/[pcre]/code/trunk/pcre.3
ViewVC logotype

Diff of /code/trunk/pcre.3

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

revision 9 by nigel, Sat Feb 24 21:38:13 2007 UTC revision 27 by nigel, Sat Feb 24 21:38:49 2007 UTC
# Line 8  pcre - Perl-compatible regular expressio Line 8  pcre - Perl-compatible regular expressio
8  .br  .br
9  .B pcre *pcre_compile(const char *\fIpattern\fR, int \fIoptions\fR,  .B pcre *pcre_compile(const char *\fIpattern\fR, int \fIoptions\fR,
10  .ti +5n  .ti +5n
11  .B char **\fIerrptr\fR, int *\fIerroffset\fR);  .B const char **\fIerrptr\fR, int *\fIerroffset\fR,
12    .ti +5n
13    .B const unsigned char *\fItableptr\fR);
14    .PP
15    .br
16    .B const unsigned char *pcre_maketables(void);
17  .PP  .PP
18  .br  .br
19  .B pcre_extra *pcre_study(const pcre *\fIcode\fR, int \fIoptions\fR,  .B pcre_extra *pcre_study(const pcre *\fIcode\fR, int \fIoptions\fR,
20  .ti +5n  .ti +5n
21  .B char **\fIerrptr\fR);  .B const char **\fIerrptr\fR);
22  .PP  .PP
23  .br  .br
24  .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"  .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"
# Line 34  pcre - Perl-compatible regular expressio Line 39  pcre - Perl-compatible regular expressio
39  .PP  .PP
40  .br  .br
41  .B void (*pcre_free)(void *);  .B void (*pcre_free)(void *);
 .PP  
 .br  
 .B unsigned char *pcre_cbits[128];  
 .PP  
 .br  
 .B unsigned char *pcre_ctypes[256];  
 .PP  
 .br  
 .B unsigned char *pcre_fcc[256];  
 .PP  
 .br  
 .B unsigned char *pcre_lcc[256];  
42    
43    
44    
45  .SH DESCRIPTION  .SH DESCRIPTION
46  The PCRE library is a set of functions that implement regular expression  The PCRE library is a set of functions that implement regular expression
47  pattern matching using the same syntax and semantics as Perl 5, with just a few  pattern matching using the same syntax and semantics as Perl 5, with just a few
48  differences (see below). The current implementation corresponds to Perl 5.004.  differences (see below). The current implementation corresponds to Perl 5.005.
49    
50  PCRE has its own native API, which is described in this man page. There is also  PCRE has its own native API, which is described in this man page. There is also
51  a set of wrapper functions that correspond to the POSIX API. See  a set of wrapper functions that correspond to the POSIX API. See
# Line 60  a set of wrapper functions that correspo Line 53  a set of wrapper functions that correspo
53    
54  The three functions \fBpcre_compile()\fR, \fBpcre_study()\fR, and  The three functions \fBpcre_compile()\fR, \fBpcre_study()\fR, and
55  \fBpcre_exec()\fR are used for compiling and matching regular expressions. The  \fBpcre_exec()\fR are used for compiling and matching regular expressions. The
56  function \fBpcre_info()\fR is used to find out information about a compiled  function \fBpcre_maketables()\fR is used (optionally) to build a set of
57    character tables in the current locale for passing to \fBpcre_compile()\fR.
58    
59    The function \fBpcre_info()\fR is used to find out information about a compiled
60  pattern, while the function \fBpcre_version()\fR returns a pointer to a string  pattern, while the function \fBpcre_version()\fR returns a pointer to a string
61  containing the version of PCRE and its date of release.  containing the version of PCRE and its date of release.
62    
# Line 70  respectively. PCRE calls the memory mana Line 66  respectively. PCRE calls the memory mana
66  so a calling program can replace them if it wishes to intercept the calls. This  so a calling program can replace them if it wishes to intercept the calls. This
67  should be done before calling any PCRE functions.  should be done before calling any PCRE functions.
68    
 The other global variables are character tables. They are initialized when PCRE  
 is compiled, from source that is generated by reference to the C character type  
 functions, but which the maintainer of PCRE is free to modify. In principle  
 they could also be modified at runtime. See PCRE's README file for more  
 details.  
   
69    
70  .SH MULTI-THREADING  .SH MULTI-THREADING
71  The PCRE functions can be used in multi-threading applications, with the  The PCRE functions can be used in multi-threading applications, with the
72  proviso that the character tables and the memory management functions pointed  proviso that the memory management functions pointed to by \fBpcre_malloc\fR
73  to by \fBpcre_malloc\fR and \fBpcre_free\fR will be shared by all threads.  and \fBpcre_free\fR are shared by all threads.
74    
75  The compiled form of a regular expression is not altered during matching, so  The compiled form of a regular expression is not altered during matching, so
76  the same compiled pattern can safely be used by several threads at once.  the same compiled pattern can safely be used by several threads at once.
# Line 89  the same compiled pattern can safely be Line 79  the same compiled pattern can safely be
79  .SH COMPILING A PATTERN  .SH COMPILING A PATTERN
80  The function \fBpcre_compile()\fR is called to compile a pattern into an  The function \fBpcre_compile()\fR is called to compile a pattern into an
81  internal form. The pattern is a C string terminated by a binary zero, and  internal form. The pattern is a C string terminated by a binary zero, and
82  is passed in the argument \fIpattern\fR. A pointer to the compiled code block  is passed in the argument \fIpattern\fR. A pointer to a single block of memory
83  is returned. The \fBpcre\fR type is defined for this for convenience, but in  that is obtained via \fBpcre_malloc\fR is returned. This contains the
84  fact \fBpcre\fR is just a typedef for \fBvoid\fR, since the contents of the  compiled code and related data. The \fBpcre\fR type is defined for this for
85  block are not defined.  convenience, but in fact \fBpcre\fR is just a typedef for \fBvoid\fR, since the
86    contents of the block are not externally defined. It is up to the caller to
87    free the memory when it is no longer required.
88  .PP  .PP
89  The size of a compiled pattern is roughly proportional to the length of the  The size of a compiled pattern is roughly proportional to the length of the
90  pattern string, except that each character class (other than those containing  pattern string, except that each character class (other than those containing
# Line 101  quantifiers with a minimum greater than Line 93  quantifiers with a minimum greater than
93  relevant portions of the compiled pattern to be replicated.  relevant portions of the compiled pattern to be replicated.
94  .PP  .PP
95  The \fIoptions\fR argument contains independent bits that affect the  The \fIoptions\fR argument contains independent bits that affect the
96  compilation. It should be zero if no options are required. Those options that  compilation. It should be zero if no options are required. Some of the options,
97  are compabible with Perl can also be set at compile time from within the  in particular, those that are compatible with Perl, can also be set and unset
98  pattern (see the detailed description of regular expressions below) and all  from within the pattern (see the detailed description of regular expressions
99  options except PCRE_EXTENDED and PCRE_EXTRA can be set at the time of matching.  below). For these options, the contents of the \fIoptions\fR argument specifies
100    their initial settings at the start of compilation and execution. The
101    PCRE_ANCHORED option can be set at the time of matching as well as at compile
102    time.
103  .PP  .PP
104  If \fIerrptr\fR is NULL, \fBpcre_compile()\fR returns NULL immediately.  If \fIerrptr\fR is NULL, \fBpcre_compile()\fR returns NULL immediately.
105  Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fR returns  Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fR returns
106  NULL, and sets the variable pointed to by \fIerrptr\fR to point to a textual  NULL, and sets the variable pointed to by \fIerrptr\fR to point to a textual
107  error message.  error message. The offset from the start of the pattern to the character where
108    the error was discovered is placed in the variable pointed to by
109  The offset from the start of the pattern to the character where the error was  \fIerroffset\fR, which must not be NULL. If it is, an immediate error is given.
110  discovered is placed in the variable pointed to by \fIerroffset\fR, which must  .PP
111  not be NULL. If it is, an immediate error is given.  If the final argument, \fItableptr\fR, is NULL, PCRE uses a default set of
112    character tables which are built when it is compiled, using the default C
113    locale. Otherwise, \fItableptr\fR must be the result of a call to
114    \fBpcre_maketables()\fR. See the section on locale support below.
115  .PP  .PP
116  The following option bits are defined in the header file:  The following option bits are defined in the header file:
117    
# Line 127  constructs in the pattern itself, which Line 125  constructs in the pattern itself, which
125    PCRE_CASELESS    PCRE_CASELESS
126    
127  If this bit is set, letters in the pattern match both upper and lower case  If this bit is set, letters in the pattern match both upper and lower case
128  letters in any subject string. It is equivalent to Perl's /i option.  letters. It is equivalent to Perl's /i option.
129    
130    PCRE_DOLLAR_ENDONLY    PCRE_DOLLAR_ENDONLY
131    
132  If this bit is set, a dollar metacharacter in the pattern matches only at the  If this bit is set, a dollar metacharacter in the pattern matches only at the
133  end of the subject string. By default, it also matches immediately before the  end of the subject string. Without this option, a dollar also matches
134  final character if it is a newline (but not before any other newlines). The  immediately before the final character if it is a newline (but not before any
135  PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set. There is no  other newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is
136  equivalent to this option in Perl.  set. There is no equivalent to this option in Perl.
137    
138    PCRE_DOTALL    PCRE_DOTALL
139    
140  If this bit is set, a dot metacharater in the pattern matches all characters,  If this bit is set, a dot metacharater in the pattern matches all characters,
141  including newlines. By default, newlines are excluded. This option is  including newlines. Without it, newlines are excluded. This option is
142  equivalent to Perl's /s option. A negative class such as [^a] always matches a  equivalent to Perl's /s option. A negative class such as [^a] always matches a
143  newline character, independent of the setting of this option.  newline character, independent of the setting of this option.
144    
145    PCRE_EXTENDED    PCRE_EXTENDED
146    
147  If this bit is set, whitespace characters in the pattern are totally ignored  If this bit is set, whitespace data characters in the pattern are totally
148  except when escaped or inside a character class, and characters between an  ignored except when escaped or inside a character class, and characters between
149  unescaped # outside a character class and the next newline character,  an unescaped # outside a character class and the next newline character,
150  inclusive, are also ignored. This is equivalent to Perl's /x option, and makes  inclusive, are also ignored. This is equivalent to Perl's /x option, and makes
151  it possible to include comments inside complicated patterns.  it possible to include comments inside complicated patterns. Note, however,
152    that this applies only to data characters. Whitespace characters may never
153    appear within special character sequences in a pattern, for example within the
154    sequence (?( which introduces a conditional subpattern.
155    
156      PCRE_EXTRA
157    
158    This option turns on additional functionality of PCRE that is incompatible with
159    Perl. Any backslash in a pattern that is followed by a letter that has no
160    special meaning causes an error, thus reserving these combinations for future
161    expansion. By default, as in Perl, a backslash followed by a letter with no
162    special meaning is treated as a literal. There are at present no other features
163    controlled by this option.
164    
165    PCRE_MULTILINE    PCRE_MULTILINE
166    
# Line 158  By default, PCRE treats the subject stri Line 168  By default, PCRE treats the subject stri
168  characters (even if it actually contains several newlines). The "start of line"  characters (even if it actually contains several newlines). The "start of line"
169  metacharacter (^) matches only at the start of the string, while the "end of  metacharacter (^) matches only at the start of the string, while the "end of
170  line" metacharacter ($) matches only at the end of the string, or before a  line" metacharacter ($) matches only at the end of the string, or before a
171  terminating newline. This is the same as Perl.  terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
172    Perl.
173    
174  When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs  When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
175  match immediately following or immediately before any newline in the subject  match immediately following or immediately before any newline in the subject
# Line 167  to Perl's /m option. If there are no "\\ Line 178  to Perl's /m option. If there are no "\\
178  no occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no  no occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no
179  effect.  effect.
180    
181    PCRE_EXTRA    PCRE_UNGREEDY
   
 This option turns on additional functionality of PCRE that is incompatible with  
 Perl. Any backslash in a pattern that is followed by a letter that has no  
 special meaning causes an error, thus reserving these combinations for future  
 expansion. By default, as in Perl, a backslash followed by a letter with no  
 special meaning is treated as a literal. There are two extra features currently  
 provided, and both are in some sense experimental additions that are useful for  
 influencing the progress of a match.  
   
   (1) The sequence \\X inserts a Prolog-like "cut" into the expression.  
   
   (2) Once a subpattern enclosed in (?>subpat) brackets has matched,  
       backtracking never goes back into the pattern.  
   
 See below for further details of both of these.  
182    
183    This option inverts the "greediness" of the quantifiers so that they are not
184    greedy by default, but become greedy if followed by "?". It is not compatible
185    with Perl. It can also be set by a (?U) option setting within the pattern.
186    
187    
188  .SH STUDYING A PATTERN  .SH STUDYING A PATTERN
# Line 195  typedef) containing additional informati Line 194  typedef) containing additional informati
194  passed to \fBpcre_exec()\fR. If no additional information is available, NULL  passed to \fBpcre_exec()\fR. If no additional information is available, NULL
195  is returned.  is returned.
196    
197  The second argument contains option bits. The only one currently supported is  The second argument contains option bits. At present, no options are defined
198  PCRE_CASELESS. It forces the studying to be done in a caseless manner, even if  for \fBpcre_study()\fR, and this argument should always be zero.
 the original pattern was compiled without PCRE_CASELESS. When the result of  
 \fBpcre_study()\fR is passed to \fBpcre_exec()\fR, it is used only if its  
 caseless state is the same as that of the matching process. A pattern that is  
 compiled without PCRE_CASELESS can be studied with and without PCRE_CASELESS,  
 and the appropriate data passed to \fBpcre_exec()\fR with and without the  
 PCRE_CASELESS flag.  
199    
200  The third argument for \fBpcre_study()\fR is a pointer to an error message. If  The third argument for \fBpcre_study()\fR is a pointer to an error message. If
201  studying succeeds (even if no data is returned), the variable it points to is  studying succeeds (even if no data is returned), the variable it points to is
# Line 213  not have a single fixed starting charact Line 206  not have a single fixed starting charact
206  characters is created.  characters is created.
207    
208    
209    .SH LOCALE SUPPORT
210    PCRE handles caseless matching, and determines whether characters are letters,
211    digits, or whatever, by reference to a set of tables. The library contains a
212    default set of tables which is created in the default C locale when PCRE is
213    compiled. This is used when the final argument of \fBpcre_compile()\fR is NULL,
214    and is sufficient for many applications.
215    
216    An alternative set of tables can, however, be supplied. Such tables are built
217    by calling the \fBpcre_maketables()\fR function, which has no arguments, in the
218    relevant locale. The result can then be passed to \fBpcre_compile()\ as often
219    as necessary. For example, to build and use tables that are appropriate for the
220    French locale (where accented characters with codes greater than 128 are
221    treated as letters), the following code could be used:
222    
223      setlocale(LC_CTYPE, "fr");
224      tables = pcre_maketables();
225      re = pcre_compile(..., tables);
226    
227    The tables are built in memory that is obtained via \fBpcre_malloc\fR. The
228    pointer that is passed to \fBpcre_compile\fR is saved with the compiled
229    pattern, and the same tables are used via this pointer by \fBpcre_study()\fR
230    and \fBpcre_match()\fR. Thus for any single pattern, compilation, studying and
231    matching all happen in the same locale, but different patterns can be compiled
232    in different locales. It is the caller's responsibility to ensure that the
233    memory containing the tables remains available for as long as it is needed.
234    
235    
236  .SH MATCHING A PATTERN  .SH MATCHING A PATTERN
237  The function \fBpcre_exec()\fR is called to match a subject string against a  The function \fBpcre_exec()\fR is called to match a subject string against a
238  pre-compiled pattern, which is passed in the \fIcode\fR argument. If the  pre-compiled pattern, which is passed in the \fIcode\fR argument. If the
# Line 222  pattern has been studied, the result of Line 242  pattern has been studied, the result of
242  The subject string is passed as a pointer in \fIsubject\fR and a length in  The subject string is passed as a pointer in \fIsubject\fR and a length in
243  \fIlength\fR. Unlike the pattern string, it may contain binary zero characters.  \fIlength\fR. Unlike the pattern string, it may contain binary zero characters.
244    
245  The options PCRE_ANCHORED, PCRE_CASELESS, PCRE_DOLLAR_ENDONLY, PCRE_DOTALL, and  The PCRE_ANCHORED option can be passed in the \fIoptions\fR argument, whose
246  PCRE_MULTILINE can be passed in the \fIoptions\fR argument, whose unused bits  unused bits must be zero. However, if a pattern was compiled with
247  must be zero. However, if a pattern is compiled with any of these options, they  PCRE_ANCHORED, or turned out to be anchored by virtue of its contents, it
248  cannot be unset when it is obeyed.  cannot be made unachored at matching time.
249    
250  There are also two further options that can be set only at matching time:  There are also two further options that can be set only at matching time:
251    
# Line 233  There are also two further options that Line 253  There are also two further options that
253    
254  The first character of the string is not the beginning of a line, so the  The first character of the string is not the beginning of a line, so the
255  circumflex metacharacter should not match before it. Setting this without  circumflex metacharacter should not match before it. Setting this without
256  PCRE_MULTILINE (at either compile or match time) causes circumflex never to  PCRE_MULTILINE (at compile time) causes circumflex never to match.
 match.  
257    
258    PCRE_NOTEOL    PCRE_NOTEOL
259    
260  The end of the string is not the end of a line, so the dollar metacharacter  The end of the string is not the end of a line, so the dollar metacharacter
261  should not match it. Setting this without PCRE_MULTILINE (at either compile or  should not match it nor (except in multiline mode) a newline immediately before
262  match time) causes dollar never to match.  it. Setting this without PCRE_MULTILINE (at compile time) causes dollar never
263    to match.
264    
265  In general, a pattern matches a certain portion of the subject, and in  In general, a pattern matches a certain portion of the subject, and in
266  addition, further substrings from the subject may be picked out by parts of the  addition, further substrings from the subject may be picked out by parts of the
# Line 251  kinds of parenthesized subpattern that d Line 271  kinds of parenthesized subpattern that d
271    
272  Captured substrings are returned to the caller via a vector of integer offsets  Captured substrings are returned to the caller via a vector of integer offsets
273  whose address is passed in \fIovector\fR. The number of elements in the vector  whose address is passed in \fIovector\fR. The number of elements in the vector
274  is passed in \fIovecsize\fR. This should always be an even number, because the  is passed in \fIovecsize\fR. The first two-thirds of the vector is used to pass
275  elements are used in pairs. If an odd number is passed, it is rounded down.  back captured substrings, each substring using a pair of integers. The
276    remaining third of the vector is used as workspace by \fBpcre_exec()\fR while
277  The first element of a pair is set to the offset of the first character in a  matching capturing subpatterns, and is not available for passing back
278  substring, and the second is set to the offset of the first character after the  information. The length passed in \fIovecsize\fR should always be a multiple of
279  end of a substring. The first pair, \fIovector[0]\fR and \fIovector[1]\fR,  three. If it is not, it is rounded down.
280  identify the portion of the subject string matched by the entire pattern. The  
281  next pair is used for the first capturing subpattern, and so on. The value  When a match has been successful, information about captured substrings is
282  returned by \fBpcre_exec()\fR is the number of pairs that have been set. If  returned in pairs of integers, starting at the beginning of \fIovector\fR, and
283  there are no capturing subpatterns, the return value from a successful match  continuing up to two-thirds of its length at the most. The first element of a
284  is 1, indicating that just the first pair of offsets has been set.  pair is set to the offset of the first character in a substring, and the second
285    is set to the offset of the first character after the end of a substring. The
286    first pair, \fIovector[0]\fR and \fIovector[1]\fR, identify the portion of the
287    subject string matched by the entire pattern. The next pair is used for the
288    first capturing subpattern, and so on. The value returned by \fBpcre_exec()\fR
289    is the number of pairs that have been set. If there are no capturing
290    subpatterns, the return value from a successful match is 1, indicating that
291    just the first pair of offsets has been set.
292    
293  It is possible for an capturing subpattern number \fIn+1\fR to match some  It is possible for an capturing subpattern number \fIn+1\fR to match some
294  part of the subject when subpattern \fIn\fR has not been used at all. For  part of the subject when subpattern \fIn\fR has not been used at all. For
295  example, if the string "abc" is matched against the pattern "(a|(z))(bc)",  example, if the string "abc" is matched against the pattern (a|(z))(bc)
296  subpatterns 1 and 3 are matched, but 2 is not. When this happens, both offset  subpatterns 1 and 3 are matched, but 2 is not. When this happens, both offset
297  values corresponding to the unused subpattern are set to -1.  values corresponding to the unused subpattern are set to -1.
298    
# Line 273  If a capturing subpattern is matched rep Line 300  If a capturing subpattern is matched rep
300  string that it matched that gets returned.  string that it matched that gets returned.
301    
302  If the vector is too small to hold all the captured substrings, it is used as  If the vector is too small to hold all the captured substrings, it is used as
303  far as possible, and the function returns a value of zero. In particular, if  far as possible (up to two-thirds of its length), and the function returns a
304  the substring offsets are not of interest, \fBpcre_exec()\fR may be called with  value of zero. In particular, if the substring offsets are not of interest,
305  \fIovector\fR passed as NULL and \fIovecsize\fR as zero. However, if the  \fBpcre_exec()\fR may be called with \fIovector\fR passed as NULL and
306  pattern contains back references and the \fIovector\fR isn't big enough to  \fIovecsize\fR as zero. However, if the pattern contains back references and
307  remember the related substrings, PCRE has to get additional memory for use  the \fIovector\fR isn't big enough to remember the related substrings, PCRE has
308  during matching. Thus it is usually advisable to supply an \fIovector\fR.  to get additional memory for use during matching. Thus it is usually advisable
309    to supply an \fIovector\fR.
310    
311  Note that \fBpcre_info()\fR can be used to find out how many capturing  Note that \fBpcre_info()\fR can be used to find out how many capturing
312  subpatterns there are in a compiled pattern.  subpatterns there are in a compiled pattern. The smallest size for
313    \fIovector\fR that will allow for \fIn\fR captured substrings in addition to
314    the offsets of the substring matched by the whole pattern is (\fIn\fR+1)*3.
315    
316  If \fBpcre_exec()\fR fails, it returns a negative number. The following are  If \fBpcre_exec()\fR fails, it returns a negative number. The following are
317  defined in the header file:  defined in the header file:
# Line 290  defined in the header file: Line 320  defined in the header file:
320    
321  The subject string did not match the pattern.  The subject string did not match the pattern.
322    
323    PCRE_ERROR_BADREF         (-2)    PCRE_ERROR_NULL           (-2)
   
 There was a back-reference in the pattern to a capturing subpattern that had  
 not previously been set.  
   
   PCRE_ERROR_NULL           (-3)  
324    
325  Either \fIcode\fR or \fIsubject\fR was passed as NULL, or \fIovector\fR was  Either \fIcode\fR or \fIsubject\fR was passed as NULL, or \fIovector\fR was
326  NULL and \fIovecsize\fR was not zero.  NULL and \fIovecsize\fR was not zero.
327    
328    PCRE_ERROR_BADOPTION      (-4)    PCRE_ERROR_BADOPTION      (-3)
329    
330  An unrecognized bit was set in the \fIoptions\fR argument.  An unrecognized bit was set in the \fIoptions\fR argument.
331    
332    PCRE_ERROR_BADMAGIC       (-5)    PCRE_ERROR_BADMAGIC       (-4)
333    
334  PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch  PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
335  the case when it is passed a junk pointer. This is the error it gives when the  the case when it is passed a junk pointer. This is the error it gives when the
336  magic number isn't present.  magic number isn't present.
337    
338    PCRE_ERROR_UNKNOWN_NODE   (-6)    PCRE_ERROR_UNKNOWN_NODE   (-5)
339    
340  While running the pattern match, an unknown item was encountered in the  While running the pattern match, an unknown item was encountered in the
341  compiled pattern. This error could be caused by a bug in PCRE or by overwriting  compiled pattern. This error could be caused by a bug in PCRE or by overwriting
342  of the compiled pattern.  of the compiled pattern.
343    
344    PCRE_ERROR_NOMEMORY       (-7)    PCRE_ERROR_NOMEMORY       (-6)
345    
346  If a pattern contains back references, but the \fIovector\fR that is passed to  If a pattern contains back references, but the \fIovector\fR that is passed to
347  \fBpcre_exec()\fR is not big enough to remember the referenced substrings, PCRE  \fBpcre_exec()\fR is not big enough to remember the referenced substrings, PCRE
# Line 353  The maximum length of a compiled pattern Line 378  The maximum length of a compiled pattern
378  All values in repeating quantifiers must be less than 65536.  All values in repeating quantifiers must be less than 65536.
379  The maximum number of capturing subpatterns is 99.  The maximum number of capturing subpatterns is 99.
380  The maximum number of all parenthesized subpatterns, including capturing  The maximum number of all parenthesized subpatterns, including capturing
381  subpatterns and assertions, is 200.  subpatterns, assertions, and other types of subpattern, is 200.
382    
383  The maximum length of a subject string is the largest positive number that an  The maximum length of a subject string is the largest positive number that an
384  integer variable can hold. However, PCRE uses recursion to handle subpatterns  integer variable can hold. However, PCRE uses recursion to handle subpatterns
# Line 362  the size of a subject string that can be Line 387  the size of a subject string that can be
387    
388    
389  .SH DIFFERENCES FROM PERL  .SH DIFFERENCES FROM PERL
390  The differences described here are with respect to Perl 5.004.  The differences described here are with respect to Perl 5.005.
391    
392  1. By default, a whitespace character is any character that the C library  1. By default, a whitespace character is any character that the C library
393  function \fBisspace()\fR recognizes, though it is possible to compile PCRE with  function \fBisspace()\fR recognizes, though it is possible to compile PCRE with
# Line 371  formfeed, newline, carriage return, hori Line 396  formfeed, newline, carriage return, hori
396  no longer includes vertical tab in its set of whitespace characters. The \\v  no longer includes vertical tab in its set of whitespace characters. The \\v
397  escape that was in the Perl documentation for a long time was never in fact  escape that was in the Perl documentation for a long time was never in fact
398  recognized. However, the character itself was treated as whitespace at least  recognized. However, the character itself was treated as whitespace at least
399  up to 5.002. In 5.004 it does not match \\s.  up to 5.002. In 5.004 and 5.005 it does not match \\s.
400    
401  2. PCRE does not allow repeat quantifiers on lookahead assertions. Perl permits  2. PCRE does not allow repeat quantifiers on lookahead assertions. Perl permits
402  them, but they do not mean what you might think. For example, "(?!a){3}" does  them, but they do not mean what you might think. For example, (?!a){3} does
403  not assert that the next three characters are not "a". It just asserts that the  not assert that the next three characters are not "a". It just asserts that the
404  next character is not "a" three times.  next character is not "a" three times.
405    
# Line 396  are not part of its pattern matching eng Line 421  are not part of its pattern matching eng
421  6. The Perl \\G assertion is not supported as it is not relevant to single  6. The Perl \\G assertion is not supported as it is not relevant to single
422  pattern matches.  pattern matches.
423    
424  7. If a backreference can never be matched, PCRE diagnoses an error. In a case  7. Fairly obviously, PCRE does not support the (?{code}) construction.
 like  
425    
426    /(123)\\2/  8. There are at the time of writing some oddities in Perl 5.005_02 concerned
427    with the settings of captured strings when part of a pattern is repeated. For
428    example, matching "aba" against the pattern /^(a(b)?)+$/ sets $2 to the value
429    "b", but matching "aabbaa" against /^(aa(bb)?)+$/ leaves $2 unset. However, if
430    the pattern is changed to /^(aa(b(b))?)+$/ then $2 (and $3) get set.
431    
432  the error occurs at compile time. Perl gives no compile time error; version  In Perl 5.004 $2 is set in both cases, and that is also true of PCRE. If in the
433  5.004 either always fails to match, or gives a segmentation fault at runtime.  future Perl changes to a consistent state that is different, PCRE may change to
434  In more complicated cases such as  follow.
435    
436    /(1)(2)(3)(4)(5)(6)(7)(8)(9)(10\\10)/  9. Another as yet unresolved discrepancy is that in Perl 5.005_02 the pattern
437    /^(a)?(?(1)a|b)+$/ matches the string "a", whereas in PCRE it does not.
438    However, in both Perl and PCRE /^(a)?a/ matched against "a" leaves $1 unset.
439    
440  PCRE returns PCRE_ERROR_BADREF at run time. Perl always fails to match.  10. PCRE provides some extensions to the Perl regular expression facilities:
441    
442  8. PCRE provides some extensions to the Perl regular expression facilities:  (a) Although lookbehind assertions must match fixed length strings, each
443    alternative branch of a lookbehind assertion can match a different length of
444    string. Perl 5.005 requires them all to have the same length.
445    
446  (a) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the $ meta-  (b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the $ meta-
447  character matches only at the very end of the string.  character matches only at the very end of the string.
448    
449  (b) If PCRE_EXTRA is set, the \\X assertion (a Prolog-like "cut") is  (c) If PCRE_EXTRA is set, a backslash followed by a letter with no special
450  recognized, and a backslash followed by a letter with no special meaning is  meaning is faulted.
451  faulted. There is also a new kind of parenthesized subpattern starting with (?>  
452  which has a block on backtracking into it once it has matched.  (d) If PCRE_UNGREEDY is set, the greediness of the repetition quantifiers is
453    inverted, that is, by default they are not greedy, but if followed by a
454    question mark they are.
455    
456    
457  .SH REGULAR EXPRESSION DETAILS  .SH REGULAR EXPRESSION DETAILS
# Line 484  non-alphameric with "\\" to specify that Line 518  non-alphameric with "\\" to specify that
518  if you want to match a backslash, you write "\\\\".  if you want to match a backslash, you write "\\\\".
519    
520  If a pattern is compiled with the PCRE_EXTENDED option, whitespace in the  If a pattern is compiled with the PCRE_EXTENDED option, whitespace in the
521  pattern and characters between a "#" outside a character class and the next  pattern (other than in a character class) and characters between a "#" outside
522  newline character are ignored. An escaping backslash can be used to include a  a character class and the next newline character are ignored. An escaping
523  whitespace or "#" character as part of the pattern.  backslash can be used to include a whitespace or "#" character as part of the
524    pattern.
525    
526  A second use of backslash provides a way of encoding non-printing characters  A second use of backslash provides a way of encoding non-printing characters
527  in patterns in a visible manner. There is no restriction on the appearance of  in patterns in a visible manner. There is no restriction on the appearance of
# Line 503  represents: Line 538  represents:
538    \\r     carriage return (hex 0D)    \\r     carriage return (hex 0D)
539    \\t     tab (hex 09)    \\t     tab (hex 09)
540    \\xhh   character with hex code hh    \\xhh   character with hex code hh
541    \\ddd   character with octal code ddd or backreference    \\ddd   character with octal code ddd, or backreference
542    
543  The precise effect of "\\cx" is as follows: if "x" is a lower case letter, it  The precise effect of "\\cx" is as follows: if "x" is a lower case letter, it
544  is converted to upper case. Then bit 6 of the character (hex 40) is inverted.  is converted to upper case. Then bit 6 of the character (hex 40) is inverted.
# Line 516  lower case). Line 551  lower case).
551  After "\\0" up to two further octal digits are read. In both cases, if there  After "\\0" up to two further octal digits are read. In both cases, if there
552  are fewer than two digits, just those that are present are used. Thus the  are fewer than two digits, just those that are present are used. Thus the
553  sequence "\\0\\x\\07" specifies two binary zeros followed by a BEL character.  sequence "\\0\\x\\07" specifies two binary zeros followed by a BEL character.
554  Make sure you supply two digits if the character that follows could otherwise  Make sure you supply two digits after the initial zero if the character that
555  be taken as another digit.  follows is itself an octal digit.
556    
557  The handling of a backslash followed by a digit other than 0 is complicated.  The handling of a backslash followed by a digit other than 0 is complicated.
558  Outside a character class, PCRE reads it and any following digits as a decimal  Outside a character class, PCRE reads it and any following digits as a decimal
# Line 567  Each pair of escape sequences partitions Line 602  Each pair of escape sequences partitions
602  two disjoint sets. Any given character matches one, and only one, of each pair.  two disjoint sets. Any given character matches one, and only one, of each pair.
603    
604  A "word" character is any letter or digit or the underscore character, that is,  A "word" character is any letter or digit or the underscore character, that is,
605  any character which can be part of a Perl "word". These character type  any character which can be part of a Perl "word". The definition of letters and
606  sequences can appear both inside and outside character classes. They each match  digits is controlled by PCRE's character tables, and may vary if locale-
607  one character of the appropriate type. If the current matching point is at the  specific matching is taking place (see "Locale support" above). For example, in
608  end of the subject string, all of them fail, since there is no character to  the "fr" (French) locale, some character codes greater than 128 are used for
609  match.  accented letters, and these are matched by \\w.
610    
611  The fourth use of backslash is for certain assertions. An assertion specifies a  These character type sequences can appear both inside and outside character
612  condition that has to be met at a particular point in a match, without  classes. They each match one character of the appropriate type. If the current
613  consuming any characters from the subject string. The backslashed assertions  matching point is at the end of the subject string, all of them fail, since
614  are  there is no character to match.
615    
616    The fourth use of backslash is for certain simple assertions. An assertion
617    specifies a condition that has to be met at a particular point in a match,
618    without consuming any characters from the subject string. The use of
619    subpatterns for more complicated assertions is described below. The backslashed
620    assertions are
621    
622    \\b     word boundary    \\b     word boundary
623    \\B     not a word boundary    \\B     not a word boundary
624    \\A     start of subject (independent of multiline mode)    \\A     start of subject (independent of multiline mode)
625    \\Z     end of subject (independent of multiline mode)    \\Z     end of subject or newline at end (independent of multiline mode)
626      \\z     end of subject (independent of multiline mode)
627    
628  Assertions may not appear in character classes (but note that "\\b" has a  These assertions may not appear in character classes (but note that "\\b" has a
629  different meaning, namely the backspace character, inside a character class).  different meaning, namely the backspace character, inside a character class).
630    
631  A word boundary is a position in the subject string where the current character  A word boundary is a position in the subject string where the current character
632  and the previous character do not both match "\\w" or "\\W" (i.e. one matches  and the previous character do not both match \\w or \\W (i.e. one matches
633  "\\w" and the other matches "\\W"), or the start or end of the string if the  \\w and the other matches \\W), or the start or end of the string if the
634  first or last character matches "\\w", respectively. More complicated  first or last character matches \\w, respectively.
635  assertions are also supported (see below).  
636    The \\A, \\Z, and \\z assertions differ from the traditional circumflex and
637  The "\\A" and "\\Z" assertions differ from the traditional "^" and "$"  dollar (described below) in that they only ever match at the very start and end
638  (described below) in that they only ever match at the very start and end of the  of the subject string, whatever options are set. They are not affected by the
639  subject string, respectively, whatever options are set.  PCRE_NOTBOL or PCRE_NOTEOL options. The difference between \\Z and \\z is that
640    \\Z matches before a newline that is the last character of the string as well
641  When the PCRE_EXTRA flag is set on a call to \fBpcre_compile()\fR, the  as at the end of the string, whereas \\z matches only at the end.
 additional assertion \\X, which has no equivalent in Perl, is recognized.  
 This operates like the "cut" operation in Prolog: it prevents the matching  
 operation from backtracking past it. For example, if the expression  
   
   .*/foo  
   
 is matched against the string "/this/string/is/not" then after the greedy .*  
 has swallowed the whole string, PCRE keeps backtracking all the way to the  
 beginning before failing. If, on the other hand, the expression is  
   
   .*/\\Xfoo  
   
 then once it has discovered that "/not" is not "/foo", backtracking ceases, and  
 the match fails. See also the section on "once-only" subpatterns below.  
   
642    
643    
644  .SH CIRCUMFLEX AND DOLLAR  .SH CIRCUMFLEX AND DOLLAR
645  Outside a character class, the circumflex character is an assertion which is  Outside a character class, in the default matching mode, the circumflex
646  true only if the current matching point is at the start of the subject string,  character is an assertion which is true only if the current matching point is
647  in the default matching mode. Inside a character class, circumflex has an  at the start of the subject string. Inside a character class, circumflex has an
648  entirely different meaning (see below).  entirely different meaning (see below).
649    
650  Circumflex need not be the first character of the pattern if a number of  Circumflex need not be the first character of the pattern if a number of
# Line 637  Dollar has no special meaning in a chara Line 664  Dollar has no special meaning in a chara
664    
665  The meaning of dollar can be changed so that it matches only at the very end of  The meaning of dollar can be changed so that it matches only at the very end of
666  the string, by setting the PCRE_DOLLAR_ENDONLY option at compile or matching  the string, by setting the PCRE_DOLLAR_ENDONLY option at compile or matching
667  time.  time. This does not affect the \\Z assertion.
668    
669  The meanings of the circumflex and dollar characters are changed if the  The meanings of the circumflex and dollar characters are changed if the
670  PCRE_MULTILINE option is set at compile or matching time. When this is the  PCRE_MULTILINE option is set. When this is the case, they match immediately
671  case, they match immediately after and immediately before an internal "\\n"  after and immediately before an internal "\\n" character, respectively, in
672  character, respectively, in addition to matching at the start and end of the  addition to matching at the start and end of the subject string. For example,
673  subject string. For example, the pattern /^abc$/ matches the subject string  the pattern /^abc$/ matches the subject string "def\\nabc" in multiline mode,
674  "def\\nabc" in multiline mode, but not otherwise. Consequently, patterns that  but not otherwise. Consequently, patterns that are anchored in single line mode
675  are anchored in single line mode because all branches start with "^" are not  because all branches start with "^" are not anchored in multiline mode. The
676  anchored in multiline mode. The PCRE_DOLLAR_ENDONLY option is ignored if  PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
677  PCRE_MULTILINE is set.  
678    Note that the sequences \\A, \\Z, and \\z can be used to match the start and
679  Note that the sequences "\\A" and "\\Z" can be used to match the start and end  end of the subject in both modes, and if all branches of a pattern start with
680  of the subject in both modes, and if all branches of a pattern start with "\\A"  \\A is it always anchored, whether PCRE_MULTILINE is set or not.
 is it always anchored.  
681    
682    
683  .SH FULL STOP (PERIOD, DOT)  .SH FULL STOP (PERIOD, DOT)
# Line 668  An opening square bracket introduces a c Line 694  An opening square bracket introduces a c
694  square bracket. A closing square bracket on its own is not special. If a  square bracket. A closing square bracket on its own is not special. If a
695  closing square bracket is required as a member of the class, it should be the  closing square bracket is required as a member of the class, it should be the
696  first data character in the class (after an initial circumflex, if present) or  first data character in the class (after an initial circumflex, if present) or
697  escaped with \\.  escaped with a backslash.
698    
699  A character class matches a single character in the subject; the character must  A character class matches a single character in the subject; the character must
700  be in the set of characters defined by the class, unless the first character in  be in the set of characters defined by the class, unless the first character in
701  the class is a circumflex, in which case the subject character must not be in  the class is a circumflex, in which case the subject character must not be in
702  the set defined by the class. If a circumflex is actually required as a member  the set defined by the class. If a circumflex is actually required as a member
703  of the class, ensure it is not the first character, or escape it with \\.  of the class, ensure it is not the first character, or escape it with a
704    backslash.
705    
706  For example, the character class [aeiou] matches any lower case vowel, while  For example, the character class [aeiou] matches any lower case vowel, while
707  [^aeiou] matches any character that is not a lower case vowel. Note that a  [^aeiou] matches any character that is not a lower case vowel. Note that a
# Line 683  are in the class by enumerating those th Line 710  are in the class by enumerating those th
710  still consumes a character from the subject string, and fails if the current  still consumes a character from the subject string, and fails if the current
711  pointer is at the end of the string.  pointer is at the end of the string.
712    
713    When caseless matching is set, any letters in a class represent both their
714    upper case and lower case versions, so for example, a caseless [aeiou] matches
715    "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a
716    caseful version would.
717    
718  The newline character is never treated in any special way in character classes,  The newline character is never treated in any special way in character classes,
719  whatever the setting of the PCRE_DOTALL or PCRE_MULTILINE options is. A class  whatever the setting of the PCRE_DOTALL or PCRE_MULTILINE options is. A class
720  such as [^a] will always match a newline.  such as [^a] will always match a newline.
# Line 690  such as [^a] will always match a newline Line 722  such as [^a] will always match a newline
722  The minus (hyphen) character can be used to specify a range of characters in a  The minus (hyphen) character can be used to specify a range of characters in a
723  character class. For example, [d-m] matches any letter between d and m,  character class. For example, [d-m] matches any letter between d and m,
724  inclusive. If a minus character is required in a class, it must be escaped with  inclusive. If a minus character is required in a class, it must be escaped with
725  \\ or appear in a position where it cannot be interpreted as indicating a  a backslash or appear in a position where it cannot be interpreted as
726  range, typically as the first or last character in the class. It is not  indicating a range, typically as the first or last character in the class. It
727  possible to have the character "]" as the end character of a range, since a  is not possible to have the character "]" as the end character of a range,
728  sequence such as [w-] is interpreted as a class of two characters. The octal or  since a sequence such as [w-] is interpreted as a class of two characters. The
729  hexadecimal representation of "]" can, however, be used to end a range.  octal or hexadecimal representation of "]" can, however, be used to end a
730    range.
731    
732  Ranges operate in ASCII collating sequence. They can also be used for  Ranges operate in ASCII collating sequence. They can also be used for
733  characters specified numerically, for example [\\000-\\037]. If a range such as  characters specified numerically, for example [\\000-\\037]. If a range that
734  [W-c] is used when PCRE_CASELESS is set, it matches the letters involved in  includes letters is used when caseless matching is set, it matches the letters
735  either case.  in either case. For example, [W-c] is equivalent to [][\\^_`wxyzabc], matched
736    caselessly, and if character tables for the "fr" locale are in use,
737    [\\xc8-\\xcb] matches accented E characters in both cases.
738    
739  The character types \\d, \\D, \\s, \\S, \\w, and \\W may also appear in a  The character types \\d, \\D, \\s, \\S, \\w, and \\W may also appear in a
740  character class, and add the characters that they match to the class. For  character class, and add the characters that they match to the class. For
741  example, the class [^\\W_] matches any letter or digit.  example, [\\dABCDEF] matches any hexadecimal digit. A circumflex can
742    conveniently be used with the upper case character types to specify a more
743    restricted set of characters than the matching lower case type. For example,
744    the class [^\\W_] matches any letter or digit, but not underscore.
745    
746  All non-alphameric characters other than \\, -, ^ (at the start) and the  All non-alphameric characters other than \\, -, ^ (at the start) and the
747  terminating ] are non-special in character classes, but it does no harm if they  terminating ] are non-special in character classes, but it does no harm if they
# Line 711  are escaped. Line 749  are escaped.
749    
750    
751  .SH VERTICAL BAR  .SH VERTICAL BAR
752  Vertical bar characters are used to separate alternative patterns. The matching  Vertical bar characters are used to separate alternative patterns. For example,
753  process tries all the alternatives in turn. For example, the pattern  the pattern
754    
755    gilbert|sullivan    gilbert|sullivan
756    
757  matches either "gilbert" or "sullivan". Any number of alternatives can be used,  matches either "gilbert" or "sullivan". Any number of alternatives may appear,
758  and an empty alternative is permitted (matching the empty string).  and an empty alternative is permitted (matching the empty string).
759    The matching process tries each alternative in turn, from left to right,
760    and the first one that succeeds is used. If the alternatives are within a
761    subpattern (defined below), "succeeds" means matching the rest of the main
762    pattern as well as the alternative in the subpattern.
763    
764    
765    .SH INTERNAL OPTION SETTING
766    The settings of PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and PCRE_EXTENDED
767    can be changed from within the pattern by a sequence of Perl option letters
768    enclosed between "(?" and ")". The option letters are
769    
770      i  for PCRE_CASELESS
771      m  for PCRE_MULTILINE
772      s  for PCRE_DOTALL
773      x  for PCRE_EXTENDED
774    
775    For example, (?im) sets caseless, multiline matching. It is also possible to
776    unset these options by preceding the letter with a hyphen, and a combined
777    setting and unsetting such as (?im-sx), which sets PCRE_CASELESS and
778    PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is also
779    permitted. If a letter appears both before and after the hyphen, the option is
780    unset.
781    
782    The scope of these option changes depends on where in the pattern the setting
783    occurs. For settings that are outside any subpattern (defined below), the
784    effect is the same as if the options were set or unset at the start of
785    matching. The following patterns all behave in exactly the same way:
786    
787      (?i)abc
788      a(?i)bc
789      ab(?i)c
790      abc(?i)
791    
792    which in turn is the same as compiling the pattern abc with PCRE_CASELESS set.
793    In other words, such "top level" settings apply to the whole pattern (unless
794    there are other changes inside subpatterns). If there is more than one setting
795    of the same option at top level, the rightmost setting is used.
796    
797    If an option change occurs inside a subpattern, the effect is different. This
798    is a change of behaviour in Perl 5.005. An option change inside a subpattern
799    affects only that part of the subpattern that follows it, so
800    
801      (a(?i)b)c
802    
803    matches abc and aBc and no other strings (assuming PCRE_CASELESS is not used).
804    By this means, options can be made to have different settings in different
805    parts of the pattern. Any changes made in one alternative do carry on
806    into subsequent branches within the same subpattern. For example,
807    
808      (a(?i)b|c)
809    
810    matches "ab", "aB", "c", and "C", even though when matching "C" the first
811    branch is abandoned before the option setting. This is because the effects of
812    option settings happen at compile time. There would be some very weird
813    behaviour otherwise.
814    
815    The PCRE-specific options PCRE_UNGREEDY and PCRE_EXTRA can be changed in the
816    same way as the Perl-compatible options by using the characters U and X
817    respectively. The (?X) flag setting is special in that it must always occur
818    earlier in the pattern than any of the additional features it turns on, even
819    when it is at top level. It is best put at the start.
820    
821    
822  .SH SUBPATTERNS  .SH SUBPATTERNS
# Line 757  the captured substrings are "white queen Line 856  the captured substrings are "white queen
856  2. The maximum number of captured substrings is 99, and the maximum number of  2. The maximum number of captured substrings is 99, and the maximum number of
857  all subpatterns, both capturing and non-capturing, is 200.  all subpatterns, both capturing and non-capturing, is 200.
858    
859    As a convenient shorthand, if any option settings are required at the start of
860  .SH BACK REFERENCES  a non-capturing subpattern, the option letters may appear between the "?" and
861  Outside a character class, a backslash followed by a digit greater than 0 (and  the ":". Thus the two patterns
862  possibly further digits) is a back reference to a capturing subpattern earlier  
863  (i.e. to its left) in the pattern, provided there have been that many previous    (?i:saturday|sunday)
864  capturing left parentheses. However, if the decimal number following the    (?:(?i)saturday|sunday)
865  backslash is less than 10, it is always taken as a back reference, and causes  
866  an error if there have not been that many previous capturing left parentheses.  match exactly the same set of strings. Because alternative branches are tried
867  See the section entitled "Backslash" above for further details of the handling  from left to right, and options are not reset until the end of the subpattern
868  of digits following a backslash.  is reached, an option setting in one branch does affect subsequent branches, so
869    the above patterns match "SUNDAY" as well as "Saturday".
 A back reference matches whatever actually matched the capturing subpattern in  
 the current subject string, rather than anything matching the subpattern  
 itself. So the pattern  
   
    (sens|respons)e and \\1ibility  
   
 matches "sense and sensibility" and "response and responsibility", but not  
 "sense and responsibility".  
   
 There may be more than one back reference to the same subpattern. If a  
 subpattern has not actually been used in a particular match, then any back  
 references to it always fail. For example, the pattern  
   
   (a|(bc))\\2  
   
 always fails if it starts to match "a" rather than "bc". Because there may be  
 up to 99 back references, all digits following the backslash are taken  
 as part of a potential back reference number. If the pattern continues with a  
 digit character, then some delimiter must be used to terminate the back  
 reference. If the PCRE_EXTENDED option is set, this can be whitespace.  
 Otherwise an empty comment can be used.  
870    
871    
872  .SH REPETITION  .SH REPETITION
# Line 798  items: Line 876  items:
876    a single character, possibly escaped    a single character, possibly escaped
877    the . metacharacter    the . metacharacter
878    a character class    a character class
879    a back reference    a back reference (see next section)
880    a parenthesized subpattern    a parenthesized subpattern (unless it is an assertion - see below)
881    
882  The general repetition quantifier specifies a minimum and maximum number of  The general repetition quantifier specifies a minimum and maximum number of
883  permitted matches, by giving the two numbers in curly brackets (braces),  permitted matches, by giving the two numbers in curly brackets (braces),
# Line 821  matches at least 3 successive vowels, bu Line 899  matches at least 3 successive vowels, bu
899    
900  matches exactly 8 digits. An opening curly bracket that appears in a position  matches exactly 8 digits. An opening curly bracket that appears in a position
901  where a quantifier is not allowed, or one that does not match the syntax of a  where a quantifier is not allowed, or one that does not match the syntax of a
902  quantifier, is taken as a literal character. For example, "{,6}" is not a  quantifier, is taken as a literal character. For example, {,6} is not a
903  quantifier, but a literal string of four characters.  quantifier, but a literal string of four characters.
904    
905  The quantifier {0} is permitted, causing the expression to behave as if the  The quantifier {0} is permitted, causing the expression to behave as if the
# Line 834  quantifiers have single-character abbrev Line 912  quantifiers have single-character abbrev
912    +    is equivalent to {1,}    +    is equivalent to {1,}
913    ?    is equivalent to {0,1}    ?    is equivalent to {0,1}
914    
915    It is possible to construct infinite loops by following a subpattern that can
916    match no characters with a quantifier that has no upper limit, for example:
917    
918      (a?)*
919    
920    Earlier versions of Perl and PCRE used to give an error at compile time for
921    such patterns. However, because there are cases where this can be useful, such
922    patterns are now accepted, but if any repetition of the subpattern does in fact
923    match no characters, the loop is forcibly broken.
924    
925  By default, the quantifiers are "greedy", that is, they match as much as  By default, the quantifiers are "greedy", that is, they match as much as
926  possible (up to the maximum number of permitted times), without causing the  possible (up to the maximum number of permitted times), without causing the
927  rest of the pattern to fail. The classic example of where this gives problems  rest of the pattern to fail. The classic example of where this gives problems
# Line 861  quantifiers is not otherwise changed, ju Line 949  quantifiers is not otherwise changed, ju
949  Do not confuse this use of question mark with its use as a quantifier in its  Do not confuse this use of question mark with its use as a quantifier in its
950  own right. Because it has two uses, it can sometimes appear doubled, as in  own right. Because it has two uses, it can sometimes appear doubled, as in
951    
952     \\d??\\d    \\d??\\d
953    
954  which matches one digit by preference, but can match two if that is the only  which matches one digit by preference, but can match two if that is the only
955  way the rest of the pattern matches.  way the rest of the pattern matches.
956    
957    If the PCRE_UNGREEDY option is set (an option which is not available in Perl)
958    then the quantifiers are not greedy by default, but individual ones can be made
959    greedy by following them with a question mark. In other words, it inverts the
960    default behaviour.
961    
962  When a parenthesized subpattern is quantified with a minimum repeat count that  When a parenthesized subpattern is quantified with a minimum repeat count that
963  is greater than 1 or with a limited maximum, more store is required for the  is greater than 1 or with a limited maximum, more store is required for the
964  compiled pattern, in proportion to the size of the minimum or maximum.  compiled pattern, in proportion to the size of the minimum or maximum.
# Line 875  follows will be tried against every char Line 968  follows will be tried against every char
968  PCRE treats this as though it were preceded by \\A.  PCRE treats this as though it were preceded by \\A.
969    
970  When a capturing subpattern is repeated, the value captured is the substring  When a capturing subpattern is repeated, the value captured is the substring
971  that matched the final iteration. For example,  that matched the final iteration. For example, after
972    
973      (tweedle[dume]{3}\\s*)+
974    
975    has matched "tweedledum tweedledee" the value of the captured substring is
976    "tweedledee". However, if there are nested capturing subpatterns, the
977    corresponding captured values may have been set in previous iterations. For
978    example, after
979    
980      /(a|(b))+/
981    
982    matches "aba" the value of the second captured substring is "b".
983    
984    
985    .SH BACK REFERENCES
986    Outside a character class, a backslash followed by a digit greater than 0 (and
987    possibly further digits) is a back reference to a capturing subpattern earlier
988    (i.e. to its left) in the pattern, provided there have been that many previous
989    capturing left parentheses.
990    
991    However, if the decimal number following the backslash is less than 10, it is
992    always taken as a back reference, and causes an error only if there are not
993    that many capturing left parentheses in the entire pattern. In other words, the
994    parentheses that are referenced need not be to the left of the reference for
995    numbers less than 10. See the section entitled "Backslash" above for further
996    details of the handling of digits following a backslash.
997    
998    A back reference matches whatever actually matched the capturing subpattern in
999    the current subject string, rather than anything matching the subpattern
1000    itself. So the pattern
1001    
1002     (\s*tweedle[dume]{3})+\\1    (sens|respons)e and \\1ibility
1003    
1004  matches "tweedledum tweedledee tweedledee" but not "tweedledum tweedledee  matches "sense and sensibility" and "response and responsibility", but not
1005  tweedledum".  "sense and responsibility". If caseful matching is in force at the time of the
1006    back reference, then the case of letters is relevant. For example,
1007    
1008      ((?i)rah)\\s+\\1
1009    
1010    matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
1011    capturing subpattern is matched caselessly.
1012    
1013    There may be more than one back reference to the same subpattern. If a
1014    subpattern has not actually been used in a particular match, then any back
1015    references to it always fail. For example, the pattern
1016    
1017      (a|(bc))\\2
1018    
1019    always fails if it starts to match "a" rather than "bc". Because there may be
1020    up to 99 back references, all digits following the backslash are taken
1021    as part of a potential back reference number. If the pattern continues with a
1022    digit character, then some delimiter must be used to terminate the back
1023    reference. If the PCRE_EXTENDED option is set, this can be whitespace.
1024    Otherwise an empty comment can be used.
1025    
1026    A back reference that occurs inside the parentheses to which it refers fails
1027    when the subpattern is first used, so, for example, (a\\1) never matches.
1028    However, such references can be useful inside repeated subpatterns. For
1029    example, the pattern
1030    
1031      (a|b\\1)+
1032    
1033    matches any number of "a"s and also "aba", "ababaa" etc. At each iteration of
1034    the subpattern, the back reference matches the character string corresponding
1035    to the previous iteration. In order for this to work, the pattern must be such
1036    that the first iteration does not need to match the back reference. This can be
1037    done using alternation, as in the example above, or by a quantifier with a
1038    minimum of zero.
1039    
1040    
1041  .SH ASSERTIONS  .SH ASSERTIONS
1042  An assertion is a test on the characters following the current matching point  An assertion is a test on the characters following or preceding the current
1043  that does not actually consume any of those characters. The simple assertions  matching point that does not actually consume any characters. The simple
1044  coded as \\b, \\B, \\A, \\Z, ^ and $ are described above. More complicated  assertions coded as \\b, \\B, \\A, \\Z, \\z, ^ and $ are described above. More
1045  assertions are coded as subpatterns starting with (?= for positive assertions,  complicated assertions are coded as subpatterns. There are two kinds: those
1046  and (?! for negative assertions. For example,  that look ahead of the current position in the subject string, and those that
1047    look behind it.
1048    
1049    An assertion subpattern is matched in the normal way, except that it does not
1050    cause the current matching position to be changed. Lookahead assertions start
1051    with (?= for positive assertions and (?! for negative assertions. For example,
1052    
1053    \\w+(?=;)    \\w+(?=;)
1054    
# Line 904  apparently similar pattern Line 1064  apparently similar pattern
1064    
1065  does not find an occurrence of "bar" that is preceded by something other than  does not find an occurrence of "bar" that is preceded by something other than
1066  "foo"; it finds any occurrence of "bar" whatsoever, because the assertion  "foo"; it finds any occurrence of "bar" whatsoever, because the assertion
1067  (?!foo) is always true when the next three characters are "bar".  (?!foo) is always true when the next three characters are "bar". A
1068    lookbehind assertion is needed to achieve this effect.
1069    
1070    Lookbehind assertions start with (?<= for positive assertions and (?<! for
1071    negative assertions. For example,
1072    
1073      (?<!foo)bar
1074    
1075    does find an occurrence of "bar" that is not preceded by "foo". The contents of
1076    a lookbehind assertion are restricted such that all the strings it matches must
1077    have a fixed length. However, if there are several alternatives, they do not
1078    all have to have the same fixed length. Thus
1079    
1080      (?<=bullock|donkey)
1081    
1082    is permitted, but
1083    
1084      (?<!dogs?|cats?)
1085    
1086    causes an error at compile time. Branches that match different length strings
1087    are permitted only at the top level of a lookbehind assertion. This is an
1088    extension compared with Perl 5.005, which requires all branches to match the
1089    same length of string. An assertion such as
1090    
1091      (?<=ab(c|de))
1092    
1093    is not permitted, because its single top-level branch can match two different
1094    lengths, but it is acceptable if rewritten to use two top-level branches:
1095    
1096      (?<=abc|abde)
1097    
1098    The implementation of lookbehind assertions is, for each alternative, to
1099    temporarily move the current position back by the fixed width and then try to
1100    match. If there are insufficient characters before the current position, the
1101    match is deemed to fail. Lookbehinds in conjunction with once-only subpatterns
1102    can be particularly useful for matching at the ends of strings; an example is
1103    given at the end of the section on once-only subpatterns.
1104    
1105    Several assertions (of any sort) may occur in succession. For example,
1106    
1107      (?<=\\d{3})(?<!999)foo
1108    
1109    matches "foo" preceded by three digits that are not "999". Furthermore,
1110    assertions can be nested in any combination. For example,
1111    
1112      (?<=(?<!foo)bar)baz
1113    
1114    matches an occurrence of "baz" that is preceded by "bar" which in turn is not
1115    preceded by "foo".
1116    
1117  Assertion subpatterns are not capturing subpatterns, and may not be repeated,  Assertion subpatterns are not capturing subpatterns, and may not be repeated,
1118  because it makes no sense to assert the same thing several times. If an  because it makes no sense to assert the same thing several times. If an
# Line 917  Assertions count towards the maximum of Line 1125  Assertions count towards the maximum of
1125    
1126    
1127  .SH ONCE-ONLY SUBPATTERNS  .SH ONCE-ONLY SUBPATTERNS
 The facility described in this section is available only when the PCRE_EXTRA  
 option is set at compile time. It is an extension to Perl regular expressions.  
   
1128  With both maximizing and minimizing repetition, failure of what follows  With both maximizing and minimizing repetition, failure of what follows
1129  normally causes the repeated item to be re-evaluated to see if a different  normally causes the repeated item to be re-evaluated to see if a different
1130  number of repeats allows the rest of the pattern to match. Sometimes it is  number of repeats allows the rest of the pattern to match. Sometimes it is
1131  useful to prevent this, either to change the nature of the match, or to cause  useful to prevent this, either to change the nature of the match, or to cause
1132  it fail earlier than it otherwise might when the author of the pattern knows  it fail earlier than it otherwise might, when the author of the pattern knows
1133  there is no point in carrying on.  there is no point in carrying on.
1134    
1135  Consider, for example, the pattern \\d+foo when applied to the subject line  Consider, for example, the pattern \\d+foo when applied to the subject line
1136    
1137     123456bar    123456bar
1138    
1139  After matching all 6 digits and then failing to match "foo", the normal  After matching all 6 digits and then failing to match "foo", the normal
1140  action of the matcher is to try again with only 5 digits matching the \\d+  action of the matcher is to try again with only 5 digits matching the \\d+
# Line 939  has matched, it is not to be re-evaluate Line 1144  has matched, it is not to be re-evaluate
1144  give up immediately on failing to match "foo" the first time. The notation is  give up immediately on failing to match "foo" the first time. The notation is
1145  another kind of special parenthesis, starting with (?> as in this example:  another kind of special parenthesis, starting with (?> as in this example:
1146    
1147    (?>\d+)bar    (?>\\d+)bar
1148    
1149  This kind of parenthesis "locks up" the  part of the pattern it contains once  This kind of parenthesis "locks up" the  part of the pattern it contains once
1150  it has matched, and a failure further into the pattern is prevented from  it has matched, and a failure further into the pattern is prevented from
1151  backtracking into it. Backtracking past it to previous items, however, works as  backtracking into it. Backtracking past it to previous items, however, works as
1152  normal.  normal.
1153    
1154  For simple cases such as the above example, this feature can be though of as a  An alternative description is that a subpattern of this type matches the string
1155  maximizing repeat that must swallow everything it can. So, while both \\d+ and  of characters that an identical standalone pattern would match, if anchored at
1156  \\d+? are prepared to adjust the number of digits they match in order to make  the current point in the subject string.
1157  the rest of the pattern match, (?>\\d+) can only match an entire sequence of  
1158  digits.  Once-only subpatterns are not capturing subpatterns. Simple cases such as the
1159    above example can be though of as a maximizing repeat that must swallow
1160    everything it can. So, while both \\d+ and \\d+? are prepared to adjust the
1161    number of digits they match in order to make the rest of the pattern match,
1162    (?>\\d+) can only match an entire sequence of digits.
1163    
1164  This construction can of course contain arbitrarily complicated subpatterns,  This construction can of course contain arbitrarily complicated subpatterns,
1165  and it can be nested. Contrast with the \\X assertion, which is a Prolog-like  and it can be nested.
1166  "cut".  
1167    Once-only subpatterns can be used in conjunction with lookbehind assertions to
1168    specify efficient matching at the end of the subject string. Consider a simple
1169    pattern such as
1170    
1171      abcd$
1172    
1173    when applied to a long string which does not match it. Because matching
1174    proceeds from left to right, PCRE will look for each "a" in the subject and
1175    then see if what follows matches the rest of the pattern. If the pattern is
1176    specified as
1177    
1178      .*abcd$
1179    
1180    then the initial .* matches the entire string at first, but when this fails, it
1181    backtracks to match all but the last character, then all but the last two
1182    characters, and so on. Once again the search for "a" covers the entire string,
1183    from right to left, so we are no better off. However, if the pattern is written
1184    as
1185    
1186      (?>.*)(?<=abcd)
1187    
1188    then there can be no backtracking for the .* item; it can match only the entire
1189    string. The subsequent lookbehind assertion does a single test on the last four
1190    characters. If it fails, the match fails immediately. For long strings, this
1191    approach makes a significant difference to the processing time.
1192    
1193    
1194    .SH CONDITIONAL SUBPATTERNS
1195    It is possible to cause the matching process to obey a subpattern
1196    conditionally or to choose between two alternative subpatterns, depending on
1197    the result of an assertion, or whether a previous capturing subpattern matched
1198    or not. The two possible forms of conditional subpattern are
1199    
1200      (?(condition)yes-pattern)
1201      (?(condition)yes-pattern|no-pattern)
1202    
1203    If the condition is satisfied, the yes-pattern is used; otherwise the
1204    no-pattern (if present) is used. If there are more than two alternatives in the
1205    subpattern, a compile-time error occurs.
1206    
1207    There are two kinds of condition. If the text between the parentheses consists
1208    of a sequence of digits, then the condition is satisfied if the capturing
1209    subpattern of that number has previously matched. Consider the following
1210    pattern, which contains non-significant white space to make it more readable
1211    (assume the PCRE_EXTENDED option) and to divide it into three parts for ease
1212    of discussion:
1213    
1214      ( \\( )?    [^()]+    (?(1) \\) )
1215    
1216    The first part matches an optional opening parenthesis, and if that
1217    character is present, sets it as the first captured substring. The second part
1218    matches one or more characters that are not parentheses. The third part is a
1219    conditional subpattern that tests whether the first set of parentheses matched
1220    or not. If they did, that is, if subject started with an opening parenthesis,
1221    the condition is true, and so the yes-pattern is executed and a closing
1222    parenthesis is required. Otherwise, since no-pattern is not present, the
1223    subpattern matches nothing. In other words, this pattern matches a sequence of
1224    non-parentheses, optionally enclosed in parentheses.
1225    
1226    If the condition is not a sequence of digits, it must be an assertion. This may
1227    be a positive or negative lookahead or lookbehind assertion. Consider this
1228    pattern, again containing non-significant white space, and with the two
1229    alternatives on the second line:
1230    
1231      (?(?=[^a-z]*[a-z])
1232      \\d{2}[a-z]{3}-\\d{2}  |  \\d{2}-\\d{2}-\\d{2} )
1233    
1234    The condition is a positive lookahead assertion that matches an optional
1235    sequence of non-letters followed by a letter. In other words, it tests for the
1236    presence of at least one letter in the subject. If a letter is found, the
1237    subject is matched against the first alternative; otherwise it is matched
1238    against the second. This pattern matches strings in one of the two forms
1239    dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
1240    
1241    
1242  .SH COMMENTS  .SH COMMENTS
# Line 967  character class introduces a comment tha Line 1249  character class introduces a comment tha
1249  character in the pattern.  character in the pattern.
1250    
1251    
 .SH INTERNAL FLAG SETTING  
 If the sequence (?i) occurs anywhere in a pattern, it has the effect of setting  
 the PCRE_CASELESS option, that is, all letters are matched in a  
 case-independent manner. The option applies to the whole pattern, not just to  
 the portion that follows it.  
   
 If the sequence (?m) occurs anywhere in a pattern, it has the effect of setting  
 the PCRE_MULTILINE option, that is, subject strings matched by this pattern are  
 treated as consisting of multiple lines.  
   
 If the sequence (?s) occurs anywhere in a pattern, it has the effect of setting  
 the PCRE_DOTALL option, so that dot metacharacters match newlines as well as  
 all other characters.  
   
 If the sequence (?x) occurs anywhere in a pattern, it has the effect of setting  
 the PCRE_EXTENDED option, that is, whitespace is ignored and # introduces a  
 comment that lasts till the next newline. The option applies to the whole  
 pattern, not just to the portion that follows it.  
   
 If more than one option is required, they can be specified jointly, for example  
 as (?ix) or (?mi).  
   
   
1252  .SH PERFORMANCE  .SH PERFORMANCE
1253  Certain items that may appear in patterns are more efficient than others. It is  Certain items that may appear in patterns are more efficient than others. It is
1254  more efficient to use a character class like [aeiou] than a set of alternatives  more efficient to use a character class like [aeiou] than a set of alternatives
# Line 998  required behaviour is usually the most e Line 1257  required behaviour is usually the most e
1257  contains a lot of discussion about optimizing regular expressions for efficient  contains a lot of discussion about optimizing regular expressions for efficient
1258  performance.  performance.
1259    
 The use of PCRE_MULTILINE causes additional processing and should be avoided  
 when it is not necessary. Caseless matching of character classes is more  
 efficient if PCRE_CASELESS is set when the pattern is compiled.  
   
1260    
1261  .SH AUTHOR  .SH AUTHOR
1262  Philip Hazel <ph10@cam.ac.uk>  Philip Hazel <ph10@cam.ac.uk>
# Line 1014  Cambridge CB2 3QG, England. Line 1269  Cambridge CB2 3QG, England.
1269  .br  .br
1270  Phone: +44 1223 334714  Phone: +44 1223 334714
1271    
1272  Copyright (c) 1997 University of Cambridge.  Copyright (c) 1997-1999 University of Cambridge.

Legend:
Removed from v.9  
changed lines
  Added in v.27

  ViewVC Help
Powered by ViewVC 1.1.5