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

Diff of /code/trunk/doc/pcreapi.3

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

revision 856 by ph10, Sat Jan 7 17:02:55 2012 UTC revision 1191 by ph10, Tue Oct 30 16:50:57 2012 UTC
# Line 1  Line 1 
1  .TH PCREAPI 3  .TH PCREAPI 3 "29 October 2012" "PCRE 8.32"
2  .SH NAME  .SH NAME
3  PCRE - Perl-compatible regular expressions  PCRE - Perl-compatible regular expressions
4  .sp  .sp
# Line 134  PCRE - Perl-compatible regular expressio Line 134  PCRE - Perl-compatible regular expressio
134  .B int (*pcre_callout)(pcre_callout_block *);  .B int (*pcre_callout)(pcre_callout_block *);
135  .  .
136  .  .
137  .SH "PCRE 8-BIT AND 16-BIT LIBRARIES"  .SH "PCRE 8-BIT, 16-BIT AND 32-BIT LIBRARIES"
138  .rs  .rs
139  .sp  .sp
140  From release 8.30, PCRE can be compiled as a library for handling 16-bit  From release 8.30, PCRE can be compiled as a library for handling 16-bit
141  character strings as well as, or instead of, the original library that handles  character strings as well as, or instead of, the original library that handles
142  8-bit character strings. To avoid too much complication, this document  8-bit character strings. From release 8.32, PCRE can also be compiled as a
143  describes the 8-bit versions of the functions, with only occasional references  library for handling 32-bit character strings. To avoid too much complication,
144  to the 16-bit library.  this document describes the 8-bit versions of the functions, with only
145  .P  occasional references to the 16-bit and 32-bit libraries.
146  The 16-bit functions operate in the same way as their 8-bit counterparts; they  .P
147  just use different data types for their arguments and results, and their names  The 16-bit and 32-bit functions operate in the same way as their 8-bit
148  start with \fBpcre16_\fP instead of \fBpcre_\fP. For every option that has UTF8  counterparts; they just use different data types for their arguments and
149  in its name (for example, PCRE_UTF8), there is a corresponding 16-bit name with  results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
150  UTF8 replaced by UTF16. This facility is in fact just cosmetic; the 16-bit  \fBpcre_\fP. For every option that has UTF8 in its name (for example,
151  option names define the same bit values.  PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
152    by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
153    16-bit and 32-bit option names define the same bit values.
154  .P  .P
155  References to bytes and UTF-8 in this document should be read as references to  References to bytes and UTF-8 in this document should be read as references to
156  16-bit data quantities and UTF-16 when using the 16-bit library, unless  16-bit data quantities and UTF-16 when using the 16-bit library, or 32-bit data
157  specified otherwise. More details of the specific differences for the 16-bit  quantities and UTF-32 when using the 32-bit library, unless specified
158  library are given in the  otherwise. More details of the specific differences for the 16-bit and 32-bit
159    libraries are given in the
160  .\" HREF  .\" HREF
161  \fBpcre16\fP  \fBpcre16\fP
162  .\"  .\"
163  page.  and
164    .\" HREF
165    \fBpcre32\fP
166    .\"
167    pages.
168  .  .
169  .  .
170  .SH "PCRE API OVERVIEW"  .SH "PCRE API OVERVIEW"
# Line 213  used if available, by setting an option Line 220  used if available, by setting an option
220  relevant. More complicated programs might need to make use of the functions  relevant. More complicated programs might need to make use of the functions
221  \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and  \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
222  \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.  \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
223  These functions are discussed in the  .P
224    From release 8.32 there is also a direct interface for JIT execution, which
225    gives improved performance. The JIT-specific functions are discussed in the
226  .\" HREF  .\" HREF
227  \fBpcrejit\fP  \fBpcrejit\fP
228  .\"  .\"
# Line 302  PCRE supports five different conventions Line 311  PCRE supports five different conventions
311  strings: a single CR (carriage return) character, a single LF (linefeed)  strings: a single CR (carriage return) character, a single LF (linefeed)
312  character, the two-character sequence CRLF, any of the three preceding, or any  character, the two-character sequence CRLF, any of the three preceding, or any
313  Unicode newline sequence. The Unicode newline sequences are the three just  Unicode newline sequence. The Unicode newline sequences are the three just
314  mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,  mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
315  U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS  U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
316  (paragraph separator, U+2029).  (paragraph separator, U+2029).
317  .P  .P
# Line 392  not recognized. The following informatio Line 401  not recognized. The following informatio
401    PCRE_CONFIG_UTF8    PCRE_CONFIG_UTF8
402  .sp  .sp
403  The output is an integer that is set to one if UTF-8 support is available;  The output is an integer that is set to one if UTF-8 support is available;
404  otherwise it is set to zero. If this option is given to the 16-bit version of  otherwise it is set to zero. This value should normally be given to the 8-bit
405  this function, \fBpcre16_config()\fP, the result is PCRE_ERROR_BADOPTION.  version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
406    or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
407  .sp  .sp
408    PCRE_CONFIG_UTF16    PCRE_CONFIG_UTF16
409  .sp  .sp
410  The output is an integer that is set to one if UTF-16 support is available;  The output is an integer that is set to one if UTF-16 support is available;
411  otherwise it is set to zero. This value should normally be given to the 16-bit  otherwise it is set to zero. This value should normally be given to the 16-bit
412  version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit  version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
413  version of this function, the result is PCRE_ERROR_BADOPTION.  or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
414    .sp
415      PCRE_CONFIG_UTF32
416    .sp
417    The output is an integer that is set to one if UTF-32 support is available;
418    otherwise it is set to zero. This value should normally be given to the 32-bit
419    version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
420    or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
421  .sp  .sp
422    PCRE_CONFIG_UNICODE_PROPERTIES    PCRE_CONFIG_UNICODE_PROPERTIES
423  .sp  .sp
# Line 412  properties is available; otherwise it is Line 429  properties is available; otherwise it is
429  The output is an integer that is set to one if support for just-in-time  The output is an integer that is set to one if support for just-in-time
430  compiling is available; otherwise it is set to zero.  compiling is available; otherwise it is set to zero.
431  .sp  .sp
432      PCRE_CONFIG_JITTARGET
433    .sp
434    The output is a pointer to a zero-terminated "const char *" string. If JIT
435    support is available, the string contains the name of the architecture for
436    which the JIT compiler is configured, for example "x86 32bit (little endian +
437    unaligned)". If JIT support is not available, the result is NULL.
438    .sp
439    PCRE_CONFIG_NEWLINE    PCRE_CONFIG_NEWLINE
440  .sp  .sp
441  The output is an integer whose value specifies the default character sequence  The output is an integer whose value specifies the default character sequence
442  that is recognized as meaning "newline". The four values that are supported  that is recognized as meaning "newline". The values that are supported in
443  are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.  ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
444  Though they are derived from ASCII, the same values are returned in EBCDIC  ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
445  environments. The default should normally correspond to the standard sequence  same values. However, the value for LF is normally 21, though some EBCDIC
446  for your operating system.  environments use 37. The corresponding values for CRLF are 3349 and 3365. The
447    default should normally correspond to the standard sequence for your operating
448    system.
449  .sp  .sp
450    PCRE_CONFIG_BSR    PCRE_CONFIG_BSR
451  .sp  .sp
# Line 433  or CRLF. The default can be overridden w Line 459  or CRLF. The default can be overridden w
459  The output is an integer that contains the number of bytes used for internal  The output is an integer that contains the number of bytes used for internal
460  linkage in compiled regular expressions. For the 8-bit library, the value can  linkage in compiled regular expressions. For the 8-bit library, the value can
461  be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still  be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
462  a number of bytes. The default value of 2 is sufficient for all but the most  a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
463  massive patterns, since it allows the compiled pattern to be up to 64K in size.  still a number of bytes. The default value of 2 is sufficient for all but the
464  Larger values allow larger regular expressions to be compiled, at the expense  most massive patterns, since it allows the compiled pattern to be up to 64K in
465  of slower matching.  size. Larger values allow larger regular expressions to be compiled, at the
466    expense of slower matching.
467  .sp  .sp
468    PCRE_CONFIG_POSIX_MALLOC_THRESHOLD    PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
469  .sp  .sp
# Line 519  documentation). For those options that c Line 546  documentation). For those options that c
546  the pattern, the contents of the \fIoptions\fP argument specifies their  the pattern, the contents of the \fIoptions\fP argument specifies their
547  settings at the start of compilation and execution. The PCRE_ANCHORED,  settings at the start of compilation and execution. The PCRE_ANCHORED,
548  PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and  PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
549  PCRE_NO_START_OPT options can be set at the time of matching as well as at  PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
550  compile time.  compile time.
551  .P  .P
552  If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.  If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
# Line 635  documentation. Line 662  documentation.
662  .sp  .sp
663    PCRE_EXTENDED    PCRE_EXTENDED
664  .sp  .sp
665  If this bit is set, whitespace data characters in the pattern are totally  If this bit is set, white space data characters in the pattern are totally
666  ignored except when escaped or inside a character class. Whitespace does not  ignored except when escaped or inside a character class. White space does not
667  include the VT character (code 11). In addition, characters between an  include the VT character (code 11). In addition, characters between an
668  unescaped # outside a character class and the next newline, inclusive, are also  unescaped # outside a character class and the next newline, inclusive, are also
669  ignored. This is equivalent to Perl's /x option, and it can be changed within a  ignored. This is equivalent to Perl's /x option, and it can be changed within a
# Line 654  comment is a literal newline sequence in Line 681  comment is a literal newline sequence in
681  happen to represent a newline do not count.  happen to represent a newline do not count.
682  .P  .P
683  This option makes it possible to include comments inside complicated patterns.  This option makes it possible to include comments inside complicated patterns.
684  Note, however, that this applies only to data characters. Whitespace characters  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  may never appear within special character sequences in a pattern, for example
686  within the sequence (?( that introduces a conditional subpattern.  within the sequence (?( that introduces a conditional subpattern.
687  .sp  .sp
# Line 732  indicated by a single character (CR or L Line 759  indicated by a single character (CR or L
759  PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character  PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
760  CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three  CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
761  preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies  preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
762  that any Unicode newline sequence should be recognized. The Unicode newline  that any Unicode newline sequence should be recognized.
763  sequences are the three just mentioned, plus the single characters VT (vertical  .P
764  tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line  In an ASCII/Unicode environment, the Unicode newline sequences are the three
765  separator, U+2028), and PS (paragraph separator, U+2029). For the 8-bit  just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
766  library, the last two are recognized only in UTF-8 mode.  feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
767    (paragraph separator, U+2029). For the 8-bit library, the last two are
768    recognized only in UTF-8 mode.
769    .P
770    When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
771    CR is 0x0d, the same as ASCII. However, the character code for LF is normally
772    0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
773    not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
774    less than 256. For more details, see the
775    .\" HREF
776    \fBpcrebuild\fP
777    .\"
778    documentation.
779  .P  .P
780  The newline setting in the options word uses three bits that are treated  The newline setting in the options word uses three bits that are treated
781  as a number, giving eight possibilities. Currently only six are used (default  as a number, giving eight possibilities. Currently only six are used (default
# Line 746  PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is Line 785  PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is
785  other combinations may yield unused numbers and cause an error.  other combinations may yield unused numbers and cause an error.
786  .P  .P
787  The only time that a line break in a pattern is specially recognized when  The only time that a line break in a pattern is specially recognized when
788  compiling is when PCRE_EXTENDED is set. CR and LF are whitespace characters,  compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
789  and so are ignored in this mode. Also, an unescaped # outside a character class  and so are ignored in this mode. Also, an unescaped # outside a character class
790  indicates a comment that lasts until after the next line break sequence. In  indicates a comment that lasts until after the next line break sequence. In
791  other circumstances, line break sequences in patterns are treated as literal  other circumstances, line break sequences in patterns are treated as literal
# Line 812  page. Line 851  page.
851  .sp  .sp
852    PCRE_NO_UTF8_CHECK    PCRE_NO_UTF8_CHECK
853  .sp  .sp
854  When PCRE_UTF8 is set, the validity of the pattern as a UTF-8  When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
855  string is automatically checked. There is a discussion about the  automatically checked. There is a discussion about the
856  .\" HTML <a href="pcreunicode.html#utf8strings">  .\" HTML <a href="pcreunicode.html#utf8strings">
857  .\" </a>  .\" </a>
858  validity of UTF-8 strings  validity of UTF-8 strings
859  .\"  .\"
860  in the  in the
861  .\" HREF  .\" HREF
# Line 828  this check for performance reasons, you Line 867  this check for performance reasons, you
867  When it is set, the effect of passing an invalid UTF-8 string as a pattern is  When it is set, the effect of passing an invalid UTF-8 string as a pattern is
868  undefined. It may cause your program to crash. Note that this option can also  undefined. It may cause your program to crash. Note that this option can also
869  be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the  be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
870  validity checking of subject strings.  validity checking of subject strings only. If the same string is being matched
871    many times, the option can be safely set for the second and subsequent
872    matchings to improve performance.
873  .  .
874  .  .
875  .SH "COMPILATION ERROR CODES"  .SH "COMPILATION ERROR CODES"
# Line 836  validity checking of subject strings. Line 877  validity checking of subject strings.
877  .sp  .sp
878  The following table lists the error codes than may be returned by  The following table lists the error codes than may be returned by
879  \fBpcre_compile2()\fP, along with the error messages that may be returned by  \fBpcre_compile2()\fP, along with the error messages that may be returned by
880  both compiling functions. Note that error messages are always 8-bit ASCII  both compiling functions. Note that error messages are always 8-bit ASCII
881  strings, even in 16-bit mode. As PCRE has developed, some error codes have  strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
882  fallen out of use. To avoid confusion, they have not been re-used.  have fallen out of use. To avoid confusion, they have not been re-used.
883  .sp  .sp
884     0  no error     0  no error
885     1  \e at end of pattern     1  \e at end of pattern
# Line 910  fallen out of use. To avoid confusion, t Line 951  fallen out of use. To avoid confusion, t
951    65  different names for subpatterns of the same number are    65  different names for subpatterns of the same number are
952          not allowed          not allowed
953    66  (*MARK) must have an argument    66  (*MARK) must have an argument
954    67  this version of PCRE is not compiled with Unicode property    67  this version of PCRE is not compiled with Unicode property
955          support          support
956    68  \ec must be followed by an ASCII character    68  \ec must be followed by an ASCII character
957    69  \ek is not followed by a braced, angle-bracketed, or quoted name    69  \ek is not followed by a braced, angle-bracketed, or quoted name
958    70  internal error: unknown opcode in find_fixedlength()    70  internal error: unknown opcode in find_fixedlength()
959    71  \eN is not supported in a class    71  \eN is not supported in a class
960    72  too many forward references    72  too many forward references
961    73  disallowed Unicode code point (>= 0xd800 && <= 0xdfff)    73  disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
962    74  invalid UTF-16 string (specifically UTF-16)    74  invalid UTF-16 string (specifically UTF-16)
963      75  name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
964      76  character value in \eu.... sequence is too large
965      77  invalid UTF-32 string (specifically UTF-32)
966  .sp  .sp
967  The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may  The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
968  be used if the limits were changed when PCRE was built.  be used if the limits were changed when PCRE was built.
# Line 951  below Line 995  below
995  in the section on matching a pattern.  in the section on matching a pattern.
996  .P  .P
997  If studying the pattern does not produce any useful information,  If studying the pattern does not produce any useful information,
998  \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program  \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
999  wants to pass any of the other fields to \fBpcre_exec()\fP or  calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1000  \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.  \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1001  .P  if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1002  The second argument of \fBpcre_study()\fP contains option bits. There is only  returns a \fBpcre_extra\fP block even if studying did not find any additional
1003  one option: PCRE_STUDY_JIT_COMPILE. If this is set, and the just-in-time  information. It may still return NULL, however, if an error occurs in
1004  compiler is available, the pattern is further compiled into machine code that  \fBpcre_study()\fP.
1005  executes much faster than the \fBpcre_exec()\fP matching function. If  .P
1006  the just-in-time compiler is not available, this option is ignored. All other  The second argument of \fBpcre_study()\fP contains option bits. There are three
1007  bits in the \fIoptions\fP argument must be zero.  further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1008    .sp
1009      PCRE_STUDY_JIT_COMPILE
1010      PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1011      PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1012    .sp
1013    If any of these are set, and the just-in-time compiler is available, the
1014    pattern is further compiled into machine code that executes much faster than
1015    the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1016    compiler is not available, these options are ignored. All undefined bits in the
1017    \fIoptions\fP argument must be zero.
1018  .P  .P
1019  JIT compilation is a heavyweight optimization. It can take some time for  JIT compilation is a heavyweight optimization. It can take some time for
1020  patterns to be analyzed, and for one-off matches and simple patterns the  patterns to be analyzed, and for one-off matches and simple patterns the
# Line 984  When you are finished with a pattern, yo Line 1038  When you are finished with a pattern, yo
1038  study data by calling \fBpcre_free_study()\fP. This function was added to the  study data by calling \fBpcre_free_study()\fP. This function was added to the
1039  API for release 8.20. For earlier versions, the memory could be freed with  API for release 8.20. For earlier versions, the memory could be freed with
1040  \fBpcre_free()\fP, just like the pattern itself. This will still work in cases  \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1041  where PCRE_STUDY_JIT_COMPILE is not used, but it is advisable to change to the  where JIT optimization is not used, but it is advisable to change to the new
1042  new function when convenient.  function when convenient.
1043  .P  .P
1044  This is a typical way in which \fBpcre_study\fP() is used (except that in a  This is a typical way in which \fBpcre_study\fP() is used (except that in a
1045  real application there should be tests for errors):  real application there should be tests for errors):
# Line 1007  real application there should be tests f Line 1061  real application there should be tests f
1061  Studying a pattern does two things: first, a lower bound for the length of  Studying a pattern does two things: first, a lower bound for the length of
1062  subject string that is needed to match the pattern is computed. This does not  subject string that is needed to match the pattern is computed. This does not
1063  mean that there are any strings of that length that match, but it does  mean that there are any strings of that length that match, but it does
1064  guarantee that no shorter strings match. The value is used by  guarantee that no shorter strings match. The value is used to avoid wasting
1065  \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to  time by trying to match strings that are shorter than the lower bound. You can
1066  match strings that are shorter than the lower bound. You can find out the value  find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
 in a calling program via the \fBpcre_fullinfo()\fP function.  
1067  .P  .P
1068  Studying a pattern is also useful for non-anchored patterns that do not have a  Studying a pattern is also useful for non-anchored patterns that do not have a
1069  single fixed starting character. A bitmap of possible starting bytes is  single fixed starting character. A bitmap of possible starting bytes is
1070  created. This speeds up finding a position in the subject at which to start  created. This speeds up finding a position in the subject at which to start
1071  matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.)  matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1072    In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1073  .P  .P
1074  These two optimizations apply to both \fBpcre_exec()\fP and  These two optimizations apply to both \fBpcre_exec()\fP and
1075  \fBpcre_dfa_exec()\fP. However, they are not used by \fBpcre_exec()\fP if  \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1076  \fBpcre_study()\fP is called with the PCRE_STUDY_JIT_COMPILE option, and  The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option
1077  just-in-time compiling is successful. The optimizations can be disabled by  when calling \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP, but if this is done,
1078  setting the PCRE_NO_START_OPTIMIZE option when calling \fBpcre_exec()\fP or  JIT execution is also disabled. You might want to do this if your pattern
1079  \fBpcre_dfa_exec()\fP. You might want to do this if your pattern contains  contains callouts or (*MARK) and you want to make use of these facilities in
1080  callouts or (*MARK) (which cannot be handled by the JIT compiler), and you want  cases where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
 to make use of these facilities in cases where matching fails. See the  
 discussion of PCRE_NO_START_OPTIMIZE  
1081  .\" HTML <a href="#execoptions">  .\" HTML <a href="#execoptions">
1082  .\" </a>  .\" </a>
1083  below.  below.
# Line 1113  the following negative numbers: Line 1165  the following negative numbers:
1165    PCRE_ERROR_NULL           the argument \fIcode\fP was NULL    PCRE_ERROR_NULL           the argument \fIcode\fP was NULL
1166                              the argument \fIwhere\fP was NULL                              the argument \fIwhere\fP was NULL
1167    PCRE_ERROR_BADMAGIC       the "magic number" was not found    PCRE_ERROR_BADMAGIC       the "magic number" was not found
1168    PCRE_ERROR_BADENDIANNESS  the pattern was compiled with different    PCRE_ERROR_BADENDIANNESS  the pattern was compiled with different
1169                              endianness                              endianness
1170    PCRE_ERROR_BADOPTION      the value of \fIwhat\fP was invalid    PCRE_ERROR_BADOPTION      the value of \fIwhat\fP was invalid
1171  .sp  .sp
1172  The "magic number" is placed at the start of each compiled pattern as an simple  The "magic number" is placed at the start of each compiled pattern as an simple
1173  check against passing an arbitrary memory pointer. The endianness error can  check against passing an arbitrary memory pointer. The endianness error can
1174  occur if a compiled pattern is saved and reloaded on a different host. Here is  occur if a compiled pattern is saved and reloaded on a different host. Here is
1175  a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled  a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1176  pattern:  pattern:
# Line 1161  where data units are bytes.) The fourth Line 1213  where data units are bytes.) The fourth
1213  variable.  variable.
1214  .P  .P
1215  If there is a fixed first value, for example, the letter "c" from a pattern  If there is a fixed first value, for example, the letter "c" from a pattern
1216  such as (cat|cow|coyote), its value is returned. In the 8-bit library, the  such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1217  value is always less than 256; in the 16-bit library the value can be up to  value is always less than 256. In the 16-bit library the value can be up to
1218  0xffff.  0xffff. In the 32-bit library the value can be up to 0x10ffff.
1219  .P  .P
1220  If there is no fixed first value, and if either  If there is no fixed first value, and if either
1221  .sp  .sp
# Line 1176  starts with "^", or Line 1228  starts with "^", or
1228  -1 is returned, indicating that the pattern matches only at the start of a  -1 is returned, indicating that the pattern matches only at the start of a
1229  subject string or after any newline within the string. Otherwise -2 is  subject string or after any newline within the string. Otherwise -2 is
1230  returned. For anchored patterns, -2 is returned.  returned. For anchored patterns, -2 is returned.
1231    .P
1232    Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1233    to return the full 32-bit range of the character, this value is deprecated;
1234    instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1235    should be used.
1236  .sp  .sp
1237    PCRE_INFO_FIRSTTABLE    PCRE_INFO_FIRSTTABLE
1238  .sp  .sp
# Line 1198  Return 1 if the (?J) or (?-J) option set Line 1255  Return 1 if the (?J) or (?-J) option set
1255  .sp  .sp
1256    PCRE_INFO_JIT    PCRE_INFO_JIT
1257  .sp  .sp
1258  Return 1 if the pattern was studied with the PCRE_STUDY_JIT_COMPILE option, and  Return 1 if the pattern was studied with one of the JIT options, and
1259  just-in-time compiling was successful. The fourth argument should point to an  just-in-time compiling was successful. The fourth argument should point to an
1260  \fBint\fP variable. A return value of 0 means that JIT support is not available  \fBint\fP variable. A return value of 0 means that JIT support is not available
1261  in this version of PCRE, or that the pattern was not studied with the  in this version of PCRE, or that the pattern was not studied with a JIT option,
1262  PCRE_STUDY_JIT_COMPILE option, or that the JIT compiler could not handle this  or that the JIT compiler could not handle this particular pattern. See the
 particular pattern. See the  
1263  .\" HREF  .\" HREF
1264  \fBpcrejit\fP  \fBpcrejit\fP
1265  .\"  .\"
# Line 1211  documentation for details of what can an Line 1267  documentation for details of what can an
1267  .sp  .sp
1268    PCRE_INFO_JITSIZE    PCRE_INFO_JITSIZE
1269  .sp  .sp
1270  If the pattern was successfully studied with the PCRE_STUDY_JIT_COMPILE option,  If the pattern was successfully studied with a JIT option, return the size of
1271  return the size of the JIT compiled code, otherwise return zero. The fourth  the JIT compiled code, otherwise return zero. The fourth argument should point
1272  argument should point to a \fBsize_t\fP variable.  to a \fBsize_t\fP variable.
1273  .sp  .sp
1274    PCRE_INFO_LASTLITERAL    PCRE_INFO_LASTLITERAL
1275  .sp  .sp
# Line 1224  value, -1 is returned. For anchored patt Line 1280  value, -1 is returned. For anchored patt
1280  only if it follows something of variable length. For example, for the pattern  only if it follows something of variable length. For example, for the pattern
1281  /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value  /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1282  is -1.  is -1.
1283    .P
1284    Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1285    to return the full 32-bit range of the character, this value is deprecated;
1286    instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1287    be used.
1288    .sp
1289      PCRE_INFO_MAXLOOKBEHIND
1290    .sp
1291    Return the number of characters (NB not bytes) in the longest lookbehind
1292    assertion in the pattern. Note that the simple assertions \eb and \eB require a
1293    one-character lookbehind. This information is useful when doing multi-segment
1294    matching using the partial matching facilities.
1295  .sp  .sp
1296    PCRE_INFO_MINLENGTH    PCRE_INFO_MINLENGTH
1297  .sp  .sp
# Line 1256  length of the longest name. PCRE_INFO_NA Line 1324  length of the longest name. PCRE_INFO_NA
1324  entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where  entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1325  the first two bytes of each entry are the number of the capturing parenthesis,  the first two bytes of each entry are the number of the capturing parenthesis,
1326  most significant byte first. In the 16-bit library, the pointer points to  most significant byte first. In the 16-bit library, the pointer points to
1327  16-bit data units, the first of which contains the parenthesis number. The rest  16-bit data units, the first of which contains the parenthesis number.
1328    In the 32-bit library, the pointer points to 32-bit data units, the first of
1329    which contains the parenthesis number. The rest
1330  of the entry is the corresponding name, zero terminated.  of the entry is the corresponding name, zero terminated.
1331  .P  .P
1332  The names are in alphabetical order. Duplicate names may appear if (?| is used  The names are in alphabetical order. Duplicate names may appear if (?| is used
# Line 1358  is made available via this option so tha Line 1428  is made available via this option so tha
1428  \fBpcreprecompile\fP  \fBpcreprecompile\fP
1429  .\"  .\"
1430  documentation for details).  documentation for details).
1431    .sp
1432      PCRE_INFO_FIRSTCHARACTERFLAGS
1433    .sp
1434    Return information about the first data unit of any matched string, for a
1435    non-anchored pattern. The fourth argument should point to an \fBint\fP
1436    variable.
1437    .P
1438    If there is a fixed first value, for example, the letter "c" from a pattern
1439    such as (cat|cow|coyote), 1 is returned, and the character value can be
1440    retrieved using PCRE_INFO_FIRSTCHARACTER.
1441    .P
1442    If there is no fixed first value, and if either
1443    .sp
1444    (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1445    starts with "^", or
1446    .sp
1447    (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1448    (if it were set, the pattern would be anchored),
1449    .sp
1450    2 is returned, indicating that the pattern matches only at the start of a
1451    subject string or after any newline within the string. Otherwise 0 is
1452    returned. For anchored patterns, 0 is returned.
1453    .sp
1454      PCRE_INFO_FIRSTCHARACTER
1455    .sp
1456    Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1457    returned 1; otherwise returns 0. The fourth argument should point to an
1458    \fBuint_t\fP variable.
1459    .P
1460    In the 8-bit library, the value is always less than 256. In the 16-bit library
1461    the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1462    can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1463    .P
1464    If there is no fixed first value, and if either
1465    .sp
1466    (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1467    starts with "^", or
1468    .sp
1469    (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1470    (if it were set, the pattern would be anchored),
1471    .sp
1472    -1 is returned, indicating that the pattern matches only at the start of a
1473    subject string or after any newline within the string. Otherwise -2 is
1474    returned. For anchored patterns, -2 is returned.
1475    .sp
1476      PCRE_INFO_REQUIREDCHARFLAGS
1477    .sp
1478    Returns 1 if there is a rightmost literal data unit that must exist in any
1479    matched string, other than at its start. The fourth argument should  point to
1480    an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1481    1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1482    .P
1483    For anchored patterns, a last literal value is recorded only if it follows
1484    something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1485    returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1486    /^a\edz\ed/ the returned value is 0.
1487    .sp
1488      PCRE_INFO_REQUIREDCHAR
1489    .sp
1490    Return the value of the rightmost literal data unit that must exist in any
1491    matched string, other than at its start, if such a value has been recorded. The
1492    fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1493    value, 0 is returned.
1494  .  .
1495  .  .
1496  .SH "REFERENCE COUNTS"  .SH "REFERENCE COUNTS"
# Line 1452  fields (not necessarily in this order): Line 1585  fields (not necessarily in this order):
1585    const unsigned char *\fItables\fP;    const unsigned char *\fItables\fP;
1586    unsigned char **\fImark\fP;    unsigned char **\fImark\fP;
1587  .sp  .sp
1588  In the 16-bit version of this structure, the \fImark\fP field has type  In the 16-bit version of this structure, the \fImark\fP field has type
1589  "PCRE_SCHAR16 **".  "PCRE_UCHAR16 **".
1590    .sp
1591    In the 32-bit version of this structure, the \fImark\fP field has type
1592    "PCRE_UCHAR32 **".
1593  .P  .P
1594  The \fIflags\fP field is a bitmap that specifies which of the other fields  The \fIflags\fP field is used to specify which of the other fields are set. The
1595  are set. The flag bits are:  flag bits are:
1596  .sp  .sp
1597    PCRE_EXTRA_STUDY_DATA    PCRE_EXTRA_CALLOUT_DATA
1598    PCRE_EXTRA_EXECUTABLE_JIT    PCRE_EXTRA_EXECUTABLE_JIT
1599      PCRE_EXTRA_MARK
1600    PCRE_EXTRA_MATCH_LIMIT    PCRE_EXTRA_MATCH_LIMIT
1601    PCRE_EXTRA_MATCH_LIMIT_RECURSION    PCRE_EXTRA_MATCH_LIMIT_RECURSION
1602    PCRE_EXTRA_CALLOUT_DATA    PCRE_EXTRA_STUDY_DATA
1603    PCRE_EXTRA_TABLES    PCRE_EXTRA_TABLES
   PCRE_EXTRA_MARK  
1604  .sp  .sp
1605  Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes  Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1606  the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is  the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1607  returned by \fBpcre_study()\fP, together with the appropriate flag bits. You  returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1608  should not set these yourself, but you may add to the block by setting the  should not set these yourself, but you may add to the block by setting other
1609  other fields and their corresponding flag bits.  fields and their corresponding flag bits.
1610  .P  .P
1611  The \fImatch_limit\fP field provides a means of preventing PCRE from using up a  The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1612  vast amount of resources when running patterns that are not going to match,  vast amount of resources when running patterns that are not going to match,
# Line 1485  patterns that are not anchored, the coun Line 1621  patterns that are not anchored, the coun
1621  in the subject string.  in the subject string.
1622  .P  .P
1623  When \fBpcre_exec()\fP is called with a pattern that was successfully studied  When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1624  with the PCRE_STUDY_JIT_COMPILE option, the way that the matching is executed  with a JIT option, the way that the matching is executed is entirely different.
1625  is entirely different. However, there is still the possibility of runaway  However, there is still the possibility of runaway matching that goes on for a
1626  matching that goes on for a very long time, and so the \fImatch_limit\fP value  very long time, and so the \fImatch_limit\fP value is also used in this case
1627  is also used in this case (but in a different way) to limit how long the  (but in a different way) to limit how long the matching can continue.
 matching can continue.  
1628  .P  .P
1629  The default value for the limit can be set when PCRE is built; the default  The default value for the limit can be set when PCRE is built; the default
1630  default is 10 million, which handles all but the most extreme cases. You can  default is 10 million, which handles all but the most extreme cases. You can
# Line 1507  This limit is of use only if it is set s Line 1642  This limit is of use only if it is set s
1642  Limiting the recursion depth limits the amount of machine stack that can be  Limiting the recursion depth limits the amount of machine stack that can be
1643  used, or, when PCRE has been compiled to use memory on the heap instead of the  used, or, when PCRE has been compiled to use memory on the heap instead of the
1644  stack, the amount of heap memory that can be used. This limit is not relevant,  stack, the amount of heap memory that can be used. This limit is not relevant,
1645  and is ignored, if the pattern was successfully studied with  and is ignored, when matching is done using JIT compiled code.
 PCRE_STUDY_JIT_COMPILE.  
1646  .P  .P
1647  The default value for \fImatch_limit_recursion\fP can be set when PCRE is  The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1648  built; the default default is the same value as the default for  built; the default default is the same value as the default for
# Line 1565  documentation. Line 1699  documentation.
1699  The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be  The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1700  zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,  zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1701  PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,  PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1702  PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and  PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1703  PCRE_PARTIAL_HARD.  PCRE_PARTIAL_SOFT.
1704  .P  .P
1705  If the pattern was successfully studied with the PCRE_STUDY_JIT_COMPILE option,  If the pattern was successfully studied with one of the just-in-time (JIT)
1706  the only supported options for JIT execution are PCRE_NO_UTF8_CHECK,  compile options, the only supported options for JIT execution are
1707  PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, and PCRE_NOTEMPTY_ATSTART. Note in  PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1708  particular that partial matching is not supported. If an unsupported option is  PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1709  used, JIT execution is disabled and the normal interpretive code in  unsupported option is used, JIT execution is disabled and the normal
1710  \fBpcre_exec()\fP is run.  interpretive code in \fBpcre_exec()\fP is run.
1711  .sp  .sp
1712    PCRE_ANCHORED    PCRE_ANCHORED
1713  .sp  .sp
# Line 1692  causing performance to suffer, but ensur Line 1826  causing performance to suffer, but ensur
1826  "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)  "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1827  are considered at every possible starting position in the subject string. If  are considered at every possible starting position in the subject string. If
1828  PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching  PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1829  time.  time. The use of PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set,
1830    matching is always done using interpretively.
1831  .P  .P
1832  Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.  Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1833  Consider the pattern  Consider the pattern
# Line 1725  returned. Line 1860  returned.
1860  .sp  .sp
1861  When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8  When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1862  string is automatically checked when \fBpcre_exec()\fP is subsequently called.  string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1863  The value of \fIstartoffset\fP is also checked to ensure that it points to the  The entire string is checked before any other processing takes place. The value
1864  start of a UTF-8 character. There is a discussion about the validity of UTF-8  of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1865  strings in the  UTF-8 character. There is a discussion about the
1866    .\" HTML <a href="pcreunicode.html#utf8strings">
1867    .\" </a>
1868    validity of UTF-8 strings
1869    .\"
1870    in the
1871  .\" HREF  .\" HREF
1872  \fBpcreunicode\fP  \fBpcreunicode\fP
1873  .\"  .\"
# Line 1875  string that it matched that is returned. Line 2015  string that it matched that is returned.
2015  .P  .P
2016  If the vector is too small to hold all the captured substring offsets, it is  If the vector is too small to hold all the captured substring offsets, it is
2017  used as far as possible (up to two-thirds of its length), and the function  used as far as possible (up to two-thirds of its length), and the function
2018  returns a value of zero. If neither the actual string matched not any captured  returns a value of zero. If neither the actual string matched nor any captured
2019  substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP  substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2020  passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains  passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2021  back references and the \fIovector\fP is not big enough to remember the related  back references and the \fIovector\fP is not big enough to remember the related
# Line 2075  time. Line 2215  time.
2215  .sp  .sp
2216    PCRE_ERROR_JIT_STACKLIMIT (-27)    PCRE_ERROR_JIT_STACKLIMIT (-27)
2217  .sp  .sp
2218  This error is returned when a pattern that was successfully studied using the  This error is returned when a pattern that was successfully studied using a
2219  PCRE_STUDY_JIT_COMPILE option is being matched, but the memory available for  JIT compile option is being matched, but the memory available for the
2220  the just-in-time processing stack is not large enough. See the  just-in-time processing stack is not large enough. See the
2221  .\" HREF  .\" HREF
2222  \fBpcrejit\fP  \fBpcrejit\fP
2223  .\"  .\"
2224  documentation for more details.  documentation for more details.
2225  .sp  .sp
2226    PCRE_ERROR_BADMODE (-28)    PCRE_ERROR_BADMODE        (-28)
2227    .sp
2228    This error is given if a pattern that was compiled by the 8-bit library is
2229    passed to a 16-bit or 32-bit library function, or vice versa.
2230  .sp  .sp
2231  This error is given if a pattern that was compiled by the 8-bit library is    PCRE_ERROR_BADENDIANNESS  (-29)
 passed to a 16-bit library function, or vice versa.  
2232  .sp  .sp
2233    PCRE_ERROR_BADENDIANNESS (-29)  This error is given if a pattern that was compiled and saved is reloaded on a
2234  .sp  host with different endianness. The utility function
2235  This error is given if a pattern that was compiled and saved is reloaded on a  \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
 host with different endianness. The utility function  
 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern  
2236  so that it runs on the new host.  so that it runs on the new host.
2237    .sp
2238      PCRE_ERROR_BADLENGTH      (-32)
2239    .sp
2240    This error is given if \fBpcre_exec()\fP is called with a negative value for
2241    the \fIlength\fP argument.
2242  .P  .P
2243  Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.  Error numbers -16 to -20, -22, 30, and -31 are not used by \fBpcre_exec()\fP.
2244  .  .
2245  .  .
2246  .\" HTML <a name="badutf8reasons"></a>  .\" HTML <a name="badutf8reasons"></a>
2247  .SS "Reason codes for invalid UTF-8 strings"  .SS "Reason codes for invalid UTF-8 strings"
2248  .rs  .rs
2249  .sp  .sp
2250  This section applies only to the 8-bit library. The corresponding information  This section applies only to the 8-bit library. The corresponding information
2251  for the 16-bit library is given in the  for the 16-bit library is given in the
2252  .\" HREF  .\" HREF
2253  \fBpcre16\fP  \fBpcre16\fP
2254  .\"  .\"
2255    page. The corresponding information for the 32-bit library is given in the
2256    .\" HREF
2257    \fBpcre32\fP
2258    .\"
2259  page.  page.
2260  .P  .P
2261  When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or  When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
# Line 2177  character. Line 2326  character.
2326  .sp  .sp
2327  The first byte of a character has the value 0xfe or 0xff. These values can  The first byte of a character has the value 0xfe or 0xff. These values can
2328  never occur in a valid UTF-8 string.  never occur in a valid UTF-8 string.
2329    .sp
2330      PCRE_UTF8_ERR2
2331    .sp
2332    Non-character. These are the last two characters in each plane (0xfffe, 0xffff,
2333    0x1fffe, 0x1ffff .. 0x10fffe, 0x10ffff), and the characters 0xfdd0..0xfdef.
2334  .  .
2335  .  .
2336  .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"  .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
# Line 2406  other alternatives. Ultimately, when it Line 2560  other alternatives. Ultimately, when it
2560  will yield PCRE_ERROR_NOMATCH.  will yield PCRE_ERROR_NOMATCH.
2561  .  .
2562  .  .
2563    .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2564    .rs
2565    .sp
2566    Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2567    stack, which in certain environments can be rather limited in size. Some users
2568    find it helpful to have an estimate of the amount of stack that is used by
2569    \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2570    .\" HREF
2571    \fBpcrestack\fP
2572    .\"
2573    documentation. The estimate that is output by \fBpcretest\fP when called with
2574    the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2575    the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2576    .P
2577    Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2578    the negative error code PCRE_ERROR_NULL, but with this special combination of
2579    arguments, it returns instead a negative number whose absolute value is the
2580    approximate stack frame size in bytes. (A negative number is used so that it is
2581    clear that no match has happened.) The value is approximate because in some
2582    cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2583    additional variables on the stack.
2584    .P
2585    If PCRE has been compiled to use the heap instead of the stack for recursion,
2586    the value returned is the size of each block that is obtained from the heap.
2587    .
2588    .
2589  .\" HTML <a name="dfamatch"></a>  .\" HTML <a name="dfamatch"></a>
2590  .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"  .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2591  .rs  .rs
# Line 2587  When a recursive subpattern is processed Line 2767  When a recursive subpattern is processed
2767  recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This  recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2768  error is given if the output vector is not large enough. This should be  error is given if the output vector is not large enough. This should be
2769  extremely rare, as a vector of size 1000 is used.  extremely rare, as a vector of size 1000 is used.
2770    .sp
2771      PCRE_ERROR_DFA_BADRESTART (-30)
2772    .sp
2773    When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2774    some plausibility checks are made on the contents of the workspace, which
2775    should contain data about the previous partial match. If any of these checks
2776    fail, this error is given.
2777  .  .
2778  .  .
2779  .SH "SEE ALSO"  .SH "SEE ALSO"
2780  .rs  .rs
2781  .sp  .sp
2782  \fBpcre16\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),  \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2783  \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),  \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2784  \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).  \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2785    \fBpcrestack\fP(3).
2786  .  .
2787  .  .
2788  .SH AUTHOR  .SH AUTHOR
# Line 2611  Cambridge CB2 3QH, England. Line 2799  Cambridge CB2 3QH, England.
2799  .rs  .rs
2800  .sp  .sp
2801  .nf  .nf
2802  Last updated: 07 January 2012  Last updated: 29 October 2012
2803  Copyright (c) 1997-2012 University of Cambridge.  Copyright (c) 1997-2012 University of Cambridge.
2804  .fi  .fi

Legend:
Removed from v.856  
changed lines
  Added in v.1191

  ViewVC Help
Powered by ViewVC 1.1.5