ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 11 by nigel, Sat Feb 24 21:38:17 2007 UTC revision 583 by ph10, Tue Jan 11 16:49:55 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.12 11-Jan-2011
5    ------------------------
7    1.  Fixed some typos in the markup of the man pages, and wrote a script that
8        checks for such things as part of the documentation building process.
11    Version 8.11 10-Dec-2010
12    ------------------------
14    1.  (*THEN) was not working properly if there were untried alternatives prior
15        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
16        backtracked to try for "b" instead of moving to the next alternative branch
17        at the same level (in this case, to look for "c"). The Perl documentation
18        is clear that when (*THEN) is backtracked onto, it goes to the "next
19        alternative in the innermost enclosing group".
21    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
22        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
23        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
24        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
25        (*THEN).
27    3.  If \s appeared in a character class, it removed the VT character from
28        the class, even if it had been included by some previous item, for example
29        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
30        of \s, but is part of the POSIX "space" class.)
32    4.  A partial match never returns an empty string (because you can always
33        match an empty string at the end of the subject); however the checking for
34        an empty string was starting at the "start of match" point. This has been
35        changed to the "earliest inspected character" point, because the returned
36        data for a partial match starts at this character. This means that, for
37        example, /(?<=abc)def/ gives a partial match for the subject "abc"
38        (previously it gave "no match").
40    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
41        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
42        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
43        has an implication that the given string is incomplete (because a partial
44        match is preferred over a full match). For this reason, these items now
45        give a partial match in this situation. [Aside: previously, the one case
46        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
47        match rather than a full match, which was wrong by the old rules, but is
48        now correct.]
50    6.  There was a bug in the handling of #-introduced comments, recognized when
51        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
52        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
53        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
54        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
55        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
56        places in pcre_compile().
58    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
59        comments when looking ahead for named forward references to subpatterns,
60        the only newline sequence it recognized was NL. It now handles newlines
61        according to the set newline convention.
63    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
64        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
65        cater for a lack of strerror(). These oversights have been fixed.
67    9.  Added --match-limit and --recursion-limit to pcregrep.
69    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
71    11. When the -o option was used, pcregrep was setting a return code of 1, even
72        when matches were found, and --line-buffered was not being honoured.
74    12. Added an optional parentheses number to the -o and --only-matching options
75        of pcregrep.
77    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
78        can match an empty string. The code to do it in pcretest and pcredemo
79        needed fixing:
81        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
82            only one byte after an empty string match just before CRLF (this case
83            just got forgotten; "any" and "anycrlf" were OK).
85        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
86            mode when an empty string match preceded an ASCII character followed by
87            a non-ASCII character. (The code for advancing by one character rather
88            than one byte was nonsense.)
90        (c) The pcredemo.c sample program did not have any code at all to handle
91            the cases when CRLF is a valid newline sequence.
93    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
94        as a starting offset was within the subject string. There is now a new
95        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
96        negative or greater than the length of the string. In order to test this,
97        pcretest is extended to allow the setting of negative starting offsets.
99    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
100        starting offset points to the beginning of a UTF-8 character was
101        unnecessarily clumsy. I tidied it up.
103    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
104        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
106    17. Nobody had reported that the --include_dir option, which was added in
107        release 7.7 should have been called --include-dir (hyphen, not underscore)
108        for compatibility with GNU grep. I have changed it to --include-dir, but
109        left --include_dir as an undocumented synonym, and the same for
110        --exclude-dir, though that is not available in GNU grep, at least as of
111        release 2.5.4.
113    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
114        characters from a string of bytes) have been redefined so as not to use
115        loops, in order to improve performance in some environments. At the same
116        time, I abstracted some of the common code into auxiliary macros to save
117        repetition (this should not affect the compiled code).
119    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
120        compile-time error is now given if \c is not followed by an ASCII
121        character, that is, a byte less than 128. (In EBCDIC mode, the code is
122        different, and any byte value is allowed.)
124    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
125        START_OPTIMIZE option, which is now allowed at compile time - but just
126        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
127        to pcregrep and other applications that have no direct access to PCRE
128        options. The new /Y option in pcretest sets this option when calling
129        pcre_compile().
131    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
132        back references. Groups containing recursive back references were forced to
133        be atomic by that change, but in the case of named groups, the amount of
134        memory required was incorrectly computed, leading to "Failed: internal
135        error: code overflow". This has been fixed.
137    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
138        pcretest.c, to avoid build problems in some Borland environments.
141    Version 8.10 25-Jun-2010
142    ------------------------
144    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
145        THEN.
147    2.  (*ACCEPT) was not working when inside an atomic group.
149    3.  Inside a character class, \B is treated as a literal by default, but
150        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
151        causes the error). The code is unchanged, but I tidied the documentation.
153    4.  Inside a character class, PCRE always treated \R and \X as literals,
154        whereas Perl faults them if its -w option is set. I have changed PCRE so
155        that it faults them when PCRE_EXTRA is set.
157    5.  Added support for \N, which always matches any character other than
158        newline. (It is the same as "." when PCRE_DOTALL is not set.)
160    6.  When compiling pcregrep with newer versions of gcc which may have
161        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
162        declared with attribute warn_unused_result" were given. Just casting the
163        result to (void) does not stop the warnings; a more elaborate fudge is
164        needed. I've used a macro to implement this.
166    7.  Minor change to pcretest.c to avoid a compiler warning.
168    8.  Added four artifical Unicode properties to help with an option to make
169        \s etc use properties (see next item). The new properties are: Xan
170        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
172    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
173        use Unicode properties. (*UCP) at the start of a pattern can be used to set
174        this option. Modified pcretest to add /W to test this facility. Added
175        REG_UCP to make it available via the POSIX interface.
177    10. Added --line-buffered to pcregrep.
179    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
180        studied, and the match started with a letter with a code point greater than
181        127 whose first byte was different to the first byte of the other case of
182        the letter, the other case of this starting letter was not recognized
183        (#976).
185    12. If a pattern that was studied started with a repeated Unicode property
186        test, for example, \p{Nd}+, there was the theoretical possibility of
187        setting up an incorrect bitmap of starting bytes, but fortunately it could
188        not have actually happened in practice until change 8 above was made (it
189        added property types that matched character-matching opcodes).
191    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
192        possible starting bytes for non-anchored patterns.
194    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
195        \R, and also a number of cases that involve Unicode properties, both
196        explicit and implicit when PCRE_UCP is set.
198    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
199        input, it could crash or give wrong results if characters with values
200        greater than 0xc0 were present in the subject string. (Detail: it assumed
201        UTF-8 input when processing these items.)
203    16. Added a lot of (int) casts to avoid compiler warnings in systems where
204        size_t is 64-bit (#991).
206    17. Added a check for running out of memory when PCRE is compiled with
207        --disable-stack-for-recursion (#990).
209    18. If the last data line in a file for pcretest does not have a newline on
210        the end, a newline was missing in the output.
212    19. The default pcre_chartables.c file recognizes only ASCII characters (values
213        less than 128) in its various bitmaps. However, there is a facility for
214        generating tables according to the current locale when PCRE is compiled. It
215        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
216        characters, are recognized by isspace() and therefore were getting set in
217        these tables, and indeed these tables seem to approximate to ISO 8859. This
218        caused a problem in UTF-8 mode when pcre_study() was used to create a list
219        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
220        which of course cannot start UTF-8 characters. I have changed the code so
221        that only real ASCII characters (less than 128) and the correct starting
222        bytes for UTF-8 encodings are set for characters greater than 127 when in
223        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
224        altogether.)
226    20. Added the /T option to pcretest so as to be able to run tests with non-
227        standard character tables, thus making it possible to include the tests
228        used for 19 above in the standard set of tests.
230    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
231        reference to a subpattern the other side of a comment that contains an
232        opening parenthesis caused either an internal compiling error, or a
233        reference to the wrong subpattern.
236    Version 8.02 19-Mar-2010
237    ------------------------
239    1.  The Unicode data tables have been updated to Unicode 5.2.0.
241    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
242        configured.
244    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
245        original author of that file, following a query about its status.
247    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
248        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
250    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
251        quantifier applied to a forward-referencing subroutine call, could compile
252        incorrect code or give the error "internal error: previously-checked
253        referenced subpattern not found".
255    6.  Both MS Visual Studio and Symbian OS have problems with initializing
256        variables to point to external functions. For these systems, therefore,
257        pcre_malloc etc. are now initialized to local functions that call the
258        relevant global functions.
260    7.  There were two entries missing in the vectors called coptable and poptable
261        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
262        I've fixed the data, and added a kludgy way of testing at compile time that
263        the lengths are correct (equal to the number of opcodes).
265    8.  Following on from 7, I added a similar kludge to check the length of the
266        eint vector in pcreposix.c.
268    9.  Error texts for pcre_compile() are held as one long string to avoid too
269        much relocation at load time. To find a text, the string is searched,
270        counting zeros. There was no check for running off the end of the string,
271        which could happen if a new error number was added without updating the
272        string.
274    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
276    11. \K was not working if it appeared in an atomic group or in a group that
277        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
278        \K is "not well defined" if used in an assertion. PCRE now accepts it if
279        the assertion is positive, but not if it is negative.
281    12. Change 11 fortuitously reduced the size of the stack frame used in the
282        "match()" function of pcre_exec.c by one pointer. Forthcoming
283        implementation of support for (*MARK) will need an extra pointer on the
284        stack; I have reserved it now, so that the stack frame size does not
285        decrease.
287    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
288        item in branch that calls a recursion is a subroutine call - as in the
289        second branch in the above example - was incorrectly given the compile-
290        time error "recursive call could loop indefinitely" because pcre_compile()
291        was not correctly checking the subroutine for matching a non-empty string.
293    14. The checks for overrunning compiling workspace could trigger after an
294        overrun had occurred. This is a "should never occur" error, but it can be
295        triggered by pathological patterns such as hundreds of nested parentheses.
296        The checks now trigger 100 bytes before the end of the workspace.
298    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
301    Version 8.01 19-Jan-2010
302    ------------------------
304    1.  If a pattern contained a conditional subpattern with only one branch (in
305        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
306        computed the wrong minimum data length (which is of course zero for such
307        subpatterns). This could cause incorrect "no match" results.
309    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
310        the pattern is reset in the first branch, pcre_compile() failed with
311        "internal error: code overflow at offset...". This happened only when
312        the reset was to the original external option setting. (An optimization
313        abstracts leading options settings into an external setting, which was the
314        cause of this.)
316    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
317        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
318        assertion pattern did not match (meaning that the assertion was true), it
319        was incorrectly treated as false if the SKIP had been reached during the
320        matching. This also applied to assertions used as conditions.
322    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
323        assertion subpattern, including such a pattern used as a condition,
324        unpredictable results occurred, instead of the error return
327    5.  The C++ GlobalReplace function was not working like Perl for the special
328        situation when an empty string is matched. It now does the fancy magic
329        stuff that is necessary.
331    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
332        removed. (These were left over from very, very early versions of PCRE.)
334    7.  Some cosmetic changes to the code to make life easier when compiling it
335        as part of something else:
337        (a) Change DEBUG to PCRE_DEBUG.
339        (b) In pcre_compile(), rename the member of the "branch_chain" structure
340            called "current" as "current_branch", to prevent a collision with the
341            Linux macro when compiled as a kernel module.
343        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
344            prevent a collision with the Linux macro when compiled as a kernel
345            module.
347    8.  In pcre_compile() there are some checks for integer overflows that used to
348        cast potentially large values to (double). This has been changed to that
349        when building, a check for int64_t is made, and if it is found, it is used
350        instead, thus avoiding the use of floating point arithmetic. (There is no
351        other use of FP in PCRE.) If int64_t is not found, the fallback is to
352        double.
354    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
355        2005 (difference between two addresses compared to an unsigned value).
357    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
358        custom one, because of the following reported problem in Windows:
360          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
361              under Win32.
362          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
363              therefore missing the function definition.
364          - The compiler thus generates a "C" signature for the test function.
365          - The linker fails to find the "C" function.
366          - PCRE fails to configure if asked to do so against libbz2.
368    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
369        messages were output:
371          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
372          rerunning libtoolize, to keep the correct libtool macros in-tree.
373          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
375        I have done both of these things.
377    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
378        most of the time, it *can* run out if it is given a pattern that contains a
379        runaway infinite recursion. I updated the discussion in the pcrestack man
380        page.
382    13. Now that we have gone to the x.xx style of version numbers, the minor
383        version may start with zero. Using 08 or 09 is a bad idea because users
384        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
385        interpreted as invalid octal numbers. I've updated the previous comment in
386        configure.ac, and also added a check that gives an error if 08 or 09 are
387        used.
389    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
390        causing partial matching to fail when the end of the subject matched \W
391        in a UTF-8 pattern where \W was quantified with a minimum of 3.
393    15. There were some discrepancies between the declarations in pcre_internal.h
394        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
395        their definitions. The declarations used "const uschar *" and the
396        definitions used USPTR. Even though USPTR is normally defined as "const
397        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
398        reported that: "This difference in casting confuses some C++ compilers, for
399        example, SunCC recognizes above declarations as different functions and
400        generates broken code for hbpcre." I have changed the declarations to use
401        USPTR.
403    16. GNU libtool is named differently on some systems. The autogen.sh script now
404        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
405        (FreeBSD).
407    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
408        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
409        comment: "Figure out how to create a longlong from a string: strtoll and
410        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
411        instance, but it only takes 2 args instead of 3!"
413    18. A subtle bug concerned with back references has been fixed by a change of
414        specification, with a corresponding code fix. A pattern such as
415        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
416        refers, was giving matches when it shouldn't. For example, xa=xaaa would
417        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
418        same bug. Such groups have to be quantified to be useful, or contained
419        inside another quantified group. (If there's no repetition, the reference
420        can never match.) The problem arises because, having left the group and
421        moved on to the rest of the pattern, a later failure that backtracks into
422        the group uses the captured value from the final iteration of the group
423        rather than the correct earlier one. I have fixed this in PCRE by forcing
424        any group that contains a reference to itself to be an atomic group; that
425        is, there cannot be any backtracking into it once it has completed. This is
426        similar to recursive and subroutine calls.
429    Version 8.00 19-Oct-09
430    ----------------------
432    1.  The table for translating pcre_compile() error codes into POSIX error codes
433        was out-of-date, and there was no check on the pcre_compile() error code
434        being within the table. This could lead to an OK return being given in
435        error.
437    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
438        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
439        in a Windows environment.
441    3.  The pcregrep --count option prints the count for each file even when it is
442        zero, as does GNU grep. However, pcregrep was also printing all files when
443        --files-with-matches was added. Now, when both options are given, it prints
444        counts only for those files that have at least one match. (GNU grep just
445        prints the file name in this circumstance, but including the count seems
446        more useful - otherwise, why use --count?) Also ensured that the
447        combination -clh just lists non-zero counts, with no names.
449    4.  The long form of the pcregrep -F option was incorrectly implemented as
450        --fixed_strings instead of --fixed-strings. This is an incompatible change,
451        but it seems right to fix it, and I didn't think it was worth preserving
452        the old behaviour.
454    5.  The command line items --regex=pattern and --regexp=pattern were not
455        recognized by pcregrep, which required --regex pattern or --regexp pattern
456        (with a space rather than an '='). The man page documented the '=' forms,
457        which are compatible with GNU grep; these now work.
459    6.  No libpcreposix.pc file was created for pkg-config; there was just
460        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
462    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
463        when UCP support is not needed, by modifying the Python script that
464        generates it from Unicode data files. This should not matter if the module
465        is correctly used as a library, but I received one complaint about 50K of
466        unwanted data. My guess is that the person linked everything into his
467        program rather than using a library. Anyway, it does no harm.
469    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
470        was a minimum greater than 1 for a wide character in a possessive
471        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
472        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
473        character. Chaos in the form of incorrect output or a compiling loop could
474        result.
476    9.  The restrictions on what a pattern can contain when partial matching is
477        requested for pcre_exec() have been removed. All patterns can now be
478        partially matched by this function. In addition, if there are at least two
479        slots in the offset vector, the offset of the earliest inspected character
480        for the match and the offset of the end of the subject are set in them when
481        PCRE_ERROR_PARTIAL is returned.
483    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
484        synonymous with PCRE_PARTIAL, for backwards compatibility, and
485        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
486        and may be more useful for multi-segment matching.
488    11. Partial matching with pcre_exec() is now more intuitive. A partial match
489        used to be given if ever the end of the subject was reached; now it is
490        given only if matching could not proceed because another character was
491        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
492        string "Z", which now yields "no match" instead of "partial match". In the
493        case of pcre_dfa_exec(), "no match" is given if every matching path for the
494        final character ended with (*FAIL).
496    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
497        if the pattern had a "must contain" character that was already found in the
498        earlier partial match, unless partial matching was again requested. For
499        example, with the pattern /dog.(body)?/, the "must contain" character is
500        "g". If the first part-match was for the string "dog", restarting with
501        "sbody" failed. This bug has been fixed.
503    13. The string returned by pcre_dfa_exec() after a partial match has been
504        changed so that it starts at the first inspected character rather than the
505        first character of the match. This makes a difference only if the pattern
506        starts with a lookbehind assertion or \b or \B (\K is not supported by
507        pcre_dfa_exec()). It's an incompatible change, but it makes the two
508        matching functions compatible, and I think it's the right thing to do.
510    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
511        so that the demonstration program is easily available in environments where
512        PCRE has not been installed from source.
514    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
515        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
516        library.
518    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
519        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
520        is not the first non-POSIX option to be added. Clearly some people find
521        these options useful.
523    17. If a caller to the POSIX matching function regexec() passes a non-zero
524        value for nmatch with a NULL value for pmatch, the value of
525        nmatch is forced to zero.
527    18. RunGrepTest did not have a test for the availability of the -u option of
528        the diff command, as RunTest does. It now checks in the same way as
529        RunTest, and also checks for the -b option.
531    19. If an odd number of negated classes containing just a single character
532        interposed, within parentheses, between a forward reference to a named
533        subpattern and the definition of the subpattern, compilation crashed with
534        an internal error, complaining that it could not find the referenced
535        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
536        [The bug was that it was starting one character too far in when skipping
537        over the character class, thus treating the ] as data rather than
538        terminating the class. This meant it could skip too much.]
540    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
541        /g option in pcretest when the pattern contains \K, which makes it possible
542        to have an empty string match not at the start, even when the pattern is
543        anchored. Updated pcretest and pcredemo to use this option.
545    21. If the maximum number of capturing subpatterns in a recursion was greater
546        than the maximum at the outer level, the higher number was returned, but
547        with unset values at the outer level. The correct (outer level) value is
548        now given.
550    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
551        PCRE did not set those parentheses (unlike Perl). I have now found a way to
552        make it do so. The string so far is captured, making this feature
553        compatible with Perl.
555    23. The tests have been re-organized, adding tests 11 and 12, to make it
556        possible to check the Perl 5.10 features against Perl 5.10.
558    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
559        pattern matches a fixed length string. PCRE did not allow this; now it
560        does. Neither allows recursion.
562    25. I finally figured out how to implement a request to provide the minimum
563        length of subject string that was needed in order to match a given pattern.
564        (It was back references and recursion that I had previously got hung up
565        on.) This code has now been added to pcre_study(); it finds a lower bound
566        to the length of subject needed. It is not necessarily the greatest lower
567        bound, but using it to avoid searching strings that are too short does give
568        some useful speed-ups. The value is available to calling programs via
569        pcre_fullinfo().
571    26. While implementing 25, I discovered to my embarrassment that pcretest had
572        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
573        study optimizations had never been tested with that matching function.
574        Oops. What is worse, even when it was passed study data, there was a bug in
575        pcre_dfa_exec() that meant it never actually used it. Double oops. There
576        were also very few tests of studied patterns with pcre_dfa_exec().
578    27. If (?| is used to create subpatterns with duplicate numbers, they are now
579        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
580        on the other side of the coin, they are no longer allowed to have different
581        names, because these cannot be distinguished in PCRE, and this has caused
582        confusion. (This is a difference from Perl.)
584    28. When duplicate subpattern names are present (necessarily with different
585        numbers, as required by 27 above), and a test is made by name in a
586        conditional pattern, either for a subpattern having been matched, or for
587        recursion in such a pattern, all the associated numbered subpatterns are
588        tested, and the overall condition is true if the condition is true for any
589        one of them. This is the way Perl works, and is also more like the way
590        testing by number works.
593    Version 7.9 11-Apr-09
594    ---------------------
596    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
597        (pcretest), all targets were linked against these libraries. This included
598        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
599        libraries. This caused unwanted dependencies to be created. This problem
600        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
601        pcretest is linked with readline.
603    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
604        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
605        moved outside it again, because FALSE and TRUE are already defined in AIX,
606        but BOOL is not.
608    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
609        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
611    4.  The pcregrep documentation said spaces were inserted as well as colons (or
612        hyphens) following file names and line numbers when outputting matching
613        lines. This is not true; no spaces are inserted. I have also clarified the
614        wording for the --colour (or --color) option.
616    5.  In pcregrep, when --colour was used with -o, the list of matching strings
617        was not coloured; this is different to GNU grep, so I have changed it to be
618        the same.
620    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
621        each matching line was coloured. Now it goes on to look for further matches
622        of any of the test patterns, which is the same behaviour as GNU grep.
624    7.  A pattern that could match an empty string could cause pcregrep to loop; it
625        doesn't make sense to accept an empty string match in pcregrep, so I have
626        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
627        seems to be how GNU grep behaves.
629    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
630        start or after a newline", because the conditional assertion was not being
631        correctly handled. The rule now is that both the assertion and what follows
632        in the first alternative must satisfy the test.
634    9.  If auto-callout was enabled in a pattern with a conditional group whose
635        condition was an assertion, PCRE could crash during matching, both with
636        pcre_exec() and pcre_dfa_exec().
638    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
639        used for matching.
641    11. Unicode property support in character classes was not working for
642        characters (bytes) greater than 127 when not in UTF-8 mode.
644    12. Added the -M command line option to pcretest.
646    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
648    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
650    16. Added comments and documentation about mis-use of no_arg in the C++
651        wrapper.
653    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
654        from Martin Jerabek that uses macro names for all relevant character and
655        string constants.
657    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
658        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
659        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
660        these, but not everybody uses configure.
662    19. A conditional group that had only one branch was not being correctly
663        recognized as an item that could match an empty string. This meant that an
664        enclosing group might also not be so recognized, causing infinite looping
665        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
666        with the subject "ab", where knowledge that the repeated group can match
667        nothing is needed in order to break the loop.
669    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
670        exec(), but without supplying a callout function, matching went wrong.
672    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
673        leak if the size of the offset vector was greater than 30. When the vector
674        is smaller, the saved offsets during recursion go onto a local stack
675        vector, but for larger vectors malloc() is used. It was failing to free
676        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
677        error, in fact).
679    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
680        heapframe that is used only when UTF-8 support is enabled. This caused no
681        problem, but was untidy.
683    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
684        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
685        included within another project.
687    24. Steven Van Ingelgem's patches to add more options to the CMake support,
688        slightly modified by me:
690          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
691              not building pcregrep.
693          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
694              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
696    25. Forward references, both numeric and by name, in patterns that made use of
697        duplicate group numbers, could behave incorrectly or give incorrect errors,
698        because when scanning forward to find the reference group, PCRE was not
699        taking into account the duplicate group numbers. A pattern such as
700        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
702    26. Changed a few more instances of "const unsigned char *" to USPTR, making
703        the feature of a custom pointer more persuasive (as requested by a user).
705    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
706        pcretest.c, inside #ifndefs, because it seems they are sometimes already
707        pre-defined.
709    28. Added support for (*UTF8) at the start of a pattern.
711    29. Arrange for flags added by the "release type" setting in CMake to be shown
712        in the configuration summary.
715    Version 7.8 05-Sep-08
716    ---------------------
718    1.  Replaced UCP searching code with optimized version as implemented for Ad
719        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
720        stage table and inline lookup instead of a function, giving speed ups of 2
721        to 5 times on some simple patterns that I tested. Permission was given to
722        distribute the MultiStage2.py script that generates the tables (it's not in
723        the tarball, but is in the Subversion repository).
725    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
726        scripts.
728    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
729        a group with a zero qualifier. The result of the study could be incorrect,
730        or the function might crash, depending on the pattern.
732    4.  Caseless matching was not working for non-ASCII characters in back
733        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
734        It now works when Unicode Property Support is available.
736    5.  In pcretest, an escape such as \x{de} in the data was always generating
737        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
738        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
739        truncation.
741    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
743    7.  Added two (int) casts to pcregrep when printing the difference of two
744        pointers, in case they are 64-bit values.
746    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
747        test 2 if it fails.
749    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
750        and a #define of that name to empty if it is not externally set. This is to
751        allow users of MSVC to set it if necessary.
753    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
754        the convenience functions in the pcre_get.c source file.
756    11. An option change at the start of a pattern that had top-level alternatives
757        could cause overwriting and/or a crash. This command provoked a crash in
758        some environments:
760          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
762        This potential security problem was recorded as CVE-2008-2371.
764    12. For a pattern where the match had to start at the beginning or immediately
765        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
766        pcre_dfa_exec() could read past the end of the passed subject if there was
767        no match. To help with detecting such bugs (e.g. with valgrind), I modified
768        pcretest so that it places the subject at the end of its malloc-ed buffer.
770    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
771        exec() might read past the end of the data buffer in UTF-8 mode.
773    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
774        the data contained the byte 0x85 as part of a UTF-8 character within its
775        first line. This applied both to normal and DFA matching.
777    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
778        /^[^d]*?$/8 failed to match "abc".
780    16. Added a missing copyright notice to pcrecpp_internal.h.
782    17. Make it more clear in the documentation that values returned from
783        pcre_exec() in ovector are byte offsets, not character counts.
785    18. Tidied a few places to stop certain compilers from issuing warnings.
787    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
788        supplied by Stefan Weber. I made a further small update for 7.8 because
789        there is a change of source arrangements: the pcre_searchfuncs.c module is
790        replaced by pcre_ucd.c.
793    Version 7.7 07-May-08
794    ---------------------
796    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
797        a string to a long long, pretend we don't even have a long long." This is
798        done by checking for the strtoq, strtoll, and _strtoi64 functions.
800    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
801        pre-7.6 versions, which defined a global no_arg variable instead of putting
802        it in the RE class. (See also #8 below.)
804    3.  Remove a line of dead code, identified by coverity and reported by Nuno
805        Lopes.
807    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
809        (1) The include/exclude patterns were being applied to the whole pathnames
810            of files, instead of just to the final components.
812        (2) If there was more than one level of directory, the subdirectories were
813            skipped unless they satisfied the include/exclude conditions. This is
814            inconsistent with GNU grep (and could even be seen as contrary to the
815            pcregrep specification - which I improved to make it absolutely clear).
816            The action now is always to scan all levels of directory, and just
817            apply the include/exclude patterns to regular files.
819    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
820        --exclude_dir in the tests to avoid scanning .svn directories.
822    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
823        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
824        doesn't support NULs in patterns.
826    7.  Added some missing "const"s to declarations of static tables in
827        pcre_compile.c and pcre_dfa_exec.c.
829    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
830        caused by fix #2  above. (Subsequently also a second patch to fix the
831        first patch. And a third patch - this was a messy problem.)
833    9.  Applied Craig's patch to remove the use of push_back().
835    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
836        matching function regexec().
838    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
839        which, however, unlike Perl's \g{...}, are subroutine calls, not back
840        references. PCRE supports relative numbers with this syntax (I don't think
841        Oniguruma does).
843    12. Previously, a group with a zero repeat such as (...){0} was completely
844        omitted from the compiled regex. However, this means that if the group
845        was called as a subroutine from elsewhere in the pattern, things went wrong
846        (an internal error was given). Such groups are now left in the compiled
847        pattern, with a new opcode that causes them to be skipped at execution
848        time.
850    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
851        to the way PCRE behaves:
853        (a) A lone ] character is dis-allowed (Perl treats it as data).
855        (b) A back reference to an unmatched subpattern matches an empty string
856            (Perl fails the current match path).
858        (c) A data ] in a character class must be notated as \] because if the
859            first data character in a class is ], it defines an empty class. (In
860            Perl it is not possible to have an empty class.) The empty class []
861            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
862            The negative empty class [^] matches any one character, independently
863            of the DOTALL setting.
865    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
866        non-existent subpattern following a character class starting with ']' and
867        containing () gave an internal compiling error instead of "reference to
868        non-existent subpattern". Fortunately, when the pattern did exist, the
869        compiled code was correct. (When scanning forwards to check for the
870        existencd of the subpattern, it was treating the data ']' as terminating
871        the class, so got the count wrong. When actually compiling, the reference
872        was subsequently set up correctly.)
874    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
875        it was being rejected as not supported by pcre_dfa_exec(), even though
876        other assertions are supported. I have made pcre_dfa_exec() support
877        (*FAIL).
879    16. The implementation of 13c above involved the invention of a new opcode,
880        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
881        cannot be changed at match time, I realized I could make a small
882        improvement to matching performance by compiling OP_ALLANY instead of
883        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
884        on the OP_ANY path.
886    17. Compiling pcretest on Windows with readline support failed without the
887        following two fixes: (1) Make the unistd.h include conditional on
888        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
890    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
891        ncurses library to be included for pcretest when ReadLine support is
892        requested, but also to allow for it to be overridden. This patch came from
893        Daniel Bergström.
895    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
896        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
897        any errors with the current Unicode tables. Thanks to Peter Kankowski for
898        spotting this.
901    Version 7.6 28-Jan-08
902    ---------------------
904    1.  A character class containing a very large number of characters with
905        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
906        overflow.
908    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
909        HAVE_LONG_LONG is not defined.
911    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
912        bring them up to date and include new features. This patch includes:
914        - Fixed PH's badly added libz and libbz2 support.
915        - Fixed a problem with static linking.
916        - Added pcredemo. [But later removed - see 7 below.]
917        - Fixed dftables problem and added an option.
918        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
919            HAVE_LONG_LONG.
920        - Added readline support for pcretest.
921        - Added an listing of the option settings after cmake has run.
923    4.  A user submitted a patch to Makefile that makes it easy to create
924        "pcre.dll" under mingw when using Configure/Make. I added stuff to
925        Makefile.am that cause it to include this special target, without
926        affecting anything else. Note that the same mingw target plus all
927        the other distribution libraries and programs are now supported
928        when configuring with CMake (see 6 below) instead of with
929        Configure/Make.
931    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
932        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
933        exported in the Windows port". It has not yet been confirmed that the patch
934        solves the problem, but it does no harm.
936    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
937        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
938        with CMake, and also correct the comment about stack recursion.
940    7.  Remove the automatic building of pcredemo from the ./configure system and
941        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
942        of a program that users should build themselves after PCRE is installed, so
943        building it automatically is not really right. What is more, it gave
944        trouble in some build environments.
946    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
949    Version 7.5 10-Jan-08
950    ---------------------
952    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
953        values in parens when parsing an RE using the C++ wrapper."
955    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
956        Characters greater than 255 were excluded from the class instead of being
957        included.
959    3.  The same bug as (2) above applied to negated POSIX classes such as
960        [:^space:].
962    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
963        defined or documented. It seems to have been a typo for PCRE_STATIC, so
964        I have changed it.
966    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
967        first named subpattern) and a construct such as (?&a) would reference the
968        first named subpattern whose name started with "a" (in other words, the
969        length check was missing). Both these problems are fixed. "Subpattern name
970        expected" is now given for (?&) (a zero-length name), and this patch also
971        makes it give the same error for \k'' (previously it complained that that
972        was a reference to a non-existent subpattern).
974    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
975        this is right because (?- can be followed by option settings as well as by
976        digits. I have, however, made the messages clearer.
978    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
979        than the number used in the conditional) now cause a compile-time error.
980        This is actually not compatible with Perl, which accepts such patterns, but
981        treats the conditional as always being FALSE (as PCRE used to), but it
982        seems to me that giving a diagnostic is better.
984    8.  Change "alphameric" to the more common word "alphanumeric" in comments
985        and messages.
987    9.  Fix two occurrences of "backslash" in comments that should have been
988        "backspace".
990    10. Remove two redundant lines of code that can never be obeyed (their function
991        was moved elsewhere).
993    11. The program that makes PCRE's Unicode character property table had a bug
994        which caused it to generate incorrect table entries for sequences of
995        characters that have the same character type, but are in different scripts.
996        It amalgamated them into a single range, with the script of the first of
997        them. In other words, some characters were in the wrong script. There were
998        thirteen such cases, affecting characters in the following ranges:
1000          U+002b0 - U+002c1
1001          U+0060c - U+0060d
1002          U+0061e - U+00612
1003          U+0064b - U+0065e
1004          U+0074d - U+0076d
1005          U+01800 - U+01805
1006          U+01d00 - U+01d77
1007          U+01d9b - U+01dbf
1008          U+0200b - U+0200f
1009          U+030fc - U+030fe
1010          U+03260 - U+0327f
1011          U+0fb46 - U+0fbb1
1012          U+10450 - U+1049d
1014    12. The -o option (show only the matching part of a line) for pcregrep was not
1015        compatible with GNU grep in that, if there was more than one match in a
1016        line, it showed only the first of them. It now behaves in the same way as
1017        GNU grep.
1019    13. If the -o and -v options were combined for pcregrep, it printed a blank
1020        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1021        does the same. The return code can be used to tell if there were any
1022        non-matching lines.
1024    14. Added --file-offsets and --line-offsets to pcregrep.
1026    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1027        infinitely looping recursion. The bug was that positive lookaheads were not
1028        being skipped when checking for a possible empty match (negative lookaheads
1029        and both kinds of lookbehind were skipped).
1031    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1032        inclusion of <windows.h> to before rather than after the definition of
1033        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1035    17. Specifying a possessive quantifier with a specific limit for a Unicode
1036        character property caused pcre_compile() to compile bad code, which led at
1037        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1038        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1039        caused the error; without that there was no problem.
1041    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1043    19. Added --enable-pcretest-libreadline.
1045    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1046        RE::GlobalReplace(). As a result, the number of replacements returned was
1047        double what it should be. I removed one of the increments, but Craig sent a
1048        later patch that removed the other one (the right fix) and added unit tests
1049        that check the return values (which was not done before).
1051    21. Several CMake things:
1053        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1054            the names libpcre and libpcreposix, not just pcre and pcreposix.
1056        (2) The above change means that pcretest and pcregrep are now correctly
1057            linked with the newly-built libraries, not previously installed ones.
1061    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1062        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1063        UTF-8 newline character). The key issue is that the pattern starts .*;
1064        this means that the match must be either at the beginning, or after a
1065        newline. The bug was in the code for advancing after a failed match and
1066        checking that the new position followed a newline. It was not taking
1067        account of UTF-8 characters correctly.
1069    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1070        character classes. PCRE was not treating the sequence [:...:] as a
1071        character class unless the ... were all letters. Perl, however, seems to
1072        allow any characters between [: and :], though of course it rejects as
1073        unknown any "names" that contain non-letters, because all the known class
1074        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1075        for example, whereas PCRE did not - it did not recognize a POSIX character
1076        class. This seemed a bit dangerous, so the code has been changed to be
1077        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1078        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1079        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1080        Perl does, and where it didn't before.
1082    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1083        Windows environments %n is disabled by default.
1086    Version 7.4 21-Sep-07
1087    ---------------------
1089    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1090        means that a class such as [\s] counted as "explicit reference to CR or
1091        LF". That isn't really right - the whole point of the change was to try to
1092        help when there was an actual mention of one of the two characters. So now
1093        the change happens only if \r or \n (or a literal CR or LF) character is
1094        encountered.
1096    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1097        of both had grown to the point where there were only 3 bits left.
1098        Fortunately, there was spare space in the data structure, and so I have
1099        moved the internal flags into a new 16-bit field to free up more option
1100        bits.
1102    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1103        but did not set the internal JCHANGED flag - either of these is enough to
1104        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1105        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1106        start sets both bits.
1108    4.  Added options (at build time, compile time, exec time) to change \R from
1109        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1111    5.  doc/pcresyntax.html was missing from the distribution.
1113    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1114        compatibility, even though it is no longer used.
1116    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1117        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1118        windows.h file is present (where different names are used). [This was
1119        reversed later after testing - see 16 below.]
1121    8.  Changed all #include <config.h> to #include "config.h". There were also
1122        some further <pcre.h> cases that I changed to "pcre.h".
1124    9.  When pcregrep was used with the --colour option, it missed the line ending
1125        sequence off the lines that it output.
1127    10. It was pointed out to me that arrays of string pointers cause lots of
1128        relocations when a shared library is dynamically loaded. A technique of
1129        using a single long string with a table of offsets can drastically reduce
1130        these. I have refactored PCRE in four places to do this. The result is
1131        dramatic:
1133          Originally:                          290
1134          After changing UCP table:            187
1135          After changing error message table:   43
1136          After changing table of "verbs"       36
1137          After changing table of Posix names   22
1139        Thanks to the folks working on Gregex for glib for this insight.
1141    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1142        unicode-properties was also set.
1144    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1146    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1147        checked only for CRLF.
1149    14. Added casts to pcretest.c to avoid compiler warnings.
1151    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1153    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1154        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1155        entirely. This removes changes made in 7 above.
1157    17. The CMake files have been updated, and there is now more information about
1158        building with CMake in the NON-UNIX-USE document.
1161    Version 7.3 28-Aug-07
1162    ---------------------
1164     1. In the rejigging of the build system that eventually resulted in 7.1, the
1165        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1166        brackets there is not right, since it causes compilers to look for an
1167        installed pcre.h, not the version that is in the source that is being
1168        compiled (which of course may be different). I have changed it back to:
1170          #include "pcre.h"
1172        I have a vague recollection that the change was concerned with compiling in
1173        different directories, but in the new build system, that is taken care of
1174        by the VPATH setting the Makefile.
1176     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1177        when the subject happened to end in the byte 0x85 (e.g. if the last
1178        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1179        characters but of course it shouldn't be taken as a newline when it is part
1180        of another character. The bug was that, for an unlimited repeat of . in
1181        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1182        characters when looking for a newline.
1184     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1186     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1187        in debug output.
1189     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1190        long printing in the pcrecpp unittest when running under MinGW.
1192     6. ESC_K was left out of the EBCDIC table.
1194     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1195        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1196        limit also applies to "virtual nesting" when a pattern is recursive, and in
1197        this case 1000 isn't so big. I have been able to remove this limit at the
1198        expense of backing off one optimization in certain circumstances. Normally,
1199        when pcre_exec() would call its internal match() function recursively and
1200        immediately return the result unconditionally, it uses a "tail recursion"
1201        feature to save stack. However, when a subpattern that can match an empty
1202        string has an unlimited repetition quantifier, it no longer makes this
1203        optimization. That gives it a stack frame in which to save the data for
1204        checking that an empty string has been matched. Previously this was taken
1205        from the 1000-entry workspace that had been reserved. So now there is no
1206        explicit limit, but more stack is used.
1208     8. Applied Daniel's patches to solve problems with the import/export magic
1209        syntax that is required for Windows, and which was going wrong for the
1210        pcreposix and pcrecpp parts of the library. These were overlooked when this
1211        problem was solved for the main library.
1213     9. There were some crude static tests to avoid integer overflow when computing
1214        the size of patterns that contain repeated groups with explicit upper
1215        limits. As the maximum quantifier is 65535, the maximum group length was
1216        set at 30,000 so that the product of these two numbers did not overflow a
1217        32-bit integer. However, it turns out that people want to use groups that
1218        are longer than 30,000 bytes (though not repeat them that many times).
1219        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1220        made it possible to implement the integer overflow checks in a much more
1221        dynamic way, which I have now done. The artificial limitation on group
1222        length has been removed - we now have only the limit on the total length of
1223        the compiled pattern, which depends on the LINK_SIZE setting.
1225    10. Fixed a bug in the documentation for get/copy named substring when
1226        duplicate names are permitted. If none of the named substrings are set, the
1227        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1228        empty string.
1230    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1231        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1232        because the ] is interpreted as the first data character and the
1233        terminating ] is not found. PCRE has been made compatible with Perl in this
1234        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1235        cause memory overwriting.
1237    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1238        string has been matched (to stop an infinite loop). It was not recognizing
1239        a conditional subpattern that could match an empty string if that
1240        subpattern was within another subpattern. For example, it looped when
1241        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1242        condition was not nested. This bug has been fixed.
1244    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1245        past the start of the subject in the presence of bytes with the top bit
1246        set, for example "\x8aBCD".
1248    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1249        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1251    14. Optimized (?!) to (*FAIL).
1253    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1254        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1255        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1256        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1257        does: it's just the validity check that is more restrictive.
1259    16. Inserted checks for integer overflows during escape sequence (backslash)
1260        processing, and also fixed erroneous offset values for syntax errors during
1261        backslash processing.
1263    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1264        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1266    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1267        caused an overrun.
1269    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1270        something other than just ASCII characters) inside a group that had an
1271        unlimited repeat caused a loop at compile time (while checking to see
1272        whether the group could match an empty string).
1274    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1275        [\P{Any}] did so. (Error in the code for printing property names.)
1277    21. An orphan \E inside a character class could cause a crash.
1279    22. A repeated capturing bracket such as (A)? could cause a wild memory
1280        reference during compilation.
1282    23. There are several functions in pcre_compile() that scan along a compiled
1283        expression for various reasons (e.g. to see if it's fixed length for look
1284        behind). There were bugs in these functions when a repeated \p or \P was
1285        present in the pattern. These operators have additional parameters compared
1286        with \d, etc, and these were not being taken into account when moving along
1287        the compiled data. Specifically:
1289        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1290            length.
1292        (b) An item such as \pL+ within a repeated group could cause crashes or
1293            loops.
1295        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1296            "reference to non-existent subpattern" error.
1298        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1300    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1301        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1303    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1305    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1306        character were causing crashes (broken optimization).
1308    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1309        \p or \P) caused a compile-time loop.
1311    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1312        break. For example, the unstudied pattern [\r\n]A does not match the string
1313        "\r\nA" because change 7.0/46 below moves the current point on by two
1314        characters after failing to match at the start. However, the pattern \nA
1315        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1316        the same is true. There doesn't seem any very clean way out of this, but
1317        what I have chosen to do makes the common cases work: PCRE now takes note
1318        of whether there can be an explicit match for \r or \n anywhere in the
1319        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1320        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1321        pattern has explicit CR or LF references.
1323    29. Added (*CR) etc for changing newline setting at start of pattern.
1326    Version 7.2 19-Jun-07
1327    ---------------------
1329     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1330        which is apparently normally available under Windows.
1332     2. Re-jig the pcregrep tests with different newline settings in an attempt
1333        to make them independent of the local environment's newline setting.
1335     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1337     4. Some of the "internals" tests were previously cut out when the link size
1338        was not 2, because the output contained actual offsets. The recent new
1339        "Z" feature of pcretest means that these can be cut out, making the tests
1340        usable with all link sizes.
1342     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1343        stack recursion. This gives a massive performance boost under BSD, but just
1344        a small improvement under Linux. However, it saves one field in the frame
1345        in all cases.
1347     6. Added more features from the forthcoming Perl 5.10:
1349        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1350            recursion call. It refers to the nth most recently opened parentheses.
1352        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1353            to be opened parentheses.
1355        (c) Conditions that refer to capturing parentheses can be specified
1356            relatively, for example, (?(-2)... or (?(+3)...
1358        (d) \K resets the start of the current match so that everything before
1359            is not part of it.
1361        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1363        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1364            reference syntax.
1366        (g) (?| introduces a group in which the numbering of parentheses in each
1367            alternative starts with the same number.
1369        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1371     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1374     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1375        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1376        for detecting groups that can match an empty string.
1378     9. A pattern with a very large number of alternatives (more than several
1379        hundred) was running out of internal workspace during the pre-compile
1380        phase, where pcre_compile() figures out how much memory will be needed. A
1381        bit of new cunning has reduced the workspace needed for groups with
1382        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1383        workspace instead of running out of the 4096 that are available.
1385    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1387    11. Applied patch from Google to remove an optimization that didn't quite work.
1388        The report of the bug said:
1390          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1391          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1392          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1394    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1395        it matched the wrong number of bytes.
1398    Version 7.1 24-Apr-07
1399    ---------------------
1401     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1402        that is more "standard", making use of automake and other Autotools. There
1403        is some re-arrangement of the files and adjustment of comments consequent
1404        on this.
1406     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1407        for recursive directory scanning broke on some systems because the files
1408        are not scanned in any specific order and on different systems the order
1409        was different. A call to "sort" has been inserted into RunGrepTest for the
1410        approprate test as a short-term fix. In the longer term there may be an
1411        alternative.
1413     3. I had an email from Eric Raymond about problems translating some of PCRE's
1414        man pages to HTML (despite the fact that I distribute HTML pages, some
1415        people do their own conversions for various reasons). The problems
1416        concerned the use of low-level troff macros .br and .in. I have therefore
1417        removed all such uses from the man pages (some were redundant, some could
1418        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1419        HTML has been updated to handle .nf/.fi and to complain if it encounters
1420        .br or .in.
1422     4. Updated comments in configure.ac that get placed in config.h.in and also
1423        arranged for config.h to be included in the distribution, with the name
1424        config.h.generic, for the benefit of those who have to compile without
1425        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1427     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1428        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1429        makevp.bat for the current PCRE, using the additional files
1430        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1432     6. A Windows user reported a minor discrepancy with test 2, which turned out
1433        to be caused by a trailing space on an input line that had got lost in his
1434        copy. The trailing space was an accident, so I've just removed it.
1436     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1437        that is needed.
1439     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1440        as "const" (a) because they are and (b) because it helps the PHP
1441        maintainers who have recently made a script to detect big data structures
1442        in the php code that should be moved to the .rodata section. I remembered
1443        to update Builducptable as well, so it won't revert if ucptable.h is ever
1444        re-created.
1446     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1447        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1448        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1449        support is not required. This saves 1.5-2K of code, which is important in
1450        some applications.
1452        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1453        so as not to refer to the tables, even though these functions will never be
1454        called when UTF-8 support is disabled. Otherwise there are problems with a
1455        shared library.
1457    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1459        (a) It was defining its arguments as char * instead of void *.
1461        (b) It was assuming that all moves were upwards in memory; this was true
1462            a long time ago when I wrote it, but is no longer the case.
1464        The emulated memove() is provided for those environments that have neither
1465        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1466        is clearly not the case, as these two bugs were recently reported.
1468    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1469        and Detrail to create the HTML documentation, the .txt form of the man
1470        pages, and it removes trailing spaces from listed files. It also creates
1471        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1472        case, it wraps all the #defines with #ifndefs. This script should be run
1473        before "make dist".
1475    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1476        with Unicode property support.
1478        (a) For a maximizing quantifier, if the two different cases of the
1479            character were of different lengths in their UTF-8 codings (there are
1480            some cases like this - I found 11), and the matching function had to
1481            back up over a mixture of the two cases, it incorrectly assumed they
1482            were both the same length.
1484        (b) When PCRE was configured to use the heap rather than the stack for
1485            recursion during matching, it was not correctly preserving the data for
1486            the other case of a UTF-8 character when checking ahead for a match
1487            while processing a minimizing repeat. If the check also involved
1488            matching a wide character, but failed, corruption could cause an
1489            erroneous result when trying to check for a repeat of the original
1490            character.
1492    13. Some tidying changes to the testing mechanism:
1494        (a) The RunTest script now detects the internal link size and whether there
1495            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1496            values substituted by "configure". (The RunGrepTest script already did
1497            this for UTF-8.) The configure.ac script no longer substitutes the
1498            relevant variables.
1500        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1501            with length and offset values. This means that the output is different
1502            for different internal link sizes. Test 2 is skipped for link sizes
1503            other than 2 because of this, bypassing the problem. Unfortunately,
1504            there was also a test in test 3 (the locale tests) that used /B and
1505            failed for link sizes other than 2. Rather than cut the whole test out,
1506            I have added a new /Z option to pcretest that replaces the length and
1507            offset values with spaces. This is now used to make test 3 independent
1508            of link size. (Test 2 will be tidied up later.)
1510    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1511        segmentation fault instead of returning the appropriate error message.
1513    15. In multiline mode when the newline sequence was set to "any", the pattern
1514        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1515        This doesn't seem right; it now treats the CRLF combination as the line
1516        ending, and so does not match in that case. It's only a pattern such as ^$
1517        that would hit this one: something like ^ABC$ would have failed after \r
1518        and then tried again after \r\n.
1520    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1521        in an attempt to make files that differ only in their line terminators
1522        compare equal. This works on Linux.
1524    17. Under certain error circumstances pcregrep might try to free random memory
1525        as it exited. This is now fixed, thanks to valgrind.
1527    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1528        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1529        was because its rules for how to advance for /g after matching an empty
1530        string at the end of a line did not allow for this case. They now check for
1531        it specially.
1533    20. pcretest is supposed to handle patterns and data of any length, by
1534        extending its buffers when necessary. It was getting this wrong when the
1535        buffer for a data line had to be extended.
1537    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1538        CRLF as a newline sequence.
1540    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1541        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1542        I have nevertheless tidied it up.
1544    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1546    24. Added a man page for pcre-config.
1549    Version 7.0 19-Dec-06
1550    ---------------------
1552     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1553        moving to gcc 4.1.1.
1555     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1556        sys/time.h, which is documented as needed for this function. It doesn't
1557        seem to matter on Linux, but it showed up on some releases of OS X.
1559     3. It seems that there are systems where bytes whose values are greater than
1560        127 match isprint() in the "C" locale. The "C" locale should be the
1561        default when a C program starts up. In most systems, only ASCII printing
1562        characters match isprint(). This difference caused the output from pcretest
1563        to vary, making some of the tests fail. I have changed pcretest so that:
1565        (a) When it is outputting text in the compiled version of a pattern, bytes
1566            other than 32-126 are always shown as hex escapes.
1568        (b) When it is outputting text that is a matched part of a subject string,
1569            it does the same, unless a different locale has been set for the match
1570            (using the /L modifier). In this case, it uses isprint() to decide.
1572     4. Fixed a major bug that caused incorrect computation of the amount of memory
1573        required for a compiled pattern when options that changed within the
1574        pattern affected the logic of the preliminary scan that determines the
1575        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1576        that the computed length was too small. The symptoms of this bug were
1577        either the PCRE error "internal error: code overflow" from pcre_compile(),
1578        or a glibc crash with a message such as "pcretest: free(): invalid next
1579        size (fast)". Examples of patterns that provoked this bug (shown in
1580        pcretest format) are:
1582          /(?-x: )/x
1583          /(?x)(?-x: \s*#\s*)/
1584          /((?i)[\x{c0}])/8
1585          /(?i:[\x{c0}])/8
1587        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1588        is now done differently.
1590     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1591        wrapper classes; (b) implement a new function in the C++ scanner that is
1592        more efficient than the old way of doing things because it avoids levels of
1593        recursion in the regex matching; (c) add a paragraph to the documentation
1594        for the FullMatch() function.
1596     6. The escape sequence \n was being treated as whatever was defined as
1597        "newline". Not only was this contrary to the documentation, which states
1598        that \n is character 10 (hex 0A), but it also went horribly wrong when
1599        "newline" was defined as CRLF. This has been fixed.
1601     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1602        was being set to -1 for the "end of line" case (supposedly a value that no
1603        character can have). Though this value is never used (the check for end of
1604        line is "zero bytes in current character"), it caused compiler complaints.
1605        I've changed it to 0xffffffff.
1607     8. In pcre_version.c, the version string was being built by a sequence of
1608        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1609        string (as it is for production releases) called a macro with an empty
1610        argument. The C standard says the result of this is undefined. The gcc
1611        compiler treats it as an empty string (which was what was wanted) but it is
1612        reported that Visual C gives an error. The source has been hacked around to
1613        avoid this problem.
1615     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1616        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1617        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1618        of them did).
1620    10. Originally, pcretest opened its input and output without "b"; then I was
1621        told that "b" was needed in some environments, so it was added for release
1622        5.0 to both the input and output. (It makes no difference on Unix-like
1623        systems.) Later I was told that it is wrong for the input on Windows. I've
1624        now abstracted the modes into two macros, to make it easier to fiddle with
1625        them, and removed "b" from the input mode under Windows.
1627    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1629    12. Added -help and --help to pcretest as an official way of being reminded
1630        of the options.
1632    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1633        and pcrecpp.cc because they annoy compilers at high warning levels.
1635    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1637    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1638        scripts are not C programs :-) and which was not noticed because it works
1639        on Linux.
1641    16. pcretest is supposed to handle any length of pattern and data line (as one
1642        line or as a continued sequence of lines) by extending its input buffer if
1643        necessary. This feature was broken for very long pattern lines, leading to
1644        a string of junk being passed to pcre_compile() if the pattern was longer
1645        than about 50K.
1647    17. I have done a major re-factoring of the way pcre_compile() computes the
1648        amount of memory needed for a compiled pattern. Previously, there was code
1649        that made a preliminary scan of the pattern in order to do this. That was
1650        OK when PCRE was new, but as the facilities have expanded, it has become
1651        harder and harder to keep it in step with the real compile phase, and there
1652        have been a number of bugs (see for example, 4 above). I have now found a
1653        cunning way of running the real compile function in a "fake" mode that
1654        enables it to compute how much memory it would need, while actually only
1655        ever using a few hundred bytes of working memory and without too many
1656        tests of the mode. This should make future maintenance and development
1657        easier. A side effect of this work is that the limit of 200 on the nesting
1658        depth of parentheses has been removed (though this was never a serious
1659        limitation, I suspect). However, there is a downside: pcre_compile() now
1660        runs more slowly than before (30% or more, depending on the pattern). I
1661        hope this isn't a big issue. There is no effect on runtime performance.
1663    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1664        newline (only possible for the last line of a file) and it was a
1665        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1667    19. Added additional timing features to pcretest. (1) The -tm option now times
1668        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1669        separate command line item, by a number that specifies the number of
1670        repeats to use when timing. The default is 50000; this gives better
1671        precision, but takes uncomfortably long for very large patterns.
1673    20. Extended pcre_study() to be more clever in cases where a branch of a
1674        subpattern has no definite first character. For example, (a*|b*)[cd] would
1675        previously give no result from pcre_study(). Now it recognizes that the
1676        first character must be a, b, c, or d.
1678    21. There was an incorrect error "recursive call could loop indefinitely" if
1679        a subpattern (or the entire pattern) that was being tested for matching an
1680        empty string contained only one non-empty item after a nested subpattern.
1681        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1682        incorrectly, because the \d was being skipped in the check.
1684    22. The pcretest program now has a new pattern option /B and a command line
1685        option -b, which is equivalent to adding /B to every pattern. This causes
1686        it to show the compiled bytecode, without the additional information that
1687        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1688        is the same as /B/I).
1690    23. A new optimization is now able automatically to treat some sequences such
1691        as a*b as a*+b. More specifically, if something simple (such as a character
1692        or a simple class like \d) has an unlimited quantifier, and is followed by
1693        something that cannot possibly match the quantified thing, the quantifier
1694        is automatically "possessified".
1696    24. A recursive reference to a subpattern whose number was greater than 39
1697        went wrong under certain circumstances in UTF-8 mode. This bug could also
1698        have affected the operation of pcre_study().
1700    25. Realized that a little bit of performance could be had by replacing
1701        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1703    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1705    27. Possessive quantifiers such as a++ were previously implemented by turning
1706        them into atomic groups such as ($>a+). Now they have their own opcodes,
1707        which improves performance. This includes the automatically created ones
1708        from 23 above.
1710    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1711        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1712        the first matched character to be a colon. This applied both to named and
1713        numbered groups.
1715    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1717    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1718        building PCRE on Windows easier, so I have included it in the distribution.
1720    31. There is now a check in pcretest against a ridiculously large number being
1721        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1722        loop, the loop is abandoned.
1724    32. Forward references to subpatterns in conditions such as (?(2)...) where
1725        subpattern 2 is defined later cause pcre_compile() to search forwards in
1726        the pattern for the relevant set of parentheses. This search went wrong
1727        when there were unescaped parentheses in a character class, parentheses
1728        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1730    33. "Subroutine" calls and backreferences were previously restricted to
1731        referencing subpatterns earlier in the regex. This restriction has now
1732        been removed.
1734    34. Added a number of extra features that are going to be in Perl 5.10. On the
1735        whole, these are just syntactic alternatives for features that PCRE had
1736        previously implemented using the Python syntax or my own invention. The
1737        other formats are all retained for compatibility.
1739        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1740            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1741            also .NET compatible.
1743        (b) A recursion or subroutine call to a named group can now be defined as
1744            (?&name) as well as (?P>name).
1746        (c) A backreference to a named group can now be defined as \k<name> or
1747            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1748            5.10, are also .NET compatible.
1750        (d) A conditional reference to a named group can now use the syntax
1751            (?(<name>) or (?('name') as well as (?(name).
1753        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1754            groups (named and numbered) that are never evaluated inline, but can be
1755            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1756            is always false. There may be only one alternative in such a group.
1758        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1759            as the simple (?(R). The condition is true only if the most recent
1760            recursion is that of the given number or name. It does not search out
1761            through the entire recursion stack.
1763        (g) The escape \gN or \g{N} has been added, where N is a positive or
1764            negative number, specifying an absolute or relative reference.
1766    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1767        some "unreachable code" warnings.
1769    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1770        things, this adds five new scripts.
1772    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1773        There were also incompatibilities regarding the handling of \Q..\E inside
1774        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1775        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1777    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1778        matches an empty string, and forcibly breaks the loop. There were bugs in
1779        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1780        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1781        separate and independent bugs (that affected different cases) have been
1782        fixed.
1784    39. Refactored the code to abolish the use of different opcodes for small
1785        capturing bracket numbers. This is a tidy that I avoided doing when I
1786        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1787        The new approach is not only tidier, it makes it possible to reduce the
1788        memory needed to fix the previous bug (38).
1790    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1791        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1792        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1793        mode.
1795    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1796        report.
1798    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1799        copy construction and assignment in the C++ wrapper.
1801    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1802        couple of bugs that could have given wrong results in the "--newline=crlf"
1803        case.
1805    44. Added a number of casts and did some reorganization of signed/unsigned int
1806        variables following suggestions from Dair Grant. Also renamed the variable
1807        "this" as "item" because it is a C++ keyword.
1809    45. Arranged for dftables to add
1811          #include "pcre_internal.h"
1813        to pcre_chartables.c because without it, gcc 4.x may remove the array
1814        definition from the final binary if PCRE is built into a static library and
1815        dead code stripping is activated.
1817    46. For an unanchored pattern, if a match attempt fails at the start of a
1818        newline sequence, and the newline setting is CRLF or ANY, and the next two
1819        characters are CRLF, advance by two characters instead of one.
1822    Version 6.7 04-Jul-06
1823    ---------------------
1825     1. In order to handle tests when input lines are enormously long, pcretest has
1826        been re-factored so that it automatically extends its buffers when
1827        necessary. The code is crude, but this _is_ just a test program. The
1828        default size has been increased from 32K to 50K.
1830     2. The code in pcre_study() was using the value of the re argument before
1831        testing it for NULL. (Of course, in any sensible call of the function, it
1832        won't be NULL.)
1834     3. The memmove() emulation function in pcre_internal.h, which is used on
1835        systems that lack both memmove() and bcopy() - that is, hardly ever -
1836        was missing a "static" storage class specifier.
1838     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1839        containing an extended class (one that cannot be represented by a bitmap
1840        because it contains high-valued characters or Unicode property items, e.g.
1841        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1842        pattern, but PCRE should not loop if you do not (it no longer does).
1843        [Detail: two cases were found: (a) a repeated subpattern containing an
1844        extended class; (b) a recursive reference to a subpattern that followed a
1845        previous extended class. It wasn't skipping over the extended class
1846        correctly when UTF-8 mode was not set.]
1848     5. A negated single-character class was not being recognized as fixed-length
1849        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1850        compile error "lookbehind assertion is not fixed length".
1852     6. The RunPerlTest auxiliary script was showing an unexpected difference
1853        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1854        write a Perl script that can interpret lines of an input file either as
1855        byte characters or as UTF-8, which is what "perltest" was being required to
1856        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1857        can't do is switch easily at run time between having the "use utf8;" pragma
1858        or not. In the end, I fudged it by using the RunPerlTest script to insert
1859        "use utf8;" explicitly for the UTF-8 tests.
1861     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1862        the end of the subject string, contrary to the documentation and to what
1863        Perl does. This was true of both matching functions. Now it matches only at
1864        the start of the subject and immediately after *internal* newlines.
1866     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1867        a pointer to an int instead of a pointer to an unsigned long int. This
1868        caused problems on 64-bit systems.
1870     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1871        instance of the 'standard' template library not being so standard".
1873    10. There was no check on the number of named subpatterns nor the maximum
1874        length of a subpattern name. The product of these values is used to compute
1875        the size of the memory block for a compiled pattern. By supplying a very
1876        long subpattern name and a large number of named subpatterns, the size
1877        computation could be caused to overflow. This is now prevented by limiting
1878        the length of names to 32 characters, and the number of named subpatterns
1879        to 10,000.
1881    11. Subpatterns that are repeated with specific counts have to be replicated in
1882        the compiled pattern. The size of memory for this was computed from the
1883        length of the subpattern and the repeat count. The latter is limited to
1884        65535, but there was no limit on the former, meaning that integer overflow
1885        could in principle occur. The compiled length of a repeated subpattern is
1886        now limited to 30,000 bytes in order to prevent this.
1888    12. Added the optional facility to have named substrings with the same name.
1890    13. Added the ability to use a named substring as a condition, using the
1891        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1892        are numbers (not recommended). Forward references are permitted.
1894    14. Added forward references in named backreferences (if you see what I mean).
1896    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1897        pattern could run off the end of the subject. For example, the pattern
1898        "(?s)(.{1,5})"8 did this with the subject "ab".
1900    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1901        PCRE_CASELESS was set when matching characters that were quantified with ?
1902        or *.
1904    17. A character class other than a single negated character that had a minimum
1905        but no maximum quantifier - for example [ab]{6,} - was not handled
1906        correctly by pce_dfa_exec(). It would match only one character.
1908    18. A valid (though odd) pattern that looked like a POSIX character
1909        class but used an invalid character after [ (for example [[,abc,]]) caused
1910        pcre_compile() to give the error "Failed: internal error: code overflow" or
1911        in some cases to crash with a glibc free() error. This could even happen if
1912        the pattern terminated after [[ but there just happened to be a sequence of
1913        letters, a binary zero, and a closing ] in the memory that followed.
1915    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1916        over the years. Originally (before any Unicode support), just the bottom 8
1917        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1918        output from "man perlunicode" includes this:
1920          The regular expression compiler produces polymorphic opcodes.  That
1921          is, the pattern adapts to the data and automatically switches to
1922          the Unicode character scheme when presented with Unicode data--or
1923          instead uses a traditional byte scheme when presented with byte
1924          data.
1926        Sadly, a wide octal escape does not cause a switch, and in a string with
1927        no other multibyte characters, these octal escapes are treated as before.
1928        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1929        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1930        Unicode string.
1932        I have not perpetrated such confusion in PCRE. Up till now, it took just
1933        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1934        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1935        translate to the appropriate multibyte character.
1937    29. Applied some refactoring to reduce the number of warnings from Microsoft
1938        and Borland compilers. This has included removing the fudge introduced
1939        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1940        a warning about an unused variable.
1942    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1943        characters since release 4.0, because Perl (from release 5.004) does not.
1944        [Or at least, is documented not to: some releases seem to be in conflict
1945        with the documentation.] However, when a pattern was studied with
1946        pcre_study() and all its branches started with \s, PCRE still included VT
1947        as a possible starting character. Of course, this did no harm; it just
1948        caused an unnecessary match attempt.
1950    22. Removed a now-redundant internal flag bit that recorded the fact that case
1951        dependency changed within the pattern. This was once needed for "required
1952        byte" processing, but is no longer used. This recovers a now-scarce options
1953        bit. Also moved the least significant internal flag bit to the most-
1954        significant bit of the word, which was not previously used (hangover from
1955        the days when it was an int rather than a uint) to free up another bit for
1956        the future.
1958    23. Added support for CRLF line endings as well as CR and LF. As well as the
1959        default being selectable at build time, it can now be changed at runtime
1960        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1961        specify that it is scanning data with non-default line endings.
1963    24. Changed the definition of CXXLINK to make it agree with the definition of
1964        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1966    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1967        recursions. This makes a big different to stack usage for some patterns.
1969    26. If a subpattern containing a named recursion or subroutine reference such
1970        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1971        the space required for the compiled pattern went wrong and gave too small a
1972        value. Depending on the environment, this could lead to "Failed: internal
1973        error: code overflow at offset 49" or "glibc detected double free or
1974        corruption" errors.
1976    27. Applied patches from Google (a) to support the new newline modes and (b) to
1977        advance over multibyte UTF-8 characters in GlobalReplace.
1979    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1980        difference for some implementation of PCRE in some Windows version.
1982    29. Added some extra testing facilities to pcretest:
1984        \q<number>   in a data line sets the "match limit" value
1985        \Q<number>   in a data line sets the "match recursion limt" value
1986        -S <number>  sets the stack size, where <number> is in megabytes
1988        The -S option isn't available for Windows.
1991    Version 6.6 06-Feb-06
1992    ---------------------
1994     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1995        in pcreposix.h. I have copied the definition from pcre.h.
1997     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1998        because pcre.h is no longer a built file.
2000     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2001        not normally included in the compiled code.
2004    Version 6.5 01-Feb-06
2005    ---------------------
2007     1. When using the partial match feature with pcre_dfa_exec(), it was not
2008        anchoring the second and subsequent partial matches at the new starting
2009        point. This could lead to incorrect results. For example, with the pattern
2010        /1234/, partially matching against "123" and then "a4" gave a match.
2012     2. Changes to pcregrep:
2014        (a) All non-match returns from pcre_exec() were being treated as failures
2015            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2016            error message is output. Some extra information is given for the
2017            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2018            probably the only errors that are likely to be caused by users (by
2019            specifying a regex that has nested indefinite repeats, for instance).
2020            If there are more than 20 of these errors, pcregrep is abandoned.
2022        (b) A binary zero was treated as data while matching, but terminated the
2023            output line if it was written out. This has been fixed: binary zeroes
2024            are now no different to any other data bytes.
2026        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2027            used to set a locale for matching. The --locale=xxxx long option has
2028            been added (no short equivalent) to specify a locale explicitly on the
2029            pcregrep command, overriding the environment variables.
2031        (d) When -B was used with -n, some line numbers in the output were one less
2032            than they should have been.
2034        (e) Added the -o (--only-matching) option.
2036        (f) If -A or -C was used with -c (count only), some lines of context were
2037            accidentally printed for the final match.
2039        (g) Added the -H (--with-filename) option.
2041        (h) The combination of options -rh failed to suppress file names for files
2042            that were found from directory arguments.
2044        (i) Added the -D (--devices) and -d (--directories) options.
2046        (j) Added the -F (--fixed-strings) option.
2048        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2050        (l) Added the --colo(u)r option.
2052        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2053            is not present by default.
2055     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2056        items such as (?R) or (?1), when the recursion could match a number of
2057        alternatives. If it matched one of the alternatives, but subsequently,
2058        outside the recursion, there was a failure, the code tried to back up into
2059        the recursion. However, because of the way PCRE is implemented, this is not
2060        possible, and the result was an incorrect result from the match.
2062        In order to prevent this happening, the specification of recursion has
2063        been changed so that all such subpatterns are automatically treated as
2064        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2066     4. I had overlooked the fact that, in some locales, there are characters for
2067        which isalpha() is true but neither isupper() nor islower() are true. In
2068        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2069        and ordfeminine) are like this. This affected the treatment of \w and \W
2070        when they appeared in character classes, but not when they appeared outside
2071        a character class. The bit map for "word" characters is now created
2072        separately from the results of isalnum() instead of just taking it from the
2073        upper, lower, and digit maps. (Plus the underscore character, of course.)
2075     5. The above bug also affected the handling of POSIX character classes such as
2076        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2077        permanent tables. Instead, the bit maps for such a class were previously
2078        created as the appropriate unions of the upper, lower, and digit bitmaps.
2079        Now they are created by subtraction from the [[:word:]] class, which has
2080        its own bitmap.
2082     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2083        It is created by subtracting the vertical space characters (\x09, \x0a,
2084        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2085        subtraction was done in the overall bitmap for a character class, meaning
2086        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2087        be recognized. This bug has been fixed.
2089     7. Patches from the folks at Google:
2091          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2092          real life, but is still worth protecting against".
2094          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2095          regular expressions".
2097          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2098          have it.
2100          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2101          "configure" and the latter not, in order to fix a problem somebody had
2102          with compiling the Arg class on HP-UX.
2104          (e) Improve the error-handling of the C++ wrapper a little bit.
2106          (f) New tests for checking recursion limiting.
2108     8. The pcre_memmove() function, which is used only if the environment does not
2109        have a standard memmove() function (and is therefore rarely compiled),
2110        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2111        returning a result (though PCRE never actually uses the result).
2113     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2114        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2115        returned instead of calling malloc() with an overflowing number that would
2116        most likely cause subsequent chaos.
2118    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2120    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2121        with this option is matched, the nmatch and pmatch options of regexec() are
2122        ignored.
2124    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2125        provided in case anyone wants to the the POSIX interface with UTF-8
2126        strings.
2128    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2129        C++ linking (needed for some HP-UX environments).
2131    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2132        (unused parameter) and in the pcre_printint() function (omitted "default"
2133        switch label when the default is to do nothing).
2135    15. Added some code to make it possible, when PCRE is compiled as a C++
2136        library, to replace subject pointers for pcre_exec() with a smart pointer
2137        class, thus making it possible to process discontinuous strings.
2139    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2140        much the same function. They were added by different people who were trying
2141        to make PCRE easy to compile on non-Unix systems. It has been suggested
2142        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2143        for compiling on Windows systems. I have therefore replaced it with
2144        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2145        defaults to "extern" for C or "extern C" for C++, which works fine on
2146        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2147        SCOPE with something explicit in config.h. In addition:
2149        (a) pcreposix.h still had just "extern" instead of either of these macros;
2150            I have replaced it with PCRE_DATA_SCOPE.
2152        (b) Functions such as _pcre_xclass(), which are internal to the library,
2153            but external in the C sense, all had PCRE_EXPORT in their definitions.
2154            This is apparently wrong for the Windows case, so I have removed it.
2155            (It makes no difference on Unix-like systems.)
2157    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2158        of recursive calls to match(). This is different to MATCH_LIMIT because
2159        that limits the total number of calls to match(), not all of which increase
2160        the depth of recursion. Limiting the recursion depth limits the amount of
2161        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2162        when PCRE is compiled, and changed at run time. A patch from Google adds
2163        this functionality to the C++ interface.
2165    18. Changes to the handling of Unicode character properties:
2167        (a) Updated the table to Unicode 4.1.0.
2169        (b) Recognize characters that are not in the table as "Cn" (undefined).
2171        (c) I revised the way the table is implemented to a much improved format
2172            which includes recognition of ranges. It now supports the ranges that
2173            are defined in UnicodeData.txt, and it also amalgamates other
2174            characters into ranges. This has reduced the number of entries in the
2175            table from around 16,000 to around 3,000, thus reducing its size
2176            considerably. I realized I did not need to use a tree structure after
2177            all - a binary chop search is just as efficient. Having reduced the
2178            number of entries, I extended their size from 6 bytes to 8 bytes to
2179            allow for more data.
2181        (d) Added support for Unicode script names via properties such as \p{Han}.
2183    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2184        matching that character.
2186    20. When matching a repeated Unicode property with a minimum greater than zero,
2187        (for example \pL{2,}), PCRE could look past the end of the subject if it
2188        reached it while seeking the minimum number of characters. This could
2189        happen only if some of the characters were more than one byte long, because
2190        there is a check for at least the minimum number of bytes.
2192    21. Refactored the implementation of \p and \P so as to be more general, to
2193        allow for more different types of property in future. This has changed the
2194        compiled form incompatibly. Anybody with saved compiled patterns that use
2195        \p or \P will have to recompile them.
2197    22. Added "Any" and "L&" to the supported property types.
2199    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2200        but give a compile time error if the value is greater than 0xff.
2202    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2203        accidentally not being installed or uninstalled.
2205    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2206        made were to insert the current release number. This seemed silly, because
2207        it made things harder for people building PCRE on systems that don't run
2208        "configure". I have turned pcre.h into a distributed file, no longer built
2209        by "configure", with the version identification directly included. There is
2210        no longer a pcre.h.in file.
2212        However, this change necessitated a change to the pcre-config script as
2213        well. It is built from pcre-config.in, and one of the substitutions was the
2214        release number. I have updated configure.ac so that ./configure now finds
2215        the release number by grepping pcre.h.
2217    26. Added the ability to run the tests under valgrind.
2220    Version 6.4 05-Sep-05
2221    ---------------------
2223     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2224        "--" to be printed when multiple files were scanned, even when none of the
2225        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2226        consider it to be a bug, and have restored the previous behaviour.
2228     2. A couple of code tidies to get rid of compiler warnings.
2230     3. The pcretest program used to cheat by referring to symbols in the library
2231        whose names begin with _pcre_. These are internal symbols that are not
2232        really supposed to be visible externally, and in some environments it is
2233        possible to suppress them. The cheating is now confined to including
2234        certain files from the library's source, which is a bit cleaner.
2236     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2237        file's purpose clearer.
2239     5. Reorganized pcre_ucp_findchar().
2242    Version 6.3 15-Aug-05
2243    ---------------------
2245     1. The file libpcre.pc.in did not have general read permission in the tarball.
2247     2. There were some problems when building without C++ support:
2249        (a) If C++ support was not built, "make install" and "make test" still
2250            tried to test it.
2252        (b) There were problems when the value of CXX was explicitly set. Some
2253            changes have been made to try to fix these, and ...
2255        (c) --disable-cpp can now be used to explicitly disable C++ support.
2257        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2258            backslash in a target when C++ was disabled. This confuses some
2259            versions of "make", apparently. Using an intermediate variable solves
2260            this. (Same for CPP_LOBJ.)
2262     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2263        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2264        necessary on certain architectures.
2266     4. Added a setting of -export-symbols-regex to the link command to remove
2267        those symbols that are exported in the C sense, but actually are local
2268        within the library, and not documented. Their names all begin with
2269        "_pcre_". This is not a perfect job, because (a) we have to except some
2270        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2271        available (and never for static libraries). I have made a note to try to
2272        find a way round (a) in the future.
2275    Version 6.2 01-Aug-05
2276    ---------------------
2278     1. There was no test for integer overflow of quantifier values. A construction
2279        such as {1111111111111111} would give undefined results. What is worse, if
2280        a minimum quantifier for a parenthesized subpattern overflowed and became
2281        negative, the calculation of the memory size went wrong. This could have
2282        led to memory overwriting.
2284     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2286     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2287        operating environments where this matters.
2289     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2290        PCRE options from within the C++ wrapper.
2292     5. Named capturing subpatterns were not being correctly counted when a pattern
2293        was compiled. This caused two problems: (a) If there were more than 100
2294        such subpatterns, the calculation of the memory needed for the whole
2295        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2296        back references of the form \12, where the number was greater than 9, were
2297        not recognized as back references, even though there were sufficient
2298        previous subpatterns.
2300     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2301        versions of gcc, e.g. 2.95.4.
2304    Version 6.1 21-Jun-05
2305    ---------------------
2307     1. There was one reference to the variable "posix" in pcretest.c that was not
2308        surrounded by "#if !defined NOPOSIX".
2310     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2311        the cross-check on the old pcre_info() function, for the benefit of the
2312        cut-down version of PCRE that is currently imported into Exim.
2314     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2315        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2316        patterns that start (?i) but I don't think that matters. The use of (?i) is
2317        just an example; this all applies to the other options as well.
2319     4. Since libtool seems to echo the compile commands it is issuing, the output
2320        from "make" can be reduced a bit by putting "@" in front of each libtool
2321        compile command.
2323     5. Patch from the folks at Google for configure.in to be a bit more thorough
2324        in checking for a suitable C++ installation before trying to compile the
2325        C++ stuff. This should fix a reported problem when a compiler was present,
2326        but no suitable headers.
2328     6. The man pages all had just "PCRE" as their title. I have changed them to
2329        be the relevant file name. I have also arranged that these names are
2330        retained in the file doc/pcre.txt, which is a concatenation in text format
2331        of all the man pages except the little individual ones for each function.
2333     7. The NON-UNIX-USE file had not been updated for the different set of source
2334        files that come with release 6. I also added a few comments about the C++
2335        wrapper.
2338    Version 6.0 07-Jun-05
2339    ---------------------
2341     1. Some minor internal re-organization to help with my DFA experiments.
2343     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2344        didn't matter for the library itself when fully configured, but did matter
2345        when compiling without UCP support, or within Exim, where the ucp files are
2346        not imported.
2348     3. Refactoring of the library code to split up the various functions into
2349        different source modules. The addition of the new DFA matching code (see
2350        below) to a single monolithic source would have made it really too
2351        unwieldy, quite apart from causing all the code to be include in a
2352        statically linked application, when only some functions are used. This is
2353        relevant even without the DFA addition now that patterns can be compiled in
2354        one application and matched in another.
2356        The downside of splitting up is that there have to be some external
2357        functions and data tables that are used internally in different modules of
2358        the library but which are not part of the API. These have all had their
2359        names changed to start with "_pcre_" so that they are unlikely to clash
2360        with other external names.
2362     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2363        a different (DFA) algorithm. Although it is slower than the original
2364        function, it does have some advantages for certain types of matching
2365        problem.
2367     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2368        including restarting after a partial match.
2370     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2371        defined when compiling for Windows was sent to me. I have put it into the
2372        code, though I have no means of testing or verifying it.
2374     7. Added the pcre_refcount() auxiliary function.
2376     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2377        match before or at the first newline in the subject string. In pcretest,
2378        the /f option on a pattern can be used to set this.
2380     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2381        would behave wrongly. This has been present in PCRE since release 4.0.
2383    10. A number of changes to the pcregrep command:
2385        (a) Refactored how -x works; insert ^(...)$ instead of setting
2386            PCRE_ANCHORED and checking the length, in preparation for adding
2387            something similar for -w.
2389        (b) Added the -w (match as a word) option.
2391        (c) Refactored the way lines are read and buffered so as to have more
2392            than one at a time available.
2394        (d) Implemented a pcregrep test script.
2396        (e) Added the -M (multiline match) option. This allows patterns to match
2397            over several lines of the subject. The buffering ensures that at least
2398            8K, or the rest of the document (whichever is the shorter) is available
2399            for matching (and similarly the previous 8K for lookbehind assertions).
2401        (f) Changed the --help output so that it now says
2403              -w, --word-regex(p)
2405            instead of two lines, one with "regex" and the other with "regexp"
2406            because that confused at least one person since the short forms are the
2407            same. (This required a bit of code, as the output is generated
2408            automatically from a table. It wasn't just a text change.)
2410        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2411            option but starts with a hyphen. Could be a pattern or a path name
2412            starting with a hyphen, for instance.
2414        (h) "-" can be given as a file name to represent stdin.
2416        (i) When file names are being printed, "(standard input)" is used for
2417            the standard input, for compatibility with GNU grep. Previously
2418            "<stdin>" was used.
2420        (j) The option --label=xxx can be used to supply a name to be used for
2421            stdin when file names are being printed. There is no short form.
2423        (k) Re-factored the options decoding logic because we are going to add
2424            two more options that take data. Such options can now be given in four
2425            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2427        (l) Added the -A, -B, and -C options for requesting that lines of context
2428            around matches be printed.
2430        (m) Added the -L option to print the names of files that do not contain
2431            any matching lines, that is, the complement of -l.
2433        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2434            continue to scan other files.
2436        (o) The -s option was incorrectly implemented. For compatibility with other
2437            greps, it now suppresses the error message for a non-existent or non-
2438            accessible file (but not the return code). There is a new option called
2439            -q that suppresses the output of matching lines, which was what -s was
2440            previously doing.
2442        (p) Added --include and --exclude options to specify files for inclusion
2443            and exclusion when recursing.
2445    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2446        Hopefully, it now does.
2448    12. Missing cast in pcre_study().
2450    13. Added an "uninstall" target to the makefile.
2452    14. Replaced "extern" in the function prototypes in Makefile.in with
2453        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2454        world, but is set differently for Windows.
2456    15. Added a second compiling function called pcre_compile2(). The only
2457        difference is that it has an extra argument, which is a pointer to an
2458        integer error code. When there is a compile-time failure, this is set
2459        non-zero, in addition to the error test pointer being set to point to an
2460        error message. The new argument may be NULL if no error number is required
2461        (but then you may as well call pcre_compile(), which is now just a
2462        wrapper). This facility is provided because some applications need a
2463        numeric error indication, but it has also enabled me to tidy up the way
2464        compile-time errors are handled in the POSIX wrapper.
2466    16. Added VPATH=.libs to the makefile; this should help when building with one
2467        prefix path and installing with another. (Or so I'm told by someone who
2468        knows more about this stuff than I do.)
2470    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2471        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2472        match everything, including newlines. This is not POSIX-compatible, but
2473        somebody wanted the feature. From pcretest it can be activated by using
2474        both the P and the s flags.
2476    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2478    19. libpcre.pc was being incorrectly installed as executable.
2480    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2481        it now also looks for '\r' so that it will work unmodified on Windows.
2483    21. Added Google's contributed C++ wrapper to the distribution.
2485    22. Added some untidy missing memory free() calls in pcretest, to keep
2486        Electric Fence happy when testing.
2490    Version 5.0 13-Sep-04
2491    ---------------------
2493     1. Internal change: literal characters are no longer packed up into items
2494        containing multiple characters in a single byte-string. Each character
2495        is now matched using a separate opcode. However, there may be more than one
2496        byte in the character in UTF-8 mode.
2498     2. The pcre_callout_block structure has two new fields: pattern_position and
2499        next_item_length. These contain the offset in the pattern to the next match
2500        item, and its length, respectively.
2502     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2503        insertion of callouts before each pattern item. Added the /C option to
2504        pcretest to make use of this.
2506     4. On the advice of a Windows user, the lines
2508          #if defined(_WIN32) || defined(WIN32)
2509          _setmode( _fileno( stdout ), 0x8000 );
2510          #endif  /* defined(_WIN32) || defined(WIN32) */
2512        have been added to the source of pcretest. This apparently does useful
2513        magic in relation to line terminators.
2515     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2516        for the benefit of those environments where the "b" makes a difference.
2518     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2519        to know about it. I have put a hack into configure.in that adds in code
2520        to set GCC=yes if CC=icc. This seems to end up at a point in the
2521        generated configure script that is early enough to affect the setting of
2522        compiler options, which is what is needed, but I have no means of testing
2523        whether it really works. (The user who reported this had patched the
2524        generated configure script, which of course I cannot do.)
2526        LATER: After change 22 below (new libtool files), the configure script
2527        seems to know about icc (and also ecc). Therefore, I have commented out
2528        this hack in configure.in.
2530     7. Added support for pkg-config (2 patches were sent in).
2532     8. Negated POSIX character classes that used a combination of internal tables
2533        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2534        [[:^ascii]]. Typically, they would match almost any characters. The other
2535        POSIX classes were not broken in this way.
2537     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2538        to find the match, as PCRE was deluded into thinking that the match had to
2539        start at the start point or following a newline. The same bug applied to
2540        patterns with negative forward assertions or any backward assertions
2541        preceding ".*" at the start, unless the pattern required a fixed first
2542        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2544    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2545        starting at the last subject character, bytes beyond the end of the subject
2546        string were read.
2548    11. Renamed the variable "class" as "classbits" to make life easier for C++
2549        users. (Previously there was a macro definition, but it apparently wasn't
2550        enough.)
2552    12. Added the new field "tables" to the extra data so that tables can be passed
2553        in at exec time, or the internal tables can be re-selected. This allows
2554        a compiled regex to be saved and re-used at a later time by a different
2555        program that might have everything at different addresses.
2557    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2558        -R library as well as a -L library.
2560    14. The debugging options of pcretest (-d on the command line or D on a
2561        pattern) showed incorrect output for anything following an extended class
2562        that contained multibyte characters and which was followed by a quantifier.
2564    15. Added optional support for general category Unicode character properties
2565        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2566        support. It adds about 90K to the size of the library. The meanings of the
2567        inbuilt class escapes such as \d and \s have NOT been changed.
2569    16. Updated pcredemo.c to include calls to free() to release the memory for the
2570        compiled pattern.
2572    17. The generated file chartables.c was being created in the source directory
2573        instead of in the building directory. This caused the build to fail if the
2574        source directory was different from the building directory, and was
2575        read-only.
2577    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2578        file. No doubt somebody will tell me if they don't make sense... Also added
2579        Dan Mooney's comments about building on OpenVMS.
2581    19. Added support for partial matching via the PCRE_PARTIAL option for
2582        pcre_exec() and the \P data escape in pcretest.
2584    20. Extended pcretest with 3 new pattern features:
2586        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2587              write the compiled pattern to the file whose name is "rest-of-line".
2588              This is a straight binary dump of the data, with the saved pointer to
2589              the character tables forced to be NULL. The study data, if any, is
2590              written too. After writing, pcretest reads a new pattern.
2592        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2593              compiled pattern from the given file. There must not be any
2594              occurrences of "<" in the file name (pretty unlikely); if there are,
2595              pcretest will instead treat the initial "<" as a pattern delimiter.
2596              After reading in the pattern, pcretest goes on to read data lines as
2597              usual.
2599        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2600              and 16-bit fields in a compiled pattern, to simulate a pattern that
2601              was compiled on a host of opposite endianness.
2603    21. The pcre-exec() function can now cope with patterns that were compiled on
2604        hosts of opposite endianness, with this restriction:
2606          As for any compiled expression that is saved and used later, the tables
2607          pointer field cannot be preserved; the extra_data field in the arguments
2608          to pcre_exec() should be used to pass in a tables address if a value
2609          other than the default internal tables were used at compile time.
2611    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2612        now diagnosed as an error. Previously, most of the time, a negative number
2613        would have been treated as zero, but if in addition "ovector" was passed as
2614        NULL, a crash could occur.
2616    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2617        new versions from the libtool 1.5 distribution (the last one is a copy of
2618        a file called libtool.m4). This seems to have fixed the need to patch
2619        "configure" to support Darwin 1.3 (which I used to do). However, I still
2620        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2621        workstation).
2623    24. Changed the PCRE licence to be the more standard "BSD" licence.
2626    Version 4.5 01-Dec-03
2627    ---------------------
2629     1. There has been some re-arrangement of the code for the match() function so
2630        that it can be compiled in a version that does not call itself recursively.
2631        Instead, it keeps those local variables that need separate instances for
2632        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2633        needs to "recurse". Keeping track of where control must go is done by means
2634        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2635        hide most of the details from the main code, and operates only if
2636        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2637        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2638        operating.
2640        To make it easier for callers to provide specially tailored get/free
2641        functions for this usage, two new functions, pcre_stack_malloc, and
2642        pcre_stack_free, are used. They are always called in strict stacking order,
2643        and the size of block requested is always the same.
2645        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2646        PCRE has been compiled to use the stack or the heap for recursion. The
2647        -C option of pcretest uses this to show which version is compiled.
2649        A new data escape \S, is added to pcretest; it causes the amounts of store
2650        obtained and freed by both kinds of malloc/free at match time to be added
2651        to the output.
2653     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2654        what's available on my current Linux desktop machine.
2656     3. When matching a UTF-8 string, the test for a valid string at the start has
2657        been extended. If start_offset is not zero, PCRE now checks that it points
2658        to a byte that is the start of a UTF-8 character. If not, it returns
2659        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2660        this is necessary because there may be backward assertions in the pattern.
2661        When matching the same subject several times, it may save resources to use
2662        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2664     4. The code for checking the validity of UTF-8 strings has been tightened so
2665        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2666        containing "overlong sequences".
2668     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2669        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2670        should have been "&", but it just so happened that all the cases this let
2671        through by mistake were picked up later in the function.
2673     6. I had used a variable called "isblank" - this is a C99 function, causing
2674        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2676     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2677        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2678        so the version is shown; (c) stop "make test" echoing "./RunTest".
2680     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2682     9. The prototype for memmove() for systems that don't have it was using
2683        size_t, but the inclusion of the header that defines size_t was later. I've
2684        moved the #includes for the C headers earlier to avoid this.
2686    10. Added some adjustments to the code to make it easier to compiler on certain
2687        special systems:
2689          (a) Some "const" qualifiers were missing.
2690          (b) Added the macro EXPORT before all exported functions; by default this
2691              is defined to be empty.
2692          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2693              that it reads its output file name as an argument instead of writing
2694              to the standard output and assuming this can be redirected.
2696    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2697        class containing characters with values greater than 255, PCRE compilation
2698        went into a loop.
2700    12. A recursive reference to a subpattern that was within another subpattern
2701        that had a minimum quantifier of zero caused PCRE to crash. For example,
2702        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2703        recursion. If the recursively-called subpattern itself had a zero repeat,
2704        that was OK.
2706    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2707        buffer into which it was copied (for escape processing) was still set at
2708        1024, so long lines caused crashes.
2710    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2711        "internal error: code overflow...". This applied to any character class
2712        that was followed by a possessive quantifier.
2714    15. Modified the Makefile to add libpcre.la as a prerequisite for
2715        libpcreposix.la because I was told this is needed for a parallel build to
2716        work.
2718    16. If a pattern that contained .* following optional items at the start was
2719        studied, the wrong optimizing data was generated, leading to matching
2720        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2721        matching string must start with a or b or c. The correct conclusion for
2722        this pattern is that a match can start with any character.
2725    Version 4.4 13-Aug-03
2726    ---------------------
2728     1. In UTF-8 mode, a character class containing characters with values between
2729        127 and 255 was not handled correctly if the compiled pattern was studied.
2730        In fixing this, I have also improved the studying algorithm for such
2731        classes (slightly).
2733     2. Three internal functions had redundant arguments passed to them. Removal
2734        might give a very teeny performance improvement.
2736     3. Documentation bug: the value of the capture_top field in a callout is *one
2737        more than* the number of the hightest numbered captured substring.
2739     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2740        in incorrectly linking with a previously installed version. They now link
2741        explicitly with libpcre.la.
2743     5. configure.in no longer needs to recognize Cygwin specially.
2745     6. A problem in pcre.in for Windows platforms is fixed.
2747     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2748        pcretest, it used to include the size of the study block as part of its
2749        output. Unfortunately, the structure contains a field that has a different
2750        size on different hardware architectures. This meant that the tests that
2751        showed this size failed. As the block is currently always of a fixed size,
2752        this information isn't actually particularly useful in pcretest output, so
2753        I have just removed it.
2755     8. Three pre-processor statements accidentally did not start in column 1.
2756        Sadly, there are *still* compilers around that complain, even though
2757        standard C has not required this for well over a decade. Sigh.
2759     9. In pcretest, the code for checking callouts passed small integers in the
2760        callout_data field, which is a void * field. However, some picky compilers
2761        complained about the casts involved for this on 64-bit systems. Now
2762        pcretest passes the address of the small integer instead, which should get
2763        rid of the warnings.
2765    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2766        both compile and run time, and gives an error if an invalid UTF-8 sequence
2767        is found. There is a option for disabling this check in cases where the
2768        string is known to be correct and/or the maximum performance is wanted.
2770    11. In response to a bug report, I changed one line in Makefile.in from
2772            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2773        to
2774            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2776        to look similar to other lines, but I have no way of telling whether this
2777        is the right thing to do, as I do not use Windows. No doubt I'll get told
2778        if it's wrong...
2781    Version 4.3 21-May-03
2782    ---------------------
2784    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2785       Makefile.
2787    2. Some refactoring to improve the quality of the code:
2789       (i)   The utf8_table... variables are now declared "const".
2791       (ii)  The code for \cx, which used the "case flipping" table to upper case
2792             lower case letters, now just substracts 32. This is ASCII-specific,
2793             but the whole concept of \cx is ASCII-specific, so it seems
2794             reasonable.
2796       (iii) PCRE was using its character types table to recognize decimal and
2797             hexadecimal digits in the pattern. This is silly, because it handles
2798             only 0-9, a-f, and A-F, but the character types table is locale-
2799             specific, which means strange things might happen. A private
2800             table is now used for this - though it costs 256 bytes, a table is
2801             much faster than multiple explicit tests. Of course, the standard
2802             character types table is still used for matching digits in subject
2803             strings against \d.
2805       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2806             ending in _t are). So I've renamed it as ESC_tee.
2808    3. The first argument for regexec() in the POSIX wrapper should have been
2809       defined as "const".
2811    4. Changed pcretest to use malloc() for its buffers so that they can be
2812       Electric Fenced for debugging.
2814    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2815       to read one or more bytes before the start of the subject string. Often this
2816       had no effect on PCRE's behaviour, but in some circumstances it could
2817       provoke a segmentation fault.
2819    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2820       to try to read one or more bytes before the start of the subject string.
2822    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2823       UTF-8 support could misbehave in various ways if the subject string
2824       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2825       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2826       back over UTF-8 characters.)
2829    Version 4.2 14-Apr-03
2830    ---------------------
2832    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2834    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2835         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2836         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2837         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2838         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2839           and BUILD_EXEEXT
2840         Note: automatic setting of the BUILD variables is not yet working
2841         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2842           compile-time but not at link-time
2843         [LINK]: use for linking executables only
2844         make different versions for Windows and non-Windows
2845         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2846           libraries
2847         [LINK_FOR_BUILD]: new variable
2848         [OBJEXT]: use throughout
2849         [EXEEXT]: use throughout
2850         <winshared>: new target
2851         <wininstall>: new target
2852         <dftables.o>: use native compiler
2853         <dftables>: use native linker
2854         <install>: handle Windows platform correctly
2855         <clean>: ditto
2856         <check>: ditto
2857         copy DLL to top builddir before testing
2859       As part of these changes, -no-undefined was removed again. This was reported
2860       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2861       in any case.
2863    3. Some tidies to get rid of compiler warnings:
2865       . In the match_data structure, match_limit was an unsigned long int, whereas
2866         match_call_count was an int. I've made them both unsigned long ints.
2868       . In pcretest the fact that a const uschar * doesn't automatically cast to
2869         a void * provoked a warning.
2871       . Turning on some more compiler warnings threw up some "shadow" variables
2872         and a few more missing casts.
2874    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2875       option, a class that contained a single character with a value between 128
2876       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2878    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2879       option, a class that contained several characters, but with at least one
2880       whose value was between 128 and 255 caused PCRE to crash.
2883    Version 4.1 12-Mar-03
2884    ---------------------
2886    1. Compiling with gcc -pedantic found a couple of places where casts were
2887    needed, and a string in dftables.c that was longer than standard compilers are
2888    required to support.
2890    2. Compiling with Sun's compiler found a few more places where the code could
2891    be tidied up in order to avoid warnings.
2893    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2894    first of these names is deprecated in the latest Autoconf in favour of the name
2895    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2896    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2897    analogy I have changed it to CFLAGS_FOR_BUILD.
2899    4. Added -no-undefined to the linking command in the Makefile, because this is
2900    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2901    linking step for the pcreposix library.
2903    5. PCRE was failing to diagnose the case of two named groups with the same
2904    name.
2906    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2907    literal character that is needed in the subject for a match, and scans along to
2908    ensure that it is present before embarking on the full matching process. This
2909    saves time in cases of nested unlimited repeats that are never going to match.
2910    Problem: the scan can take a lot of time if the subject is very long (e.g.
2911    megabytes), thus penalizing straightforward matches. It is now done only if the
2912    amount of subject to be scanned is less than 1000 bytes.
2914    7. A lesser problem with the same optimization is that it was recording the
2915    first character of an anchored pattern as "needed", thus provoking a search
2916    right along the subject, even when the first match of the pattern was going to
2917    fail. The "needed" character is now not set for anchored patterns, unless it
2918    follows something in the pattern that is of non-fixed length. Thus, it still
2919    fulfils its original purpose of finding quick non-matches in cases of nested
2920    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2923    Version 4.0 17-Feb-03
2924    ---------------------
2926    1. If a comment in an extended regex that started immediately after a meta-item
2927    extended to the end of string, PCRE compiled incorrect data. This could lead to
2928    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2930    2. Moved to autoconf 2.53 and libtool 1.4.2.
2932    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2933    the special perltest8 script is no longer needed - all the tests can be run
2934    from a single perltest script.
2936    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2937    by \s. It has now been removed in PCRE. This means it isn't recognized as
2938    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2939    class [:space:] *does* include VT, thereby creating a mess.
2941    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2942    space and tab.
2944    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2945    its new features into the main test script, reducing the number of scripts.
2947    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2948    were backward compatible, and made the (?i) apply to the whole pattern, as if
2949    /i were given. Now it behaves more logically, and applies the option setting
2950    only to what follows. PCRE has been changed to follow suit. However, if it
2951    finds options settings right at the start of the pattern, it extracts them into
2952    the global options, as before. Thus, they show up in the info data.
2954    8. Added support for the \Q...\E escape sequence. Characters in between are
2955    treated as literals. This is slightly different from Perl in that $ and @ are
2956    also handled as literals inside the quotes. In Perl, they will cause variable
2957    interpolation. Note the following examples:
2959        Pattern            PCRE matches      Perl matches
2961        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2962        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2963        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2965    For compatibility with Perl, \Q...\E sequences are recognized inside character
2966    classes as well as outside them.
2968    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2969    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2970    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2971    signed/unsigned warnings.
2973    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2974    option for pcretest, so I've replaced it by a simple function that does just
2975    that job.
2977    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2978    "pcregrep -".
2980    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2981    Java package. This provides some syntactic sugar for simple cases of what my
2982    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2983    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2984    item, you can use this simplified notation. Note that only makes sense with
2985    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2986    greediness, whatever the setting of the PCRE_UNGREEDY option.
2988    13. A change of greediness default within a pattern was not taking effect at
2989    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2990    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2991    was abstracted outside.
2993    14. PCRE now supports the \G assertion. It is true when the current matching
2994    position is at the start point of the match. This differs from \A when the
2995    starting offset is non-zero. Used with the /g option of pcretest (or similar
2996    code), it works in the same way as it does for Perl's /g option. If all
2997    alternatives of a regex begin with \G, the expression is anchored to the start
2998    match position, and the "anchored" flag is set in the compiled expression.
3000    15. Some bugs concerning the handling of certain option changes within patterns
3001    have been fixed. These applied to options other than (?ims). For example,
3002    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3003    been the other way round. Some of this was related to change 7 above.
3005    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3006    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3007    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3008    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3010    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3011    mode. Unlike ".", it always matches newline, whatever the setting of
3012    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3013    assertions. Perl allows it, but it doesn't (in general) work because it can't
3014    calculate the length of the lookbehind. At least, that's the case for Perl
3015    5.8.0 - I've been told they are going to document that it doesn't work in
3016    future.
3018    18. Added an error diagnosis for escapes that PCRE does not support: these are
3019    \L, \l, \N, \P, \p, \U, \u, and \X.
3021    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3022    reading past the end of the pattern in cases such as /[abcd/.
3024    20. PCRE was getting more memory than necessary for patterns with classes that
3025    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3027    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3028    compiling PCRE for use with Virtual Pascal.
3030    22. Small fix to the Makefile to make it work properly if the build is done
3031    outside the source tree.
3033    23. Added a new extension: a condition to go with recursion. If a conditional
3034    subpattern starts with (?(R) the "true" branch is used if recursion has
3035    happened, whereas the "false" branch is used only at the top level.
3037    24. When there was a very long string of literal characters (over 255 bytes
3038    without UTF support, over 250 bytes with UTF support), the computation of how
3039    much memory was required could be incorrect, leading to segfaults or other
3040    strange effects.
3042    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3043    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3044    there was a subsequent back reference to those brackets. This meant that, for
3045    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3046    possible to check for precisely this case. All we can do is abandon the
3047    optimization if .* occurs inside capturing brackets when there are any back
3048    references whatsoever. (See below for a better fix that came later.)
3050    26. The handling of the optimization for finding the first character of a
3051    non-anchored pattern, and for finding a character that is required later in the
3052    match were failing in some cases. This didn't break the matching; it just
3053    failed to optimize when it could. The way this is done has been re-implemented.
3055    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3057    28. Added a new feature that provides some of the functionality that Perl
3058    provides with (?{...}). The facility is termed a "callout". The way it is done
3059    in PCRE is for the caller to provide an optional function, by setting
3060    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3061    global variable. By default it is unset, which disables all calling out. To get
3062    the function called, the regex must include (?C) at appropriate points. This
3063    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3064    This provides a means of identifying different callout points. When PCRE
3065    reaches such a point in the regex, if pcre_callout has been set, the external
3066    function is called. It is provided with data in a structure called
3067    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3068    matching continues; if it returns a non-zero value, the match at the current
3069    point fails. However, backtracking will occur if possible. [This was changed
3070    later and other features added - see item 49 below.]
3072    29. pcretest is upgraded to test the callout functionality. It provides a
3073    callout function that displays information. By default, it shows the start of
3074    the match and the current position in the text. There are some new data escapes
3075    to vary what happens:
3077        \C+         in addition, show current contents of captured substrings
3078        \C-         do not supply a callout function
3079        \C!n        return 1 when callout number n is reached
3080        \C!n!m      return 1 when callout number n is reached for the mth time
3082    30. If pcregrep was called with the -l option and just a single file name, it
3083    output "<stdin>" if a match was found, instead of the file name.
3085    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3086    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3087    pcre_exec(). This saves a malloc/free per call. The default value of
3088    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3089    when configuring.
3091    32. The default maximum size of a compiled pattern is 64K. There have been a
3092    few cases of people hitting this limit. The code now uses macros to handle the
3093    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3094    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3095    configuring. Tests 2 and 5 work only with 2-byte links because they output
3096    debugging information about compiled patterns.
3098    33. Internal code re-arrangements:
3100    (a) Moved the debugging function for printing out a compiled regex into
3101        its own source file (printint.c) and used #include to pull it into
3102        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3103        separate copies.
3105    (b) Defined the list of op-code names for debugging as a macro in
3106        internal.h so that it is next to the definition of the opcodes.
3108    (c) Defined a table of op-code lengths for simpler skipping along compiled
3109        code. This is again a macro in internal.h so that it is next to the
3110        definition of the opcodes.
3112    34. Added support for recursive calls to individual subpatterns, along the
3113    lines of Robin Houston's patch (but implemented somewhat differently).
3115    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3116    allow it to read and process whole directories in Win32. This code was
3117    contributed by Lionel Fourquaux; it has not been tested by me.
3119    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3120    used to name a group. Names consist of alphanumerics and underscores, and must
3121    be unique. Back references use the syntax (?P=name) and recursive calls use
3122    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3123    numbers. The function pcre_fullinfo() can be used after compilation to extract
3124    a name/number map. There are three relevant calls:
3126      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3127      PCRE_INFO_NAMECOUNT            yields the number of entries
3128      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3130    The map is a vector of fixed-size entries. The size of each entry depends on
3131    the length of the longest name used. The first two bytes of each entry are the
3132    group number, most significant byte first. There follows the corresponding
3133    name, zero terminated. The names are in alphabetical order.
3135    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3136    case instead of 255. Making it the same both with and without UTF-8 support
3137    means that the same test output works with both.
3139    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3140    calling malloc() with a zero argument.
3142    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3143    optimization. I've improved things by keeping a bitmap of backreferences with
3144    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3145    fact referenced, the optimization can be applied. It is unlikely that a
3146    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3147    the match to follow \n) will appear inside brackets with a number greater than
3148    31, but if it does, any back reference > 31 suppresses the optimization.
3150    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3151    of disabling numbered capturing parentheses. Any opening parenthesis that is
3152    not followed by ? behaves as if it were followed by ?: but named parentheses
3153    can still be used for capturing (and they will acquire numbers in the usual
3154    way).
3156    41. Redesigned the return codes from the match() function into yes/no/error so
3157    that errors can be passed back from deep inside the nested calls. A malloc
3158    failure while inside a recursive subpattern call now causes the
3159    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3161    42. It is now possible to set a limit on the number of times the match()
3162    function is called in a call to pcre_exec(). This facility makes it possible to
3163    limit the amount of recursion and backtracking, though not in a directly
3164    obvious way, because the match() function is used in a number of different
3165    circumstances. The count starts from zero for each position in the subject
3166    string (for non-anchored patterns). The default limit is, for compatibility, a
3167    large number, namely 10 000 000. You can change this in two ways:
3169    (a) When configuring PCRE before making, you can use --with-match-limit=n
3170        to set a default value for the compiled library.
3172    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3173        a different value is set. See 45 below.
3175    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3177    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3178    of things that can be changed at compile time. The first argument specifies
3179    what is wanted and the second points to where the information is to be placed.
3180    The current list of available information is:
3182      PCRE_CONFIG_UTF8
3184    The output is an integer that is set to one if UTF-8 support is available;
3185    otherwise it is set to zero.
3189    The output is an integer that it set to the value of the code that is used for
3190    newline. It is either LF (10) or CR (13).
3194    The output is an integer that contains the number of bytes used for internal
3195    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3199    The output is an integer that contains the threshold above which the POSIX
3200    interface uses malloc() for output vectors. See item 31 above.
3204    The output is an unsigned integer that contains the default limit of the number
3205    of match() calls in a pcre_exec() execution. See 42 above.
3207    44. pcretest has been upgraded by the addition of the -C option. This causes it
3208    to extract all the available output from the new pcre_config() function, and to
3209    output it. The program then exits immediately.
3211    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3212    order to support additional features. One way would have been to define
3213    pcre_exec2() (for example) with extra arguments, but this would not have been
3214    extensible, and would also have required all calls to the original function to
3215    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3216    is used for passing in "extra" data from pcre_study().
3218    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3219    contains the following fields:
3221      flags         a bitmap indicating which of the following fields are set
3222      study_data    opaque data from pcre_study()
3223      match_limit   a way of specifying a limit on match() calls for a specific
3224                      call to pcre_exec()
3225      callout_data  data for callouts (see 49 below)
3227    The flag bits are also defined in pcre.h, and are
3233    The pcre_study() function now returns one of these new pcre_extra blocks, with
3234    the actual study data pointed to by the study_data field, and the
3235    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3236    before. That is, this change is entirely upwards-compatible and requires no
3237    change to existing code.
3239    If you want to pass in additional data to pcre_exec(), you can either place it
3240    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3241    block.
3243    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3244    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3245    times with different match limits, until it finds the minimum value needed for
3246    pcre_exec() to complete. The value is then output. This can be instructive; for
3247    most simple matches the number is quite small, but for pathological cases it
3248    gets very large very quickly.
3250    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3251    returns the size of the data block pointed to by the study_data field in a
3252    pcre_extra block, that is, the value that was passed as the argument to
3253    pcre_malloc() when PCRE was getting memory in which to place the information
3254    created by pcre_study(). The fourth argument should point to a size_t variable.
3255    pcretest has been extended so that this information is shown after a successful
3256    pcre_study() call when information about the compiled regex is being displayed.
3258    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3259    because what follows is always an absolute path. (Later: it turns out that this
3260    is more than cosmetic for MinGW, because it doesn't like empty path
3261    components.)
3263    49. Some changes have been made to the callout feature (see 28 above):
3265    (i)  A callout function now has three choices for what it returns:
3267           0  =>  success, carry on matching
3268         > 0  =>  failure at this point, but backtrack if possible
3269         < 0  =>  serious error, return this value from pcre_exec()
3271         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3272         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3273         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3274         use by callout functions. It will never be used by PCRE itself.
3276    (ii) The pcre_extra structure (see 45 above) has a void * field called
3277         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3278         pcre_callout_block structure has a field of the same name. The contents of
3279         the field passed in the pcre_extra structure are passed to the callout
3280         function in the corresponding field in the callout block. This makes it
3281         easier to use the same callout-containing regex from multiple threads. For
3282         testing, the pcretest program has a new data escape
3284           \C*n        pass the number n (may be negative) as callout_data
3286         If the callout function in pcretest receives a non-zero value as
3287         callout_data, it returns that value.
3289    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3290    there were some redundant $(CFLAGS) in commands that are now specified as
3291    $(LINK), which already includes $(CFLAGS).
3293    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3294    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3295    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3296    one-byte characters throughout. Note that case-insensitive matching applies
3297    only to characters whose values are less than 256. PCRE doesn't support the
3298    notion of cases for higher-valued characters.
3300    (i)   A character class whose characters are all within 0-255 is handled as
3301          a bit map, and the map is inverted for negative classes. Previously, a
3302          character > 255 always failed to match such a class; however it should
3303          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3305    (ii)  A negated character class with a single character < 255 is coded as
3306          "not this character" (OP_NOT). This wasn't working properly when the test
3307          character was multibyte, either singly or repeated.
3309    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3310          mode, for example: \x{100}{2,3}.
3312    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3313          singly or repeated) now correctly test multibyte characters. However,
3314          PCRE doesn't recognize any characters with values greater than 255 as
3315          digits, spaces, or word characters. Such characters always match \D, \S,
3316          and \W, and never match \d, \s, or \w.
3318    (v)   Classes may now contain characters and character ranges with values
3319          greater than 255. For example: [ab\x{100}-\x{400}].
3321    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3322          PCRE in UTF-8 mode.
3324    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3325    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3326    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3327    value.)
3329    53. The single man page has become too large. I have therefore split it up into
3330    a number of separate man pages. These also give rise to individual HTML pages;
3331    these are now put in a separate directory, and there is an index.html page that
3332    lists them all. Some hyperlinking between the pages has been installed.
3334    54. Added convenience functions for handling named capturing parentheses.
3336    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3337    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3338    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3339    are faulted.
3341    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3342    calling configure. These values are used when compiling the dftables.c program
3343    which is run to generate the source of the default character tables. They
3344    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3345    you will need to set these values.
3347    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3350    Version 3.9 02-Jan-02
3351    ---------------------
3353    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3355    2. If --disable-static was given, the building process failed when trying to
3356    build pcretest and pcregrep. (For some reason it was using libtool to compile
3357    them, which is not right, as they aren't part of the library.)
3360    Version 3.8 18-Dec-01
3361    ---------------------
3363    1. The experimental UTF-8 code was completely screwed up. It was packing the
3364    bytes in the wrong order. How dumb can you get?
3367    Version 3.7 29-Oct-01
3368    ---------------------
3370    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3371    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3372    this didn't happen under Solaris 8, where I normally test things.
3374    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3375    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3376    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3377    this fix an hour or so after the initial 3.7 release.)
3380    Version 3.6 23-Oct-01
3381    ---------------------
3383    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3384    offsets passed as NULL with zero offset count.
3386    2. The config.guess and config.sub files had not been updated when I moved to
3387    the latest autoconf.
3390    Version 3.5 15-Aug-01
3391    ---------------------
3393    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3394    had been forgotten.
3396    2. By using declared but undefined structures, we can avoid using "void"
3397    definitions in pcre.h while keeping the internal definitions of the structures
3398    private.
3400    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3401    user point of view, this means that both static and shared libraries are built
3402    by default, but this can be individually controlled. More of the work of
3403    handling this static/shared cases is now inside libtool instead of PCRE's make
3404    file.
3406    4. The pcretest utility is now installed along with pcregrep because it is
3407    useful for users (to test regexs) and by doing this, it automatically gets
3408    relinked by libtool. The documentation has been turned into a man page, so
3409    there are now .1, .txt, and .html versions in /doc.
3411    5. Upgrades to pcregrep:
3412       (i)   Added long-form option names like gnu grep.
3413       (ii)  Added --help to list all options with an explanatory phrase.
3414       (iii) Added -r, --recursive to recurse into sub-directories.
3415       (iv)  Added -f, --file to read patterns from a file.
3417    6. pcre_exec() was referring to its "code" argument before testing that
3418    argument for NULL (and giving an error if it was NULL).
3420    7. Upgraded Makefile.in to allow for compiling in a different directory from
3421    the source directory.
3423    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3424    options bits, the pointer it was passed was to an int instead of to an unsigned
3425    long int. This mattered only on 64-bit systems.
3427    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3428    generated) instead of pcre.in, which it its source. Also made the same change
3429    in several of the .c files.
3431    10. A new release of gcc defines printf() as a macro, which broke pcretest
3432    because it had an ifdef in the middle of a string argument for printf(). Fixed
3433    by using separate calls to printf().
3435    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3436    script, to force use of CR or LF instead of \n in the source. On non-Unix
3437    systems, the value can be set in config.h.
3439    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3440    absolute limit. Changed the text of the error message to make this clear, and
3441    likewise updated the man page.
3443    13. The limit of 99 on the number of capturing subpatterns has been removed.
3444    The new limit is 65535, which I hope will not be a "real" limit.
3447    Version 3.4 22-Aug-00
3448    ---------------------
3450    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3452    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3455    Version 3.3 01-Aug-00
3456    ---------------------
3458    1. If an octal character was given, but the value was greater than \377, it
3459    was not getting masked to the least significant bits, as documented. This could
3460    lead to crashes in some systems.
3462    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3463    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3465    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3466    These just pass their arguments on to (pcre_free)(), but they are provided
3467    because some uses of PCRE bind it to non-C systems that can call its functions,
3468    but cannot call free() or pcre_free() directly.
3470    4. Add "make test" as a synonym for "make check". Corrected some comments in
3471    the Makefile.
3473    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3474    Makefile.
3476    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3477    command called pgrep for grepping around the active processes.
3479    7. Added the beginnings of support for UTF-8 character strings.
3481    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3482    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3483    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3484    out for the ar command.)
3487    Version 3.2 12-May-00
3488    ---------------------
3490    This is purely a bug fixing release.
3492    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3493    of ZA. This was just one example of several cases that could provoke this bug,
3494    which was introduced by change 9 of version 2.00. The code for breaking
3495    infinite loops after an iteration that matches an empty string was't working
3496    correctly.
3498    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3499    when matched against abbab (for example). After matching an empty string, it
3500    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3501    caused it to match further down the string than it should.
3503    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3504    was there because it doesn't seem to be needed, and it causes trouble on some
3505    systems, as it is not a Standard C header. It has been removed.
3507    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3508    were reported on the Macintosh. The changes were from
3510      while ((c = *(++ptr)) != 0 && c != '\n');
3511    to
3512      while ((c = *(++ptr)) != 0 && c != '\n') ;
3514    Totally extraordinary, but if that's what it takes...
3516    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3517    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3518    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3519    assumes the way PCRE uses memmove() (always moving upwards).
3521    6. PCRE is being used in one environment where strchr() is not available. There
3522    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3523    faster code anyway.
3526    Version 3.1 09-Feb-00
3527    ---------------------
3529    The only change in this release is the fixing of some bugs in Makefile.in for
3530    the "install" target:
3532    (1) It was failing to install pcreposix.h.
3534    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3537    Version 3.0 01-Feb-00
3538    ---------------------
3540    1. Add support for the /+ modifier to perltest (to output $` like it does in
3541    pcretest).
3543    2. Add support for the /g modifier to perltest.
3545    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3546    matches null strings.
3548    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3549    pattern. Perl treats empty patterns specially - it reuses the most recent
3550    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3551    effect.
3553    5. The POSIX interface was broken in that it was just handing over the POSIX
3554    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3555    required a bigger vector, with some working space on the end. This means that
3556    the POSIX wrapper now has to get and free some memory, and copy the results.
3558    6. Added some simple autoconf support, placing the test data and the
3559    documentation in separate directories, re-organizing some of the
3560    information files, and making it build pcre-config (a GNU standard). Also added
3561    libtool support for building PCRE as a shared library, which is now the
3562    default.
3564    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3565    09 are not valid octal constants. Single digits will be used for minor values
3566    less than 10.
3568    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3569    existing programs that set these in the POSIX interface can use PCRE without
3570    modification.
3572    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3573    return all that pcre_info() returns, plus additional data. The pcre_info()
3574    function is retained for compatibility, but is considered to be obsolete.
3576    10. Added experimental recursion feature (?R) to handle one common case that
3577    Perl 5.6 will be able to do with (?p{...}).
3579    11. Added support for POSIX character classes like [:alpha:], which Perl is
3580    adopting.
3583    Version 2.08 31-Aug-99
3584    ----------------------
3586    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3587    trying to match at the startoffset position, but instead was moving forward to
3588    the next newline as if a previous match had failed.
3590    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3591    and could get into a loop if a null string was matched other than at the start
3592    of the subject.
3594    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3595    be distinguished at compile time, and for completeness also added PCRE_DATE.
3597    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3598    in GnuWin32 environments.
3601    Version 2.07 29-Jul-99
3602    ----------------------
3604    1. The documentation is now supplied in plain text form and HTML as well as in
3605    the form of man page sources.
3607    2. C++ compilers don't like assigning (void *) values to other pointer types.
3608    In particular this affects malloc(). Although there is no problem in Standard
3609    C, I've put in casts to keep C++ compilers happy.
3611    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3612    should be (const char *).
3614    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3615    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3616    However, I haven't made this a standard facility. The documentation doesn't
3617    mention it, and the Makefile doesn't support it.
3619    5. The Makefile now contains an "install" target, with editable destinations at
3620    the top of the file. The pcretest program is not installed.
3622    6. pgrep -V now gives the PCRE version number and date.
3624    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3625    causing the entire string to be ignored, instead of just the last character.
3627    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3628    non-matching string, it can take a very, very long time, even for strings of
3629    quite modest length, because of the nested recursion. PCRE now does better in
3630    some of these cases. It does this by remembering the last required literal
3631    character in the pattern, and pre-searching the subject to ensure it is present