: The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2695: a compiled pattern, using a matching algorithm that scans the subject string
2696: just once, and does not backtrack. This has different characteristics to the
2697: normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2698: patterns are not supported. Nevertheless, there are times when this kind of
2699: matching can be useful. For a discussion of the two matching algorithms, and a
2700: list of features that \fBpcre_dfa_exec()\fP does not support, see the
2701: .\" HREF
2702: \fBpcrematching\fP
2703: .\"
2704: documentation.
2705: .P
2706: The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2707: \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2708: different way, and this is described below. The other common arguments are used
2709: in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2710: here.
2711: .P
2712: The two additional arguments provide workspace for the function. The workspace
2713: vector should contain at least 20 elements. It is used for keeping track of
2714: multiple paths through the pattern tree. More workspace will be needed for
2715: patterns and subjects where there are a lot of potential matches.
2716: .P
2717: Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2718: .sp
2719: int rc;
2720: int ovector[10];
2721: int wspace[20];
2722: rc = pcre_dfa_exec(
2723: re, /* result of pcre_compile() */
2724: NULL, /* we didn't study the pattern */
2725: "some string", /* the subject string */
2726: 11, /* the length of the subject string */
2727: 0, /* start at offset 0 in the subject */
2728: 0, /* default options */
2729: ovector, /* vector of integers for substring information */
2730: 10, /* number of elements (NOT size in bytes) */
2731: wspace, /* working space vector */
2732: 20); /* number of elements (NOT size in bytes) */
2733: .
2734: .SS "Option bits for \fBpcre_dfa_exec()\fP"
2735: .rs
2736: .sp
2737: The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2738: zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2739: PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2740: PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2741: PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2742: All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2743: so their description is not repeated here.
2744: .sp
2745: PCRE_PARTIAL_HARD
2746: PCRE_PARTIAL_SOFT
2747: .sp
2748: These have the same general effect as they do for \fBpcre_exec()\fP, but the
2749: details are slightly different. When PCRE_PARTIAL_HARD is set for
2750: \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2751: is reached and there is still at least one matching possibility that requires
2752: additional characters. This happens even if some complete matches have also
2753: been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2754: is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2755: there have been no complete matches, but there is still at least one matching
2756: possibility. The portion of the string that was inspected when the longest
2757: partial match was found is set as the first matching string in both cases.
2758: There is a more detailed discussion of partial and multi-segment matching, with
2759: examples, in the
2760: .\" HREF
2761: \fBpcrepartial\fP
2762: .\"
2763: documentation.
2764: .sp
2765: PCRE_DFA_SHORTEST
2766: .sp
2767: Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2768: soon as it has found one match. Because of the way the alternative algorithm
2769: works, this is necessarily the shortest possible match at the first possible
2770: matching point in the subject string.
2771: .sp
2772: PCRE_DFA_RESTART
2773: .sp
2774: When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2775: again, with additional subject characters, and have it continue with the same
2776: match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2777: \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2778: before because data about the match so far is left in them after a partial
2779: match. There is more discussion of this facility in the
2780: .\" HREF
2781: \fBpcrepartial\fP
2782: .\"
2783: documentation.
2784: .
2785: .
2786: .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2787: .rs
2788: .sp
2789: When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2790: substring in the subject. Note, however, that all the matches from one run of
2791: the function start at the same point in the subject. The shorter matches are
2792: all initial substrings of the longer matches. For example, if the pattern
2793: .sp
2794: <.*>
2795: .sp
2796: is matched against the string
2797: .sp
2798: This is <something> <something else> <something further> no more
2799: .sp
2800: the three matched strings are
2801: .sp
2802: <something>
2803: <something> <something else>
2804: <something> <something else> <something further>
2805: .sp
2806: On success, the yield of the function is a number greater than zero, which is
2807: the number of matched substrings. The substrings themselves are returned in
2808: \fIovector\fP. Each string uses two elements; the first is the offset to the
2809: start, and the second is the offset to the end. In fact, all the strings have
2810: the same start offset. (Space could have been saved by giving this only once,
2811: but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2812: returns data, even though the meaning of the strings is different.)
2813: .P
2814: The strings are returned in reverse order of length; that is, the longest
2815: matching string is given first. If there were too many matches to fit into
2816: \fIovector\fP, the yield of the function is zero, and the vector is filled with
2817: the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2818: the entire \fIovector\fP for returning matched strings.
2819: .P
2820: NOTE: PCRE's "auto-possessification" optimization usually applies to character
2821: repeats at the end of a pattern (as well as internally). For example, the
2822: pattern "a\ed+" is compiled as if it were "a\ed++" because there is no point
2823: even considering the possibility of backtracking into the repeated digits. For
2824: DFA matching, this means that only one possible match is found. If you really
2825: do want multiple matches in such cases, either use an ungreedy repeat
2826: ("a\ed+?") or set the PCRE_NO_AUTO_POSSESS option when compiling.
2827: .
2828: .
2829: .SS "Error returns from \fBpcre_dfa_exec()\fP"
2830: .rs
2831: .sp
2832: The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2833: Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2834: described
2835: .\" HTML <a href="#errorlist">
2836: .\" </a>
2837: above.
2838: .\"
2839: There are in addition the following errors that are specific to
2840: \fBpcre_dfa_exec()\fP:
2841: .sp
2842: PCRE_ERROR_DFA_UITEM (-16)
2843: .sp
2844: This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2845: that it does not support, for instance, the use of \eC or a back reference.
2846: .sp
2847: PCRE_ERROR_DFA_UCOND (-17)
2848: .sp
2849: This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2850: uses a back reference for the condition, or a test for recursion in a specific
2851: group. These are not supported.
2852: .sp
2853: PCRE_ERROR_DFA_UMLIMIT (-18)
2854: .sp
2855: This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2856: block that contains a setting of the \fImatch_limit\fP or
2857: \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2858: meaningless for DFA matching).
2859: .sp
2860: PCRE_ERROR_DFA_WSSIZE (-19)
2861: .sp
2862: This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2863: \fIworkspace\fP vector.
2864: .sp
2865: PCRE_ERROR_DFA_RECURSE (-20)
2866: .sp
2867: When a recursive subpattern is processed, the matching function calls itself
2868: recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2869: error is given if the output vector is not large enough. This should be
2870: extremely rare, as a vector of size 1000 is used.
2871: .sp
2872: PCRE_ERROR_DFA_BADRESTART (-30)
2873: .sp
2874: When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2875: some plausibility checks are made on the contents of the workspace, which
2876: should contain data about the previous partial match. If any of these checks
2877: fail, this error is given.
2878: .
2879: .
2880: .SH "SEE ALSO"
2881: .rs
2882: .sp
2883: \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2884: \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2885: \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2886: \fBpcrestack\fP(3).
2887: .
2888: .
2889: .SH AUTHOR
2890: .rs
2891: .sp
2892: .nf
2893: Philip Hazel
2894: University Computing Service
2895: Cambridge CB2 3QH, England.
2896: .fi
2897: .
2898: .
2899: .SH REVISION
2900: .rs
2901: .sp
2902: .nf
2903: Last updated: 12 November 2013
2904: Copyright (c) 1997-2013 University of Cambridge.
2905: .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