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 558 by ph10, Tue Oct 26 15:26:45 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.11 10-Oct-2010
5    ------------------------
7    1.  (*THEN) was not working properly if there were untried alternatives prior
8        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
9        backtracked to try for "b" instead of moving to the next alternative branch
10        at the same level (in this case, to look for "c"). The Perl documentation
11        is clear that when (*THEN) is backtracked onto, it goes to the "next
12        alternative in the innermost enclosing group".
14    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
15        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
16        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
17        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
18        (*THEN).
20    3.  If \s appeared in a character class, it removed the VT character from
21        the class, even if it had been included by some previous item, for example
22        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
23        of \s, but is part of the POSIX "space" class.)
25    4.  A partial match never returns an empty string (because you can always
26        match an empty string at the end of the subject); however the checking for
27        an empty string was starting at the "start of match" point. This has been
28        changed to the "earliest inspected character" point, because the returned
29        data for a partial match starts at this character. This means that, for
30        example, /(?<=abc)def/ gives a partial match for the subject "abc"
31        (previously it gave "no match").
33    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
34        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
35        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
36        has an implication that the given string is incomplete (because a partial
37        match is preferred over a full match). For this reason, these items now
38        give a partial match in this situation. [Aside: previously, the one case
39        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
40        match rather than a full match, which was wrong by the old rules, but is
41        now correct.]
43    6.  There was a bug in the handling of #-introduced comments, recognized when
44        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
45        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
46        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
47        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
48        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
49        places in pcre_compile().
51    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
52        comments when looking ahead for named forward references to subpatterns,
53        the only newline sequence it recognized was NL. It now handles newlines
54        according to the set newline convention.
56    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
57        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
58        cater for a lack of strerror(). These oversights have been fixed.
61    Version 8.10 25-Jun-2010
62    ------------------------
64    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
65        THEN.
67    2.  (*ACCEPT) was not working when inside an atomic group.
69    3.  Inside a character class, \B is treated as a literal by default, but
70        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
71        causes the error). The code is unchanged, but I tidied the documentation.
73    4.  Inside a character class, PCRE always treated \R and \X as literals,
74        whereas Perl faults them if its -w option is set. I have changed PCRE so
75        that it faults them when PCRE_EXTRA is set.
77    5.  Added support for \N, which always matches any character other than
78        newline. (It is the same as "." when PCRE_DOTALL is not set.)
80    6.  When compiling pcregrep with newer versions of gcc which may have
81        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
82        declared with attribute warn_unused_result" were given. Just casting the
83        result to (void) does not stop the warnings; a more elaborate fudge is
84        needed. I've used a macro to implement this.
86    7.  Minor change to pcretest.c to avoid a compiler warning.
88    8.  Added four artifical Unicode properties to help with an option to make
89        \s etc use properties (see next item). The new properties are: Xan
90        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
92    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
93        use Unicode properties. (*UCP) at the start of a pattern can be used to set
94        this option. Modified pcretest to add /W to test this facility. Added
95        REG_UCP to make it available via the POSIX interface.
97    10. Added --line-buffered to pcregrep.
99    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
100        studied, and the match started with a letter with a code point greater than
101        127 whose first byte was different to the first byte of the other case of
102        the letter, the other case of this starting letter was not recognized
103        (#976).
105    12. If a pattern that was studied started with a repeated Unicode property
106        test, for example, \p{Nd}+, there was the theoretical possibility of
107        setting up an incorrect bitmap of starting bytes, but fortunately it could
108        not have actually happened in practice until change 8 above was made (it
109        added property types that matched character-matching opcodes).
111    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
112        possible starting bytes for non-anchored patterns.
114    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
115        \R, and also a number of cases that involve Unicode properties, both
116        explicit and implicit when PCRE_UCP is set.
118    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
119        input, it could crash or give wrong results if characters with values
120        greater than 0xc0 were present in the subject string. (Detail: it assumed
121        UTF-8 input when processing these items.)
123    16. Added a lot of (int) casts to avoid compiler warnings in systems where
124        size_t is 64-bit (#991).
126    17. Added a check for running out of memory when PCRE is compiled with
127        --disable-stack-for-recursion (#990).
129    18. If the last data line in a file for pcretest does not have a newline on
130        the end, a newline was missing in the output.
132    19. The default pcre_chartables.c file recognizes only ASCII characters (values
133        less than 128) in its various bitmaps. However, there is a facility for
134        generating tables according to the current locale when PCRE is compiled. It
135        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
136        characters, are recognized by isspace() and therefore were getting set in
137        these tables, and indeed these tables seem to approximate to ISO 8859. This
138        caused a problem in UTF-8 mode when pcre_study() was used to create a list
139        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
140        which of course cannot start UTF-8 characters. I have changed the code so
141        that only real ASCII characters (less than 128) and the correct starting
142        bytes for UTF-8 encodings are set for characters greater than 127 when in
143        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
144        altogether.)
146    20. Added the /T option to pcretest so as to be able to run tests with non-
147        standard character tables, thus making it possible to include the tests
148        used for 19 above in the standard set of tests.
150    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
151        reference to a subpattern the other side of a comment that contains an
152        opening parenthesis caused either an internal compiling error, or a
153        reference to the wrong subpattern.
156    Version 8.02 19-Mar-2010
157    ------------------------
159    1.  The Unicode data tables have been updated to Unicode 5.2.0.
161    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
162        configured.
164    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
165        original author of that file, following a query about its status.
167    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
168        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
170    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
171        quantifier applied to a forward-referencing subroutine call, could compile
172        incorrect code or give the error "internal error: previously-checked
173        referenced subpattern not found".
175    6.  Both MS Visual Studio and Symbian OS have problems with initializing
176        variables to point to external functions. For these systems, therefore,
177        pcre_malloc etc. are now initialized to local functions that call the
178        relevant global functions.
180    7.  There were two entries missing in the vectors called coptable and poptable
181        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
182        I've fixed the data, and added a kludgy way of testing at compile time that
183        the lengths are correct (equal to the number of opcodes).
185    8.  Following on from 7, I added a similar kludge to check the length of the
186        eint vector in pcreposix.c.
188    9.  Error texts for pcre_compile() are held as one long string to avoid too
189        much relocation at load time. To find a text, the string is searched,
190        counting zeros. There was no check for running off the end of the string,
191        which could happen if a new error number was added without updating the
192        string.
194    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
196    11. \K was not working if it appeared in an atomic group or in a group that
197        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
198        \K is "not well defined" if used in an assertion. PCRE now accepts it if
199        the assertion is positive, but not if it is negative.
201    12. Change 11 fortuitously reduced the size of the stack frame used in the
202        "match()" function of pcre_exec.c by one pointer. Forthcoming
203        implementation of support for (*MARK) will need an extra pointer on the
204        stack; I have reserved it now, so that the stack frame size does not
205        decrease.
207    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
208        item in branch that calls a recursion is a subroutine call - as in the
209        second branch in the above example - was incorrectly given the compile-
210        time error "recursive call could loop indefinitely" because pcre_compile()
211        was not correctly checking the subroutine for matching a non-empty string.
213    14. The checks for overrunning compiling workspace could trigger after an
214        overrun had occurred. This is a "should never occur" error, but it can be
215        triggered by pathological patterns such as hundreds of nested parentheses.
216        The checks now trigger 100 bytes before the end of the workspace.
218    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
221    Version 8.01 19-Jan-2010
222    ------------------------
224    1.  If a pattern contained a conditional subpattern with only one branch (in
225        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
226        computed the wrong minimum data length (which is of course zero for such
227        subpatterns). This could cause incorrect "no match" results.
229    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
230        the pattern is reset in the first branch, pcre_compile() failed with
231        "internal error: code overflow at offset...". This happened only when
232        the reset was to the original external option setting. (An optimization
233        abstracts leading options settings into an external setting, which was the
234        cause of this.)
236    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
237        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
238        assertion pattern did not match (meaning that the assertion was true), it
239        was incorrectly treated as false if the SKIP had been reached during the
240        matching. This also applied to assertions used as conditions.
242    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
243        assertion subpattern, including such a pattern used as a condition,
244        unpredictable results occurred, instead of the error return
247    5.  The C++ GlobalReplace function was not working like Perl for the special
248        situation when an empty string is matched. It now does the fancy magic
249        stuff that is necessary.
251    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
252        removed. (These were left over from very, very early versions of PCRE.)
254    7.  Some cosmetic changes to the code to make life easier when compiling it
255        as part of something else:
257        (a) Change DEBUG to PCRE_DEBUG.
259        (b) In pcre_compile(), rename the member of the "branch_chain" structure
260            called "current" as "current_branch", to prevent a collision with the
261            Linux macro when compiled as a kernel module.
263        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
264            prevent a collision with the Linux macro when compiled as a kernel
265            module.
267    8.  In pcre_compile() there are some checks for integer overflows that used to
268        cast potentially large values to (double). This has been changed to that
269        when building, a check for int64_t is made, and if it is found, it is used
270        instead, thus avoiding the use of floating point arithmetic. (There is no
271        other use of FP in PCRE.) If int64_t is not found, the fallback is to
272        double.
274    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
275        2005 (difference between two addresses compared to an unsigned value).
277    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
278        custom one, because of the following reported problem in Windows:
280          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
281              under Win32.
282          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
283              therefore missing the function definition.
284          - The compiler thus generates a "C" signature for the test function.
285          - The linker fails to find the "C" function.
286          - PCRE fails to configure if asked to do so against libbz2.
288    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
289        messages were output:
291          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
292          rerunning libtoolize, to keep the correct libtool macros in-tree.
293          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
295        I have done both of these things.
297    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
298        most of the time, it *can* run out if it is given a pattern that contains a
299        runaway infinite recursion. I updated the discussion in the pcrestack man
300        page.
302    13. Now that we have gone to the x.xx style of version numbers, the minor
303        version may start with zero. Using 08 or 09 is a bad idea because users
304        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
305        interpreted as invalid octal numbers. I've updated the previous comment in
306        configure.ac, and also added a check that gives an error if 08 or 09 are
307        used.
309    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
310        causing partial matching to fail when the end of the subject matched \W
311        in a UTF-8 pattern where \W was quantified with a minimum of 3.
313    15. There were some discrepancies between the declarations in pcre_internal.h
314        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
315        their definitions. The declarations used "const uschar *" and the
316        definitions used USPTR. Even though USPTR is normally defined as "const
317        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
318        reported that: "This difference in casting confuses some C++ compilers, for
319        example, SunCC recognizes above declarations as different functions and
320        generates broken code for hbpcre." I have changed the declarations to use
321        USPTR.
323    16. GNU libtool is named differently on some systems. The autogen.sh script now
324        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
325        (FreeBSD).
327    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
328        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
329        comment: "Figure out how to create a longlong from a string: strtoll and
330        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
331        instance, but it only takes 2 args instead of 3!"
333    18. A subtle bug concerned with back references has been fixed by a change of
334        specification, with a corresponding code fix. A pattern such as
335        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
336        refers, was giving matches when it shouldn't. For example, xa=xaaa would
337        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
338        same bug. Such groups have to be quantified to be useful, or contained
339        inside another quantified group. (If there's no repetition, the reference
340        can never match.) The problem arises because, having left the group and
341        moved on to the rest of the pattern, a later failure that backtracks into
342        the group uses the captured value from the final iteration of the group
343        rather than the correct earlier one. I have fixed this in PCRE by forcing
344        any group that contains a reference to itself to be an atomic group; that
345        is, there cannot be any backtracking into it once it has completed. This is
346        similar to recursive and subroutine calls.
349    Version 8.00 19-Oct-09
350    ----------------------
352    1.  The table for translating pcre_compile() error codes into POSIX error codes
353        was out-of-date, and there was no check on the pcre_compile() error code
354        being within the table. This could lead to an OK return being given in
355        error.
357    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
358        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
359        in a Windows environment.
361    3.  The pcregrep --count option prints the count for each file even when it is
362        zero, as does GNU grep. However, pcregrep was also printing all files when
363        --files-with-matches was added. Now, when both options are given, it prints
364        counts only for those files that have at least one match. (GNU grep just
365        prints the file name in this circumstance, but including the count seems
366        more useful - otherwise, why use --count?) Also ensured that the
367        combination -clh just lists non-zero counts, with no names.
369    4.  The long form of the pcregrep -F option was incorrectly implemented as
370        --fixed_strings instead of --fixed-strings. This is an incompatible change,
371        but it seems right to fix it, and I didn't think it was worth preserving
372        the old behaviour.
374    5.  The command line items --regex=pattern and --regexp=pattern were not
375        recognized by pcregrep, which required --regex pattern or --regexp pattern
376        (with a space rather than an '='). The man page documented the '=' forms,
377        which are compatible with GNU grep; these now work.
379    6.  No libpcreposix.pc file was created for pkg-config; there was just
380        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
382    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
383        when UCP support is not needed, by modifying the Python script that
384        generates it from Unicode data files. This should not matter if the module
385        is correctly used as a library, but I received one complaint about 50K of
386        unwanted data. My guess is that the person linked everything into his
387        program rather than using a library. Anyway, it does no harm.
389    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
390        was a minimum greater than 1 for a wide character in a possessive
391        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
392        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
393        character. Chaos in the form of incorrect output or a compiling loop could
394        result.
396    9.  The restrictions on what a pattern can contain when partial matching is
397        requested for pcre_exec() have been removed. All patterns can now be
398        partially matched by this function. In addition, if there are at least two
399        slots in the offset vector, the offset of the earliest inspected character
400        for the match and the offset of the end of the subject are set in them when
401        PCRE_ERROR_PARTIAL is returned.
403    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
404        synonymous with PCRE_PARTIAL, for backwards compatibility, and
405        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
406        and may be more useful for multi-segment matching.
408    11. Partial matching with pcre_exec() is now more intuitive. A partial match
409        used to be given if ever the end of the subject was reached; now it is
410        given only if matching could not proceed because another character was
411        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
412        string "Z", which now yields "no match" instead of "partial match". In the
413        case of pcre_dfa_exec(), "no match" is given if every matching path for the
414        final character ended with (*FAIL).
416    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
417        if the pattern had a "must contain" character that was already found in the
418        earlier partial match, unless partial matching was again requested. For
419        example, with the pattern /dog.(body)?/, the "must contain" character is
420        "g". If the first part-match was for the string "dog", restarting with
421        "sbody" failed. This bug has been fixed.
423    13. The string returned by pcre_dfa_exec() after a partial match has been
424        changed so that it starts at the first inspected character rather than the
425        first character of the match. This makes a difference only if the pattern
426        starts with a lookbehind assertion or \b or \B (\K is not supported by
427        pcre_dfa_exec()). It's an incompatible change, but it makes the two
428        matching functions compatible, and I think it's the right thing to do.
430    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
431        so that the demonstration program is easily available in environments where
432        PCRE has not been installed from source.
434    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
435        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
436        library.
438    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
439        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
440        is not the first non-POSIX option to be added. Clearly some people find
441        these options useful.
443    17. If a caller to the POSIX matching function regexec() passes a non-zero
444        value for nmatch with a NULL value for pmatch, the value of
445        nmatch is forced to zero.
447    18. RunGrepTest did not have a test for the availability of the -u option of
448        the diff command, as RunTest does. It now checks in the same way as
449        RunTest, and also checks for the -b option.
451    19. If an odd number of negated classes containing just a single character
452        interposed, within parentheses, between a forward reference to a named
453        subpattern and the definition of the subpattern, compilation crashed with
454        an internal error, complaining that it could not find the referenced
455        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
456        [The bug was that it was starting one character too far in when skipping
457        over the character class, thus treating the ] as data rather than
458        terminating the class. This meant it could skip too much.]
460    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
461        /g option in pcretest when the pattern contains \K, which makes it possible
462        to have an empty string match not at the start, even when the pattern is
463        anchored. Updated pcretest and pcredemo to use this option.
465    21. If the maximum number of capturing subpatterns in a recursion was greater
466        than the maximum at the outer level, the higher number was returned, but
467        with unset values at the outer level. The correct (outer level) value is
468        now given.
470    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
471        PCRE did not set those parentheses (unlike Perl). I have now found a way to
472        make it do so. The string so far is captured, making this feature
473        compatible with Perl.
475    23. The tests have been re-organized, adding tests 11 and 12, to make it
476        possible to check the Perl 5.10 features against Perl 5.10.
478    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
479        pattern matches a fixed length string. PCRE did not allow this; now it
480        does. Neither allows recursion.
482    25. I finally figured out how to implement a request to provide the minimum
483        length of subject string that was needed in order to match a given pattern.
484        (It was back references and recursion that I had previously got hung up
485        on.) This code has now been added to pcre_study(); it finds a lower bound
486        to the length of subject needed. It is not necessarily the greatest lower
487        bound, but using it to avoid searching strings that are too short does give
488        some useful speed-ups. The value is available to calling programs via
489        pcre_fullinfo().
491    26. While implementing 25, I discovered to my embarrassment that pcretest had
492        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
493        study optimizations had never been tested with that matching function.
494        Oops. What is worse, even when it was passed study data, there was a bug in
495        pcre_dfa_exec() that meant it never actually used it. Double oops. There
496        were also very few tests of studied patterns with pcre_dfa_exec().
498    27. If (?| is used to create subpatterns with duplicate numbers, they are now
499        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
500        on the other side of the coin, they are no longer allowed to have different
501        names, because these cannot be distinguished in PCRE, and this has caused
502        confusion. (This is a difference from Perl.)
504    28. When duplicate subpattern names are present (necessarily with different
505        numbers, as required by 27 above), and a test is made by name in a
506        conditional pattern, either for a subpattern having been matched, or for
507        recursion in such a pattern, all the associated numbered subpatterns are
508        tested, and the overall condition is true if the condition is true for any
509        one of them. This is the way Perl works, and is also more like the way
510        testing by number works.
513    Version 7.9 11-Apr-09
514    ---------------------
516    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
517        (pcretest), all targets were linked against these libraries. This included
518        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
519        libraries. This caused unwanted dependencies to be created. This problem
520        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
521        pcretest is linked with readline.
523    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
524        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
525        moved outside it again, because FALSE and TRUE are already defined in AIX,
526        but BOOL is not.
528    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
529        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
531    4.  The pcregrep documentation said spaces were inserted as well as colons (or
532        hyphens) following file names and line numbers when outputting matching
533        lines. This is not true; no spaces are inserted. I have also clarified the
534        wording for the --colour (or --color) option.
536    5.  In pcregrep, when --colour was used with -o, the list of matching strings
537        was not coloured; this is different to GNU grep, so I have changed it to be
538        the same.
540    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
541        each matching line was coloured. Now it goes on to look for further matches
542        of any of the test patterns, which is the same behaviour as GNU grep.
544    7.  A pattern that could match an empty string could cause pcregrep to loop; it
545        doesn't make sense to accept an empty string match in pcregrep, so I have
546        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
547        seems to be how GNU grep behaves.
549    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
550        start or after a newline", because the conditional assertion was not being
551        correctly handled. The rule now is that both the assertion and what follows
552        in the first alternative must satisfy the test.
554    9.  If auto-callout was enabled in a pattern with a conditional group whose
555        condition was an assertion, PCRE could crash during matching, both with
556        pcre_exec() and pcre_dfa_exec().
558    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
559        used for matching.
561    11. Unicode property support in character classes was not working for
562        characters (bytes) greater than 127 when not in UTF-8 mode.
564    12. Added the -M command line option to pcretest.
566    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
568    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
570    16. Added comments and documentation about mis-use of no_arg in the C++
571        wrapper.
573    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
574        from Martin Jerabek that uses macro names for all relevant character and
575        string constants.
577    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
578        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
579        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
580        these, but not everybody uses configure.
582    19. A conditional group that had only one branch was not being correctly
583        recognized as an item that could match an empty string. This meant that an
584        enclosing group might also not be so recognized, causing infinite looping
585        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
586        with the subject "ab", where knowledge that the repeated group can match
587        nothing is needed in order to break the loop.
589    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
590        exec(), but without supplying a callout function, matching went wrong.
592    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
593        leak if the size of the offset vector was greater than 30. When the vector
594        is smaller, the saved offsets during recursion go onto a local stack
595        vector, but for larger vectors malloc() is used. It was failing to free
596        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
597        error, in fact).
599    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
600        heapframe that is used only when UTF-8 support is enabled. This caused no
601        problem, but was untidy.
603    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
604        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
605        included within another project.
607    24. Steven Van Ingelgem's patches to add more options to the CMake support,
608        slightly modified by me:
610          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
611              not building pcregrep.
613          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
614              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
616    25. Forward references, both numeric and by name, in patterns that made use of
617        duplicate group numbers, could behave incorrectly or give incorrect errors,
618        because when scanning forward to find the reference group, PCRE was not
619        taking into account the duplicate group numbers. A pattern such as
620        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
622    26. Changed a few more instances of "const unsigned char *" to USPTR, making
623        the feature of a custom pointer more persuasive (as requested by a user).
625    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
626        pcretest.c, inside #ifndefs, because it seems they are sometimes already
627        pre-defined.
629    28. Added support for (*UTF8) at the start of a pattern.
631    29. Arrange for flags added by the "release type" setting in CMake to be shown
632        in the configuration summary.
635    Version 7.8 05-Sep-08
636    ---------------------
638    1.  Replaced UCP searching code with optimized version as implemented for Ad
639        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
640        stage table and inline lookup instead of a function, giving speed ups of 2
641        to 5 times on some simple patterns that I tested. Permission was given to
642        distribute the MultiStage2.py script that generates the tables (it's not in
643        the tarball, but is in the Subversion repository).
645    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
646        scripts.
648    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
649        a group with a zero qualifier. The result of the study could be incorrect,
650        or the function might crash, depending on the pattern.
652    4.  Caseless matching was not working for non-ASCII characters in back
653        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
654        It now works when Unicode Property Support is available.
656    5.  In pcretest, an escape such as \x{de} in the data was always generating
657        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
658        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
659        truncation.
661    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
663    7.  Added two (int) casts to pcregrep when printing the difference of two
664        pointers, in case they are 64-bit values.
666    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
667        test 2 if it fails.
669    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
670        and a #define of that name to empty if it is not externally set. This is to
671        allow users of MSVC to set it if necessary.
673    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
674        the convenience functions in the pcre_get.c source file.
676    11. An option change at the start of a pattern that had top-level alternatives
677        could cause overwriting and/or a crash. This command provoked a crash in
678        some environments:
680          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
682        This potential security problem was recorded as CVE-2008-2371.
684    12. For a pattern where the match had to start at the beginning or immediately
685        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
686        pcre_dfa_exec() could read past the end of the passed subject if there was
687        no match. To help with detecting such bugs (e.g. with valgrind), I modified
688        pcretest so that it places the subject at the end of its malloc-ed buffer.
690    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
691        exec() might read past the end of the data buffer in UTF-8 mode.
693    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
694        the data contained the byte 0x85 as part of a UTF-8 character within its
695        first line. This applied both to normal and DFA matching.
697    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
698        /^[^d]*?$/8 failed to match "abc".
700    16. Added a missing copyright notice to pcrecpp_internal.h.
702    17. Make it more clear in the documentation that values returned from
703        pcre_exec() in ovector are byte offsets, not character counts.
705    18. Tidied a few places to stop certain compilers from issuing warnings.
707    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
708        supplied by Stefan Weber. I made a further small update for 7.8 because
709        there is a change of source arrangements: the pcre_searchfuncs.c module is
710        replaced by pcre_ucd.c.
713    Version 7.7 07-May-08
714    ---------------------
716    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
717        a string to a long long, pretend we don't even have a long long." This is
718        done by checking for the strtoq, strtoll, and _strtoi64 functions.
720    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
721        pre-7.6 versions, which defined a global no_arg variable instead of putting
722        it in the RE class. (See also #8 below.)
724    3.  Remove a line of dead code, identified by coverity and reported by Nuno
725        Lopes.
727    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
729        (1) The include/exclude patterns were being applied to the whole pathnames
730            of files, instead of just to the final components.
732        (2) If there was more than one level of directory, the subdirectories were
733            skipped unless they satisfied the include/exclude conditions. This is
734            inconsistent with GNU grep (and could even be seen as contrary to the
735            pcregrep specification - which I improved to make it absolutely clear).
736            The action now is always to scan all levels of directory, and just
737            apply the include/exclude patterns to regular files.
739    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
740        --exclude_dir in the tests to avoid scanning .svn directories.
742    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
743        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
744        doesn't support NULs in patterns.
746    7.  Added some missing "const"s to declarations of static tables in
747        pcre_compile.c and pcre_dfa_exec.c.
749    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
750        caused by fix #2  above. (Subsequently also a second patch to fix the
751        first patch. And a third patch - this was a messy problem.)
753    9.  Applied Craig's patch to remove the use of push_back().
755    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
756        matching function regexec().
758    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
759        which, however, unlike Perl's \g{...}, are subroutine calls, not back
760        references. PCRE supports relative numbers with this syntax (I don't think
761        Oniguruma does).
763    12. Previously, a group with a zero repeat such as (...){0} was completely
764        omitted from the compiled regex. However, this means that if the group
765        was called as a subroutine from elsewhere in the pattern, things went wrong
766        (an internal error was given). Such groups are now left in the compiled
767        pattern, with a new opcode that causes them to be skipped at execution
768        time.
770    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
771        to the way PCRE behaves:
773        (a) A lone ] character is dis-allowed (Perl treats it as data).
775        (b) A back reference to an unmatched subpattern matches an empty string
776            (Perl fails the current match path).
778        (c) A data ] in a character class must be notated as \] because if the
779            first data character in a class is ], it defines an empty class. (In
780            Perl it is not possible to have an empty class.) The empty class []
781            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
782            The negative empty class [^] matches any one character, independently
783            of the DOTALL setting.
785    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
786        non-existent subpattern following a character class starting with ']' and
787        containing () gave an internal compiling error instead of "reference to
788        non-existent subpattern". Fortunately, when the pattern did exist, the
789        compiled code was correct. (When scanning forwards to check for the
790        existencd of the subpattern, it was treating the data ']' as terminating
791        the class, so got the count wrong. When actually compiling, the reference
792        was subsequently set up correctly.)
794    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
795        it was being rejected as not supported by pcre_dfa_exec(), even though
796        other assertions are supported. I have made pcre_dfa_exec() support
797        (*FAIL).
799    16. The implementation of 13c above involved the invention of a new opcode,
800        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
801        cannot be changed at match time, I realized I could make a small
802        improvement to matching performance by compiling OP_ALLANY instead of
803        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
804        on the OP_ANY path.
806    17. Compiling pcretest on Windows with readline support failed without the
807        following two fixes: (1) Make the unistd.h include conditional on
808        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
810    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
811        ncurses library to be included for pcretest when ReadLine support is
812        requested, but also to allow for it to be overridden. This patch came from
813        Daniel Bergström.
815    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
816        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
817        any errors with the current Unicode tables. Thanks to Peter Kankowski for
818        spotting this.
821    Version 7.6 28-Jan-08
822    ---------------------
824    1.  A character class containing a very large number of characters with
825        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
826        overflow.
828    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
829        HAVE_LONG_LONG is not defined.
831    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
832        bring them up to date and include new features. This patch includes:
834        - Fixed PH's badly added libz and libbz2 support.
835        - Fixed a problem with static linking.
836        - Added pcredemo. [But later removed - see 7 below.]
837        - Fixed dftables problem and added an option.
838        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
839            HAVE_LONG_LONG.
840        - Added readline support for pcretest.
841        - Added an listing of the option settings after cmake has run.
843    4.  A user submitted a patch to Makefile that makes it easy to create
844        "pcre.dll" under mingw when using Configure/Make. I added stuff to
845        Makefile.am that cause it to include this special target, without
846        affecting anything else. Note that the same mingw target plus all
847        the other distribution libraries and programs are now supported
848        when configuring with CMake (see 6 below) instead of with
849        Configure/Make.
851    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
852        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
853        exported in the Windows port". It has not yet been confirmed that the patch
854        solves the problem, but it does no harm.
856    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
857        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
858        with CMake, and also correct the comment about stack recursion.
860    7.  Remove the automatic building of pcredemo from the ./configure system and
861        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
862        of a program that users should build themselves after PCRE is installed, so
863        building it automatically is not really right. What is more, it gave
864        trouble in some build environments.
866    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
869    Version 7.5 10-Jan-08
870    ---------------------
872    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
873        values in parens when parsing an RE using the C++ wrapper."
875    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
876        Characters greater than 255 were excluded from the class instead of being
877        included.
879    3.  The same bug as (2) above applied to negated POSIX classes such as
880        [:^space:].
882    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
883        defined or documented. It seems to have been a typo for PCRE_STATIC, so
884        I have changed it.
886    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
887        first named subpattern) and a construct such as (?&a) would reference the
888        first named subpattern whose name started with "a" (in other words, the
889        length check was missing). Both these problems are fixed. "Subpattern name
890        expected" is now given for (?&) (a zero-length name), and this patch also
891        makes it give the same error for \k'' (previously it complained that that
892        was a reference to a non-existent subpattern).
894    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
895        this is right because (?- can be followed by option settings as well as by
896        digits. I have, however, made the messages clearer.
898    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
899        than the number used in the conditional) now cause a compile-time error.
900        This is actually not compatible with Perl, which accepts such patterns, but
901        treats the conditional as always being FALSE (as PCRE used to), but it
902        seems to me that giving a diagnostic is better.
904    8.  Change "alphameric" to the more common word "alphanumeric" in comments
905        and messages.
907    9.  Fix two occurrences of "backslash" in comments that should have been
908        "backspace".
910    10. Remove two redundant lines of code that can never be obeyed (their function
911        was moved elsewhere).
913    11. The program that makes PCRE's Unicode character property table had a bug
914        which caused it to generate incorrect table entries for sequences of
915        characters that have the same character type, but are in different scripts.
916        It amalgamated them into a single range, with the script of the first of
917        them. In other words, some characters were in the wrong script. There were
918        thirteen such cases, affecting characters in the following ranges:
920          U+002b0 - U+002c1
921          U+0060c - U+0060d
922          U+0061e - U+00612
923          U+0064b - U+0065e
924          U+0074d - U+0076d
925          U+01800 - U+01805
926          U+01d00 - U+01d77
927          U+01d9b - U+01dbf
928          U+0200b - U+0200f
929          U+030fc - U+030fe
930          U+03260 - U+0327f
931          U+0fb46 - U+0fbb1
932          U+10450 - U+1049d
934    12. The -o option (show only the matching part of a line) for pcregrep was not
935        compatible with GNU grep in that, if there was more than one match in a
936        line, it showed only the first of them. It now behaves in the same way as
937        GNU grep.
939    13. If the -o and -v options were combined for pcregrep, it printed a blank
940        line for every non-matching line. GNU grep prints nothing, and pcregrep now
941        does the same. The return code can be used to tell if there were any
942        non-matching lines.
944    14. Added --file-offsets and --line-offsets to pcregrep.
946    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
947        infinitely looping recursion. The bug was that positive lookaheads were not
948        being skipped when checking for a possible empty match (negative lookaheads
949        and both kinds of lookbehind were skipped).
951    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
952        inclusion of <windows.h> to before rather than after the definition of
953        INVALID_FILE_ATTRIBUTES (patch from David Byron).
955    17. Specifying a possessive quantifier with a specific limit for a Unicode
956        character property caused pcre_compile() to compile bad code, which led at
957        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
958        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
959        caused the error; without that there was no problem.
961    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
963    19. Added --enable-pcretest-libreadline.
965    20. In pcrecpp.cc, the variable 'count' was incremented twice in
966        RE::GlobalReplace(). As a result, the number of replacements returned was
967        double what it should be. I removed one of the increments, but Craig sent a
968        later patch that removed the other one (the right fix) and added unit tests
969        that check the return values (which was not done before).
971    21. Several CMake things:
973        (1) Arranged that, when cmake is used on Unix, the libraries end up with
974            the names libpcre and libpcreposix, not just pcre and pcreposix.
976        (2) The above change means that pcretest and pcregrep are now correctly
977            linked with the newly-built libraries, not previously installed ones.
981    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
982        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
983        UTF-8 newline character). The key issue is that the pattern starts .*;
984        this means that the match must be either at the beginning, or after a
985        newline. The bug was in the code for advancing after a failed match and
986        checking that the new position followed a newline. It was not taking
987        account of UTF-8 characters correctly.
989    23. PCRE was behaving differently from Perl in the way it recognized POSIX
990        character classes. PCRE was not treating the sequence [:...:] as a
991        character class unless the ... were all letters. Perl, however, seems to
992        allow any characters between [: and :], though of course it rejects as
993        unknown any "names" that contain non-letters, because all the known class
994        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
995        for example, whereas PCRE did not - it did not recognize a POSIX character
996        class. This seemed a bit dangerous, so the code has been changed to be
997        closer to Perl. The behaviour is not identical to Perl, because PCRE will
998        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
999        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1000        Perl does, and where it didn't before.
1002    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1003        Windows environments %n is disabled by default.
1006    Version 7.4 21-Sep-07
1007    ---------------------
1009    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1010        means that a class such as [\s] counted as "explicit reference to CR or
1011        LF". That isn't really right - the whole point of the change was to try to
1012        help when there was an actual mention of one of the two characters. So now
1013        the change happens only if \r or \n (or a literal CR or LF) character is
1014        encountered.
1016    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1017        of both had grown to the point where there were only 3 bits left.
1018        Fortunately, there was spare space in the data structure, and so I have
1019        moved the internal flags into a new 16-bit field to free up more option
1020        bits.
1022    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1023        but did not set the internal JCHANGED flag - either of these is enough to
1024        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1025        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1026        start sets both bits.
1028    4.  Added options (at build time, compile time, exec time) to change \R from
1029        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1031    5.  doc/pcresyntax.html was missing from the distribution.
1033    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1034        compatibility, even though it is no longer used.
1036    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1037        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1038        windows.h file is present (where different names are used). [This was
1039        reversed later after testing - see 16 below.]
1041    8.  Changed all #include <config.h> to #include "config.h". There were also
1042        some further <pcre.h> cases that I changed to "pcre.h".
1044    9.  When pcregrep was used with the --colour option, it missed the line ending
1045        sequence off the lines that it output.
1047    10. It was pointed out to me that arrays of string pointers cause lots of
1048        relocations when a shared library is dynamically loaded. A technique of
1049        using a single long string with a table of offsets can drastically reduce
1050        these. I have refactored PCRE in four places to do this. The result is
1051        dramatic:
1053          Originally:                          290
1054          After changing UCP table:            187
1055          After changing error message table:   43
1056          After changing table of "verbs"       36
1057          After changing table of Posix names   22
1059        Thanks to the folks working on Gregex for glib for this insight.
1061    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1062        unicode-properties was also set.
1064    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1066    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1067        checked only for CRLF.
1069    14. Added casts to pcretest.c to avoid compiler warnings.
1071    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1073    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1074        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1075        entirely. This removes changes made in 7 above.
1077    17. The CMake files have been updated, and there is now more information about
1078        building with CMake in the NON-UNIX-USE document.
1081    Version 7.3 28-Aug-07
1082    ---------------------
1084     1. In the rejigging of the build system that eventually resulted in 7.1, the
1085        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1086        brackets there is not right, since it causes compilers to look for an
1087        installed pcre.h, not the version that is in the source that is being
1088        compiled (which of course may be different). I have changed it back to:
1090          #include "pcre.h"
1092        I have a vague recollection that the change was concerned with compiling in
1093        different directories, but in the new build system, that is taken care of
1094        by the VPATH setting the Makefile.
1096     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1097        when the subject happened to end in the byte 0x85 (e.g. if the last
1098        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1099        characters but of course it shouldn't be taken as a newline when it is part
1100        of another character. The bug was that, for an unlimited repeat of . in
1101        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1102        characters when looking for a newline.
1104     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1106     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1107        in debug output.
1109     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1110        long printing in the pcrecpp unittest when running under MinGW.
1112     6. ESC_K was left out of the EBCDIC table.
1114     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1115        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1116        limit also applies to "virtual nesting" when a pattern is recursive, and in
1117        this case 1000 isn't so big. I have been able to remove this limit at the
1118        expense of backing off one optimization in certain circumstances. Normally,
1119        when pcre_exec() would call its internal match() function recursively and
1120        immediately return the result unconditionally, it uses a "tail recursion"
1121        feature to save stack. However, when a subpattern that can match an empty
1122        string has an unlimited repetition quantifier, it no longer makes this
1123        optimization. That gives it a stack frame in which to save the data for
1124        checking that an empty string has been matched. Previously this was taken
1125        from the 1000-entry workspace that had been reserved. So now there is no
1126        explicit limit, but more stack is used.
1128     8. Applied Daniel's patches to solve problems with the import/export magic
1129        syntax that is required for Windows, and which was going wrong for the
1130        pcreposix and pcrecpp parts of the library. These were overlooked when this
1131        problem was solved for the main library.
1133     9. There were some crude static tests to avoid integer overflow when computing
1134        the size of patterns that contain repeated groups with explicit upper
1135        limits. As the maximum quantifier is 65535, the maximum group length was
1136        set at 30,000 so that the product of these two numbers did not overflow a
1137        32-bit integer. However, it turns out that people want to use groups that
1138        are longer than 30,000 bytes (though not repeat them that many times).
1139        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1140        made it possible to implement the integer overflow checks in a much more
1141        dynamic way, which I have now done. The artificial limitation on group
1142        length has been removed - we now have only the limit on the total length of
1143        the compiled pattern, which depends on the LINK_SIZE setting.
1145    10. Fixed a bug in the documentation for get/copy named substring when
1146        duplicate names are permitted. If none of the named substrings are set, the
1147        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1148        empty string.
1150    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1151        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1152        because the ] is interpreted as the first data character and the
1153        terminating ] is not found. PCRE has been made compatible with Perl in this
1154        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1155        cause memory overwriting.
1157    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1158        string has been matched (to stop an infinite loop). It was not recognizing
1159        a conditional subpattern that could match an empty string if that
1160        subpattern was within another subpattern. For example, it looped when
1161        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1162        condition was not nested. This bug has been fixed.
1164    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1165        past the start of the subject in the presence of bytes with the top bit
1166        set, for example "\x8aBCD".
1168    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1169        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1171    14. Optimized (?!) to (*FAIL).
1173    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1174        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1175        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1176        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1177        does: it's just the validity check that is more restrictive.
1179    16. Inserted checks for integer overflows during escape sequence (backslash)
1180        processing, and also fixed erroneous offset values for syntax errors during
1181        backslash processing.
1183    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1184        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1186    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1187        caused an overrun.
1189    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1190        something other than just ASCII characters) inside a group that had an
1191        unlimited repeat caused a loop at compile time (while checking to see
1192        whether the group could match an empty string).
1194    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1195        [\P{Any}] did so. (Error in the code for printing property names.)
1197    21. An orphan \E inside a character class could cause a crash.
1199    22. A repeated capturing bracket such as (A)? could cause a wild memory
1200        reference during compilation.
1202    23. There are several functions in pcre_compile() that scan along a compiled
1203        expression for various reasons (e.g. to see if it's fixed length for look
1204        behind). There were bugs in these functions when a repeated \p or \P was
1205        present in the pattern. These operators have additional parameters compared
1206        with \d, etc, and these were not being taken into account when moving along
1207        the compiled data. Specifically:
1209        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1210            length.
1212        (b) An item such as \pL+ within a repeated group could cause crashes or
1213            loops.
1215        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1216            "reference to non-existent subpattern" error.
1218        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1220    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1221        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1223    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1225    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1226        character were causing crashes (broken optimization).
1228    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1229        \p or \P) caused a compile-time loop.
1231    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1232        break. For example, the unstudied pattern [\r\n]A does not match the string
1233        "\r\nA" because change 7.0/46 below moves the current point on by two
1234        characters after failing to match at the start. However, the pattern \nA
1235        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1236        the same is true. There doesn't seem any very clean way out of this, but
1237        what I have chosen to do makes the common cases work: PCRE now takes note
1238        of whether there can be an explicit match for \r or \n anywhere in the
1239        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1240        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1241        pattern has explicit CR or LF references.
1243    29. Added (*CR) etc for changing newline setting at start of pattern.
1246    Version 7.2 19-Jun-07
1247    ---------------------
1249     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1250        which is apparently normally available under Windows.
1252     2. Re-jig the pcregrep tests with different newline settings in an attempt
1253        to make them independent of the local environment's newline setting.
1255     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1257     4. Some of the "internals" tests were previously cut out when the link size
1258        was not 2, because the output contained actual offsets. The recent new
1259        "Z" feature of pcretest means that these can be cut out, making the tests
1260        usable with all link sizes.
1262     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1263        stack recursion. This gives a massive performance boost under BSD, but just
1264        a small improvement under Linux. However, it saves one field in the frame
1265        in all cases.
1267     6. Added more features from the forthcoming Perl 5.10:
1269        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1270            recursion call. It refers to the nth most recently opened parentheses.
1272        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1273            to be opened parentheses.
1275        (c) Conditions that refer to capturing parentheses can be specified
1276            relatively, for example, (?(-2)... or (?(+3)...
1278        (d) \K resets the start of the current match so that everything before
1279            is not part of it.
1281        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1283        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1284            reference syntax.
1286        (g) (?| introduces a group in which the numbering of parentheses in each
1287            alternative starts with the same number.
1289        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1291     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1294     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1295        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1296        for detecting groups that can match an empty string.
1298     9. A pattern with a very large number of alternatives (more than several
1299        hundred) was running out of internal workspace during the pre-compile
1300        phase, where pcre_compile() figures out how much memory will be needed. A
1301        bit of new cunning has reduced the workspace needed for groups with
1302        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1303        workspace instead of running out of the 4096 that are available.
1305    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1307    11. Applied patch from Google to remove an optimization that didn't quite work.
1308        The report of the bug said:
1310          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1311          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1312          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1314    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1315        it matched the wrong number of bytes.
1318    Version 7.1 24-Apr-07
1319    ---------------------
1321     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1322        that is more "standard", making use of automake and other Autotools. There
1323        is some re-arrangement of the files and adjustment of comments consequent
1324        on this.
1326     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1327        for recursive directory scanning broke on some systems because the files
1328        are not scanned in any specific order and on different systems the order
1329        was different. A call to "sort" has been inserted into RunGrepTest for the
1330        approprate test as a short-term fix. In the longer term there may be an
1331        alternative.
1333     3. I had an email from Eric Raymond about problems translating some of PCRE's
1334        man pages to HTML (despite the fact that I distribute HTML pages, some
1335        people do their own conversions for various reasons). The problems
1336        concerned the use of low-level troff macros .br and .in. I have therefore
1337        removed all such uses from the man pages (some were redundant, some could
1338        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1339        HTML has been updated to handle .nf/.fi and to complain if it encounters
1340        .br or .in.
1342     4. Updated comments in configure.ac that get placed in config.h.in and also
1343        arranged for config.h to be included in the distribution, with the name
1344        config.h.generic, for the benefit of those who have to compile without
1345        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1347     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1348        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1349        makevp.bat for the current PCRE, using the additional files
1350        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1352     6. A Windows user reported a minor discrepancy with test 2, which turned out
1353        to be caused by a trailing space on an input line that had got lost in his
1354        copy. The trailing space was an accident, so I've just removed it.
1356     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1357        that is needed.
1359     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1360        as "const" (a) because they are and (b) because it helps the PHP
1361        maintainers who have recently made a script to detect big data structures
1362        in the php code that should be moved to the .rodata section. I remembered
1363        to update Builducptable as well, so it won't revert if ucptable.h is ever
1364        re-created.
1366     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1367        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1368        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1369        support is not required. This saves 1.5-2K of code, which is important in
1370        some applications.
1372        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1373        so as not to refer to the tables, even though these functions will never be
1374        called when UTF-8 support is disabled. Otherwise there are problems with a
1375        shared library.
1377    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1379        (a) It was defining its arguments as char * instead of void *.
1381        (b) It was assuming that all moves were upwards in memory; this was true
1382            a long time ago when I wrote it, but is no longer the case.
1384        The emulated memove() is provided for those environments that have neither
1385        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1386        is clearly not the case, as these two bugs were recently reported.
1388    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1389        and Detrail to create the HTML documentation, the .txt form of the man
1390        pages, and it removes trailing spaces from listed files. It also creates
1391        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1392        case, it wraps all the #defines with #ifndefs. This script should be run
1393        before "make dist".
1395    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1396        with Unicode property support.
1398        (a) For a maximizing quantifier, if the two different cases of the
1399            character were of different lengths in their UTF-8 codings (there are
1400            some cases like this - I found 11), and the matching function had to
1401            back up over a mixture of the two cases, it incorrectly assumed they
1402            were both the same length.
1404        (b) When PCRE was configured to use the heap rather than the stack for
1405            recursion during matching, it was not correctly preserving the data for
1406            the other case of a UTF-8 character when checking ahead for a match
1407            while processing a minimizing repeat. If the check also involved
1408            matching a wide character, but failed, corruption could cause an
1409            erroneous result when trying to check for a repeat of the original
1410            character.
1412    13. Some tidying changes to the testing mechanism:
1414        (a) The RunTest script now detects the internal link size and whether there
1415            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1416            values substituted by "configure". (The RunGrepTest script already did
1417            this for UTF-8.) The configure.ac script no longer substitutes the
1418            relevant variables.
1420        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1421            with length and offset values. This means that the output is different
1422            for different internal link sizes. Test 2 is skipped for link sizes
1423            other than 2 because of this, bypassing the problem. Unfortunately,
1424            there was also a test in test 3 (the locale tests) that used /B and
1425            failed for link sizes other than 2. Rather than cut the whole test out,
1426            I have added a new /Z option to pcretest that replaces the length and
1427            offset values with spaces. This is now used to make test 3 independent
1428            of link size. (Test 2 will be tidied up later.)
1430    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1431        segmentation fault instead of returning the appropriate error message.
1433    15. In multiline mode when the newline sequence was set to "any", the pattern
1434        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1435        This doesn't seem right; it now treats the CRLF combination as the line
1436        ending, and so does not match in that case. It's only a pattern such as ^$
1437        that would hit this one: something like ^ABC$ would have failed after \r
1438        and then tried again after \r\n.
1440    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1441        in an attempt to make files that differ only in their line terminators
1442        compare equal. This works on Linux.
1444    17. Under certain error circumstances pcregrep might try to free random memory
1445        as it exited. This is now fixed, thanks to valgrind.
1447    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1448        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1449        was because its rules for how to advance for /g after matching an empty
1450        string at the end of a line did not allow for this case. They now check for
1451        it specially.
1453    20. pcretest is supposed to handle patterns and data of any length, by
1454        extending its buffers when necessary. It was getting this wrong when the
1455        buffer for a data line had to be extended.
1457    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1458        CRLF as a newline sequence.
1460    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1461        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1462        I have nevertheless tidied it up.
1464    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1466    24. Added a man page for pcre-config.
1469    Version 7.0 19-Dec-06
1470    ---------------------
1472     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1473        moving to gcc 4.1.1.
1475     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1476        sys/time.h, which is documented as needed for this function. It doesn't
1477        seem to matter on Linux, but it showed up on some releases of OS X.
1479     3. It seems that there are systems where bytes whose values are greater than
1480        127 match isprint() in the "C" locale. The "C" locale should be the
1481        default when a C program starts up. In most systems, only ASCII printing
1482        characters match isprint(). This difference caused the output from pcretest
1483        to vary, making some of the tests fail. I have changed pcretest so that:
1485        (a) When it is outputting text in the compiled version of a pattern, bytes
1486            other than 32-126 are always shown as hex escapes.
1488        (b) When it is outputting text that is a matched part of a subject string,
1489            it does the same, unless a different locale has been set for the match
1490            (using the /L modifier). In this case, it uses isprint() to decide.
1492     4. Fixed a major bug that caused incorrect computation of the amount of memory
1493        required for a compiled pattern when options that changed within the
1494        pattern affected the logic of the preliminary scan that determines the
1495        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1496        that the computed length was too small. The symptoms of this bug were
1497        either the PCRE error "internal error: code overflow" from pcre_compile(),
1498        or a glibc crash with a message such as "pcretest: free(): invalid next
1499        size (fast)". Examples of patterns that provoked this bug (shown in
1500        pcretest format) are:
1502          /(?-x: )/x
1503          /(?x)(?-x: \s*#\s*)/
1504          /((?i)[\x{c0}])/8
1505          /(?i:[\x{c0}])/8
1507        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1508        is now done differently.
1510     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1511        wrapper classes; (b) implement a new function in the C++ scanner that is
1512        more efficient than the old way of doing things because it avoids levels of
1513        recursion in the regex matching; (c) add a paragraph to the documentation
1514        for the FullMatch() function.
1516     6. The escape sequence \n was being treated as whatever was defined as
1517        "newline". Not only was this contrary to the documentation, which states
1518        that \n is character 10 (hex 0A), but it also went horribly wrong when
1519        "newline" was defined as CRLF. This has been fixed.
1521     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1522        was being set to -1 for the "end of line" case (supposedly a value that no
1523        character can have). Though this value is never used (the check for end of
1524        line is "zero bytes in current character"), it caused compiler complaints.
1525        I've changed it to 0xffffffff.
1527     8. In pcre_version.c, the version string was being built by a sequence of
1528        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1529        string (as it is for production releases) called a macro with an empty
1530        argument. The C standard says the result of this is undefined. The gcc
1531        compiler treats it as an empty string (which was what was wanted) but it is
1532        reported that Visual C gives an error. The source has been hacked around to
1533        avoid this problem.
1535     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1536        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1537        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1538        of them did).
1540    10. Originally, pcretest opened its input and output without "b"; then I was
1541        told that "b" was needed in some environments, so it was added for release
1542        5.0 to both the input and output. (It makes no difference on Unix-like
1543        systems.) Later I was told that it is wrong for the input on Windows. I've
1544        now abstracted the modes into two macros, to make it easier to fiddle with
1545        them, and removed "b" from the input mode under Windows.
1547    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1549    12. Added -help and --help to pcretest as an official way of being reminded
1550        of the options.
1552    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1553        and pcrecpp.cc because they annoy compilers at high warning levels.
1555    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1557    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1558        scripts are not C programs :-) and which was not noticed because it works
1559        on Linux.
1561    16. pcretest is supposed to handle any length of pattern and data line (as one
1562        line or as a continued sequence of lines) by extending its input buffer if
1563        necessary. This feature was broken for very long pattern lines, leading to
1564        a string of junk being passed to pcre_compile() if the pattern was longer
1565        than about 50K.
1567    17. I have done a major re-factoring of the way pcre_compile() computes the
1568        amount of memory needed for a compiled pattern. Previously, there was code
1569        that made a preliminary scan of the pattern in order to do this. That was
1570        OK when PCRE was new, but as the facilities have expanded, it has become
1571        harder and harder to keep it in step with the real compile phase, and there
1572        have been a number of bugs (see for example, 4 above). I have now found a
1573        cunning way of running the real compile function in a "fake" mode that
1574        enables it to compute how much memory it would need, while actually only
1575        ever using a few hundred bytes of working memory and without too many
1576        tests of the mode. This should make future maintenance and development
1577        easier. A side effect of this work is that the limit of 200 on the nesting
1578        depth of parentheses has been removed (though this was never a serious
1579        limitation, I suspect). However, there is a downside: pcre_compile() now
1580        runs more slowly than before (30% or more, depending on the pattern). I
1581        hope this isn't a big issue. There is no effect on runtime performance.
1583    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1584        newline (only possible for the last line of a file) and it was a
1585        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1587    19. Added additional timing features to pcretest. (1) The -tm option now times
1588        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1589        separate command line item, by a number that specifies the number of
1590        repeats to use when timing. The default is 50000; this gives better
1591        precision, but takes uncomfortably long for very large patterns.
1593    20. Extended pcre_study() to be more clever in cases where a branch of a
1594        subpattern has no definite first character. For example, (a*|b*)[cd] would
1595        previously give no result from pcre_study(). Now it recognizes that the
1596        first character must be a, b, c, or d.
1598    21. There was an incorrect error "recursive call could loop indefinitely" if
1599        a subpattern (or the entire pattern) that was being tested for matching an
1600        empty string contained only one non-empty item after a nested subpattern.
1601        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1602        incorrectly, because the \d was being skipped in the check.
1604    22. The pcretest program now has a new pattern option /B and a command line
1605        option -b, which is equivalent to adding /B to every pattern. This causes
1606        it to show the compiled bytecode, without the additional information that
1607        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1608        is the same as /B/I).
1610    23. A new optimization is now able automatically to treat some sequences such
1611        as a*b as a*+b. More specifically, if something simple (such as a character
1612        or a simple class like \d) has an unlimited quantifier, and is followed by
1613        something that cannot possibly match the quantified thing, the quantifier
1614        is automatically "possessified".
1616    24. A recursive reference to a subpattern whose number was greater than 39
1617        went wrong under certain circumstances in UTF-8 mode. This bug could also
1618        have affected the operation of pcre_study().
1620    25. Realized that a little bit of performance could be had by replacing
1621        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1623    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1625    27. Possessive quantifiers such as a++ were previously implemented by turning
1626        them into atomic groups such as ($>a+). Now they have their own opcodes,
1627        which improves performance. This includes the automatically created ones
1628        from 23 above.
1630    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1631        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1632        the first matched character to be a colon. This applied both to named and
1633        numbered groups.
1635    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1637    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1638        building PCRE on Windows easier, so I have included it in the distribution.
1640    31. There is now a check in pcretest against a ridiculously large number being
1641        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1642        loop, the loop is abandoned.
1644    32. Forward references to subpatterns in conditions such as (?(2)...) where
1645        subpattern 2 is defined later cause pcre_compile() to search forwards in
1646        the pattern for the relevant set of parentheses. This search went wrong
1647        when there were unescaped parentheses in a character class, parentheses
1648        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1650    33. "Subroutine" calls and backreferences were previously restricted to
1651        referencing subpatterns earlier in the regex. This restriction has now
1652        been removed.
1654    34. Added a number of extra features that are going to be in Perl 5.10. On the
1655        whole, these are just syntactic alternatives for features that PCRE had
1656        previously implemented using the Python syntax or my own invention. The
1657        other formats are all retained for compatibility.
1659        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1660            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1661            also .NET compatible.
1663        (b) A recursion or subroutine call to a named group can now be defined as
1664            (?&name) as well as (?P>name).
1666        (c) A backreference to a named group can now be defined as \k<name> or
1667            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1668            5.10, are also .NET compatible.
1670        (d) A conditional reference to a named group can now use the syntax
1671            (?(<name>) or (?('name') as well as (?(name).
1673        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1674            groups (named and numbered) that are never evaluated inline, but can be
1675            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1676            is always false. There may be only one alternative in such a group.
1678        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1679            as the simple (?(R). The condition is true only if the most recent
1680            recursion is that of the given number or name. It does not search out
1681            through the entire recursion stack.
1683        (g) The escape \gN or \g{N} has been added, where N is a positive or
1684            negative number, specifying an absolute or relative reference.
1686    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1687        some "unreachable code" warnings.
1689    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1690        things, this adds five new scripts.
1692    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1693        There were also incompatibilities regarding the handling of \Q..\E inside
1694        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1695        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1697    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1698        matches an empty string, and forcibly breaks the loop. There were bugs in
1699        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1700        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1701        separate and independent bugs (that affected different cases) have been
1702        fixed.
1704    39. Refactored the code to abolish the use of different opcodes for small
1705        capturing bracket numbers. This is a tidy that I avoided doing when I
1706        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1707        The new approach is not only tidier, it makes it possible to reduce the
1708        memory needed to fix the previous bug (38).
1710    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1711        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1712        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1713        mode.
1715    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1716        report.
1718    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1719        copy construction and assignment in the C++ wrapper.
1721    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1722        couple of bugs that could have given wrong results in the "--newline=crlf"
1723        case.
1725    44. Added a number of casts and did some reorganization of signed/unsigned int
1726        variables following suggestions from Dair Grant. Also renamed the variable
1727        "this" as "item" because it is a C++ keyword.
1729    45. Arranged for dftables to add
1731          #include "pcre_internal.h"
1733        to pcre_chartables.c because without it, gcc 4.x may remove the array
1734        definition from the final binary if PCRE is built into a static library and
1735        dead code stripping is activated.
1737    46. For an unanchored pattern, if a match attempt fails at the start of a
1738        newline sequence, and the newline setting is CRLF or ANY, and the next two
1739        characters are CRLF, advance by two characters instead of one.
1742    Version 6.7 04-Jul-06
1743    ---------------------
1745     1. In order to handle tests when input lines are enormously long, pcretest has
1746        been re-factored so that it automatically extends its buffers when
1747        necessary. The code is crude, but this _is_ just a test program. The
1748        default size has been increased from 32K to 50K.
1750     2. The code in pcre_study() was using the value of the re argument before
1751        testing it for NULL. (Of course, in any sensible call of the function, it
1752        won't be NULL.)
1754     3. The memmove() emulation function in pcre_internal.h, which is used on
1755        systems that lack both memmove() and bcopy() - that is, hardly ever -
1756        was missing a "static" storage class specifier.
1758     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1759        containing an extended class (one that cannot be represented by a bitmap
1760        because it contains high-valued characters or Unicode property items, e.g.
1761        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1762        pattern, but PCRE should not loop if you do not (it no longer does).
1763        [Detail: two cases were found: (a) a repeated subpattern containing an
1764        extended class; (b) a recursive reference to a subpattern that followed a
1765        previous extended class. It wasn't skipping over the extended class
1766        correctly when UTF-8 mode was not set.]
1768     5. A negated single-character class was not being recognized as fixed-length
1769        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1770        compile error "lookbehind assertion is not fixed length".
1772     6. The RunPerlTest auxiliary script was showing an unexpected difference
1773        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1774        write a Perl script that can interpret lines of an input file either as
1775        byte characters or as UTF-8, which is what "perltest" was being required to
1776        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1777        can't do is switch easily at run time between having the "use utf8;" pragma
1778        or not. In the end, I fudged it by using the RunPerlTest script to insert
1779        "use utf8;" explicitly for the UTF-8 tests.
1781     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1782        the end of the subject string, contrary to the documentation and to what
1783        Perl does. This was true of both matching functions. Now it matches only at
1784        the start of the subject and immediately after *internal* newlines.
1786     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1787        a pointer to an int instead of a pointer to an unsigned long int. This
1788        caused problems on 64-bit systems.
1790     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1791        instance of the 'standard' template library not being so standard".
1793    10. There was no check on the number of named subpatterns nor the maximum
1794        length of a subpattern name. The product of these values is used to compute
1795        the size of the memory block for a compiled pattern. By supplying a very
1796        long subpattern name and a large number of named subpatterns, the size
1797        computation could be caused to overflow. This is now prevented by limiting
1798        the length of names to 32 characters, and the number of named subpatterns
1799        to 10,000.
1801    11. Subpatterns that are repeated with specific counts have to be replicated in
1802        the compiled pattern. The size of memory for this was computed from the
1803        length of the subpattern and the repeat count. The latter is limited to
1804        65535, but there was no limit on the former, meaning that integer overflow
1805        could in principle occur. The compiled length of a repeated subpattern is
1806        now limited to 30,000 bytes in order to prevent this.
1808    12. Added the optional facility to have named substrings with the same name.
1810    13. Added the ability to use a named substring as a condition, using the
1811        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1812        are numbers (not recommended). Forward references are permitted.
1814    14. Added forward references in named backreferences (if you see what I mean).
1816    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1817        pattern could run off the end of the subject. For example, the pattern
1818        "(?s)(.{1,5})"8 did this with the subject "ab".
1820    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1821        PCRE_CASELESS was set when matching characters that were quantified with ?
1822        or *.
1824    17. A character class other than a single negated character that had a minimum
1825        but no maximum quantifier - for example [ab]{6,} - was not handled
1826        correctly by pce_dfa_exec(). It would match only one character.
1828    18. A valid (though odd) pattern that looked like a POSIX character
1829        class but used an invalid character after [ (for example [[,abc,]]) caused
1830        pcre_compile() to give the error "Failed: internal error: code overflow" or
1831        in some cases to crash with a glibc free() error. This could even happen if
1832        the pattern terminated after [[ but there just happened to be a sequence of
1833        letters, a binary zero, and a closing ] in the memory that followed.
1835    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1836        over the years. Originally (before any Unicode support), just the bottom 8
1837        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1838        output from "man perlunicode" includes this:
1840          The regular expression compiler produces polymorphic opcodes.  That
1841          is, the pattern adapts to the data and automatically switches to
1842          the Unicode character scheme when presented with Unicode data--or
1843          instead uses a traditional byte scheme when presented with byte
1844          data.
1846        Sadly, a wide octal escape does not cause a switch, and in a string with
1847        no other multibyte characters, these octal escapes are treated as before.
1848        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1849        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1850        Unicode string.
1852        I have not perpetrated such confusion in PCRE. Up till now, it took just
1853        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1854        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1855        translate to the appropriate multibyte character.
1857    29. Applied some refactoring to reduce the number of warnings from Microsoft
1858        and Borland compilers. This has included removing the fudge introduced
1859        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1860        a warning about an unused variable.
1862    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1863        characters since release 4.0, because Perl (from release 5.004) does not.
1864        [Or at least, is documented not to: some releases seem to be in conflict
1865        with the documentation.] However, when a pattern was studied with
1866        pcre_study() and all its branches started with \s, PCRE still included VT
1867        as a possible starting character. Of course, this did no harm; it just
1868        caused an unnecessary match attempt.
1870    22. Removed a now-redundant internal flag bit that recorded the fact that case
1871        dependency changed within the pattern. This was once needed for "required
1872        byte" processing, but is no longer used. This recovers a now-scarce options
1873        bit. Also moved the least significant internal flag bit to the most-
1874        significant bit of the word, which was not previously used (hangover from
1875        the days when it was an int rather than a uint) to free up another bit for
1876        the future.
1878    23. Added support for CRLF line endings as well as CR and LF. As well as the
1879        default being selectable at build time, it can now be changed at runtime
1880        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1881        specify that it is scanning data with non-default line endings.
1883    24. Changed the definition of CXXLINK to make it agree with the definition of
1884        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1886    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1887        recursions. This makes a big different to stack usage for some patterns.
1889    26. If a subpattern containing a named recursion or subroutine reference such
1890        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1891        the space required for the compiled pattern went wrong and gave too small a
1892        value. Depending on the environment, this could lead to "Failed: internal
1893        error: code overflow at offset 49" or "glibc detected double free or
1894        corruption" errors.
1896    27. Applied patches from Google (a) to support the new newline modes and (b) to
1897        advance over multibyte UTF-8 characters in GlobalReplace.
1899    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1900        difference for some implementation of PCRE in some Windows version.
1902    29. Added some extra testing facilities to pcretest:
1904        \q<number>   in a data line sets the "match limit" value
1905        \Q<number>   in a data line sets the "match recursion limt" value
1906        -S <number>  sets the stack size, where <number> is in megabytes
1908        The -S option isn't available for Windows.
1911    Version 6.6 06-Feb-06
1912    ---------------------
1914     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1915        in pcreposix.h. I have copied the definition from pcre.h.
1917     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1918        because pcre.h is no longer a built file.
1920     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1921        not normally included in the compiled code.
1924    Version 6.5 01-Feb-06
1925    ---------------------
1927     1. When using the partial match feature with pcre_dfa_exec(), it was not
1928        anchoring the second and subsequent partial matches at the new starting
1929        point. This could lead to incorrect results. For example, with the pattern
1930        /1234/, partially matching against "123" and then "a4" gave a match.
1932     2. Changes to pcregrep:
1934        (a) All non-match returns from pcre_exec() were being treated as failures
1935            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1936            error message is output. Some extra information is given for the
1937            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1938            probably the only errors that are likely to be caused by users (by
1939            specifying a regex that has nested indefinite repeats, for instance).
1940            If there are more than 20 of these errors, pcregrep is abandoned.
1942        (b) A binary zero was treated as data while matching, but terminated the
1943            output line if it was written out. This has been fixed: binary zeroes
1944            are now no different to any other data bytes.
1946        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1947            used to set a locale for matching. The --locale=xxxx long option has
1948            been added (no short equivalent) to specify a locale explicitly on the
1949            pcregrep command, overriding the environment variables.
1951        (d) When -B was used with -n, some line numbers in the output were one less
1952            than they should have been.
1954        (e) Added the -o (--only-matching) option.
1956        (f) If -A or -C was used with -c (count only), some lines of context were
1957            accidentally printed for the final match.
1959        (g) Added the -H (--with-filename) option.
1961        (h) The combination of options -rh failed to suppress file names for files
1962            that were found from directory arguments.
1964        (i) Added the -D (--devices) and -d (--directories) options.
1966        (j) Added the -F (--fixed-strings) option.
1968        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1970        (l) Added the --colo(u)r option.
1972        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1973            is not present by default.
1975     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1976        items such as (?R) or (?1), when the recursion could match a number of
1977        alternatives. If it matched one of the alternatives, but subsequently,
1978        outside the recursion, there was a failure, the code tried to back up into
1979        the recursion. However, because of the way PCRE is implemented, this is not
1980        possible, and the result was an incorrect result from the match.
1982        In order to prevent this happening, the specification of recursion has
1983        been changed so that all such subpatterns are automatically treated as
1984        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1986     4. I had overlooked the fact that, in some locales, there are characters for
1987        which isalpha() is true but neither isupper() nor islower() are true. In
1988        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1989        and ordfeminine) are like this. This affected the treatment of \w and \W
1990        when they appeared in character classes, but not when they appeared outside
1991        a character class. The bit map for "word" characters is now created
1992        separately from the results of isalnum() instead of just taking it from the
1993        upper, lower, and digit maps. (Plus the underscore character, of course.)
1995     5. The above bug also affected the handling of POSIX character classes such as
1996        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1997        permanent tables. Instead, the bit maps for such a class were previously
1998        created as the appropriate unions of the upper, lower, and digit bitmaps.
1999        Now they are created by subtraction from the [[:word:]] class, which has
2000        its own bitmap.
2002     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2003        It is created by subtracting the vertical space characters (\x09, \x0a,
2004        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2005        subtraction was done in the overall bitmap for a character class, meaning
2006        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2007        be recognized. This bug has been fixed.
2009     7. Patches from the folks at Google:
2011          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2012          real life, but is still worth protecting against".
2014          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2015          regular expressions".
2017          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2018          have it.
2020          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2021          "configure" and the latter not, in order to fix a problem somebody had
2022          with compiling the Arg class on HP-UX.
2024          (e) Improve the error-handling of the C++ wrapper a little bit.
2026          (f) New tests for checking recursion limiting.
2028     8. The pcre_memmove() function, which is used only if the environment does not
2029        have a standard memmove() function (and is therefore rarely compiled),
2030        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2031        returning a result (though PCRE never actually uses the result).
2033     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2034        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2035        returned instead of calling malloc() with an overflowing number that would
2036        most likely cause subsequent chaos.
2038    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2040    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2041        with this option is matched, the nmatch and pmatch options of regexec() are
2042        ignored.
2044    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2045        provided in case anyone wants to the the POSIX interface with UTF-8
2046        strings.
2048    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2049        C++ linking (needed for some HP-UX environments).
2051    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2052        (unused parameter) and in the pcre_printint() function (omitted "default"
2053        switch label when the default is to do nothing).
2055    15. Added some code to make it possible, when PCRE is compiled as a C++
2056        library, to replace subject pointers for pcre_exec() with a smart pointer
2057        class, thus making it possible to process discontinuous strings.
2059    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2060        much the same function. They were added by different people who were trying
2061        to make PCRE easy to compile on non-Unix systems. It has been suggested
2062        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2063        for compiling on Windows systems. I have therefore replaced it with
2064        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2065        defaults to "extern" for C or "extern C" for C++, which works fine on
2066        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2067        SCOPE with something explicit in config.h. In addition:
2069        (a) pcreposix.h still had just "extern" instead of either of these macros;
2070            I have replaced it with PCRE_DATA_SCOPE.
2072        (b) Functions such as _pcre_xclass(), which are internal to the library,
2073            but external in the C sense, all had PCRE_EXPORT in their definitions.
2074            This is apparently wrong for the Windows case, so I have removed it.
2075            (It makes no difference on Unix-like systems.)
2077    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2078        of recursive calls to match(). This is different to MATCH_LIMIT because
2079        that limits the total number of calls to match(), not all of which increase
2080        the depth of recursion. Limiting the recursion depth limits the amount of
2081        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2082        when PCRE is compiled, and changed at run time. A patch from Google adds
2083        this functionality to the C++ interface.
2085    18. Changes to the handling of Unicode character properties:
2087        (a) Updated the table to Unicode 4.1.0.
2089        (b) Recognize characters that are not in the table as "Cn" (undefined).
2091        (c) I revised the way the table is implemented to a much improved format
2092            which includes recognition of ranges. It now supports the ranges that
2093            are defined in UnicodeData.txt, and it also amalgamates other
2094            characters into ranges. This has reduced the number of entries in the
2095            table from around 16,000 to around 3,000, thus reducing its size
2096            considerably. I realized I did not need to use a tree structure after
2097            all - a binary chop search is just as efficient. Having reduced the
2098            number of entries, I extended their size from 6 bytes to 8 bytes to
2099            allow for more data.
2101        (d) Added support for Unicode script names via properties such as \p{Han}.
2103    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2104        matching that character.
2106    20. When matching a repeated Unicode property with a minimum greater than zero,
2107        (for example \pL{2,}), PCRE could look past the end of the subject if it
2108        reached it while seeking the minimum number of characters. This could
2109        happen only if some of the characters were more than one byte long, because
2110        there is a check for at least the minimum number of bytes.
2112    21. Refactored the implementation of \p and \P so as to be more general, to
2113        allow for more different types of property in future. This has changed the
2114        compiled form incompatibly. Anybody with saved compiled patterns that use
2115        \p or \P will have to recompile them.
2117    22. Added "Any" and "L&" to the supported property types.
2119    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2120        but give a compile time error if the value is greater than 0xff.
2122    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2123        accidentally not being installed or uninstalled.
2125    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2126        made were to insert the current release number. This seemed silly, because
2127        it made things harder for people building PCRE on systems that don't run
2128        "configure". I have turned pcre.h into a distributed file, no longer built
2129        by "configure", with the version identification directly included. There is
2130        no longer a pcre.h.in file.
2132        However, this change necessitated a change to the pcre-config script as
2133        well. It is built from pcre-config.in, and one of the substitutions was the
2134        release number. I have updated configure.ac so that ./configure now finds
2135        the release number by grepping pcre.h.
2137    26. Added the ability to run the tests under valgrind.
2140    Version 6.4 05-Sep-05
2141    ---------------------
2143     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2144        "--" to be printed when multiple files were scanned, even when none of the
2145        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2146        consider it to be a bug, and have restored the previous behaviour.
2148     2. A couple of code tidies to get rid of compiler warnings.
2150     3. The pcretest program used to cheat by referring to symbols in the library
2151        whose names begin with _pcre_. These are internal symbols that are not
2152        really supposed to be visible externally, and in some environments it is
2153        possible to suppress them. The cheating is now confined to including
2154        certain files from the library's source, which is a bit cleaner.
2156     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2157        file's purpose clearer.
2159     5. Reorganized pcre_ucp_findchar().
2162    Version 6.3 15-Aug-05
2163    ---------------------
2165     1. The file libpcre.pc.in did not have general read permission in the tarball.
2167     2. There were some problems when building without C++ support:
2169        (a) If C++ support was not built, "make install" and "make test" still
2170            tried to test it.
2172        (b) There were problems when the value of CXX was explicitly set. Some
2173            changes have been made to try to fix these, and ...
2175        (c) --disable-cpp can now be used to explicitly disable C++ support.
2177        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2178            backslash in a target when C++ was disabled. This confuses some
2179            versions of "make", apparently. Using an intermediate variable solves
2180            this. (Same for CPP_LOBJ.)
2182     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2183        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2184        necessary on certain architectures.
2186     4. Added a setting of -export-symbols-regex to the link command to remove
2187        those symbols that are exported in the C sense, but actually are local
2188        within the library, and not documented. Their names all begin with
2189        "_pcre_". This is not a perfect job, because (a) we have to except some
2190        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2191        available (and never for static libraries). I have made a note to try to
2192        find a way round (a) in the future.
2195    Version 6.2 01-Aug-05
2196    ---------------------
2198     1. There was no test for integer overflow of quantifier values. A construction
2199        such as {1111111111111111} would give undefined results. What is worse, if
2200        a minimum quantifier for a parenthesized subpattern overflowed and became
2201        negative, the calculation of the memory size went wrong. This could have
2202        led to memory overwriting.
2204     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2206     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2207        operating environments where this matters.
2209     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2210        PCRE options from within the C++ wrapper.
2212     5. Named capturing subpatterns were not being correctly counted when a pattern
2213        was compiled. This caused two problems: (a) If there were more than 100
2214        such subpatterns, the calculation of the memory needed for the whole
2215        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2216        back references of the form \12, where the number was greater than 9, were
2217        not recognized as back references, even though there were sufficient
2218        previous subpatterns.
2220     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2221        versions of gcc, e.g. 2.95.4.
2224    Version 6.1 21-Jun-05
2225    ---------------------
2227     1. There was one reference to the variable "posix" in pcretest.c that was not
2228        surrounded by "#if !defined NOPOSIX".
2230     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2231        the cross-check on the old pcre_info() function, for the benefit of the
2232        cut-down version of PCRE that is currently imported into Exim.
2234     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2235        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2236        patterns that start (?i) but I don't think that matters. The use of (?i) is
2237        just an example; this all applies to the other options as well.
2239     4. Since libtool seems to echo the compile commands it is issuing, the output
2240        from "make" can be reduced a bit by putting "@" in front of each libtool
2241        compile command.
2243     5. Patch from the folks at Google for configure.in to be a bit more thorough
2244        in checking for a suitable C++ installation before trying to compile the
2245        C++ stuff. This should fix a reported problem when a compiler was present,
2246        but no suitable headers.
2248     6. The man pages all had just "PCRE" as their title. I have changed them to
2249        be the relevant file name. I have also arranged that these names are
2250        retained in the file doc/pcre.txt, which is a concatenation in text format
2251        of all the man pages except the little individual ones for each function.
2253     7. The NON-UNIX-USE file had not been updated for the different set of source
2254        files that come with release 6. I also added a few comments about the C++
2255        wrapper.
2258    Version 6.0 07-Jun-05
2259    ---------------------
2261     1. Some minor internal re-organization to help with my DFA experiments.
2263     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2264        didn't matter for the library itself when fully configured, but did matter
2265        when compiling without UCP support, or within Exim, where the ucp files are
2266        not imported.
2268     3. Refactoring of the library code to split up the various functions into
2269        different source modules. The addition of the new DFA matching code (see
2270        below) to a single monolithic source would have made it really too
2271        unwieldy, quite apart from causing all the code to be include in a
2272        statically linked application, when only some functions are used. This is
2273        relevant even without the DFA addition now that patterns can be compiled in
2274        one application and matched in another.
2276        The downside of splitting up is that there have to be some external
2277        functions and data tables that are used internally in different modules of
2278        the library but which are not part of the API. These have all had their
2279        names changed to start with "_pcre_" so that they are unlikely to clash
2280        with other external names.
2282     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2283        a different (DFA) algorithm. Although it is slower than the original
2284        function, it does have some advantages for certain types of matching
2285        problem.
2287     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2288        including restarting after a partial match.
2290     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2291        defined when compiling for Windows was sent to me. I have put it into the
2292        code, though I have no means of testing or verifying it.
2294     7. Added the pcre_refcount() auxiliary function.
2296     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2297        match before or at the first newline in the subject string. In pcretest,
2298        the /f option on a pattern can be used to set this.
2300     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2301        would behave wrongly. This has been present in PCRE since release 4.0.
2303    10. A number of changes to the pcregrep command:
2305        (a) Refactored how -x works; insert ^(...)$ instead of setting
2306            PCRE_ANCHORED and checking the length, in preparation for adding
2307            something similar for -w.
2309        (b) Added the -w (match as a word) option.
2311        (c) Refactored the way lines are read and buffered so as to have more
2312            than one at a time available.
2314        (d) Implemented a pcregrep test script.
2316        (e) Added the -M (multiline match) option. This allows patterns to match
2317            over several lines of the subject. The buffering ensures that at least
2318            8K, or the rest of the document (whichever is the shorter) is available
2319            for matching (and similarly the previous 8K for lookbehind assertions).
2321        (f) Changed the --help output so that it now says
2323              -w, --word-regex(p)
2325            instead of two lines, one with "regex" and the other with "regexp"
2326            because that confused at least one person since the short forms are the
2327            same. (This required a bit of code, as the output is generated
2328            automatically from a table. It wasn't just a text change.)
2330        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2331            option but starts with a hyphen. Could be a pattern or a path name
2332            starting with a hyphen, for instance.
2334        (h) "-" can be given as a file name to represent stdin.
2336        (i) When file names are being printed, "(standard input)" is used for
2337            the standard input, for compatibility with GNU grep. Previously
2338            "<stdin>" was used.
2340        (j) The option --label=xxx can be used to supply a name to be used for
2341            stdin when file names are being printed. There is no short form.
2343        (k) Re-factored the options decoding logic because we are going to add
2344            two more options that take data. Such options can now be given in four
2345            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2347        (l) Added the -A, -B, and -C options for requesting that lines of context
2348            around matches be printed.
2350        (m) Added the -L option to print the names of files that do not contain
2351            any matching lines, that is, the complement of -l.
2353        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2354            continue to scan other files.
2356        (o) The -s option was incorrectly implemented. For compatibility with other
2357            greps, it now suppresses the error message for a non-existent or non-
2358            accessible file (but not the return code). There is a new option called
2359            -q that suppresses the output of matching lines, which was what -s was
2360            previously doing.
2362        (p) Added --include and --exclude options to specify files for inclusion
2363            and exclusion when recursing.
2365    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2366        Hopefully, it now does.
2368    12. Missing cast in pcre_study().
2370    13. Added an "uninstall" target to the makefile.
2372    14. Replaced "extern" in the function prototypes in Makefile.in with
2373        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2374        world, but is set differently for Windows.
2376    15. Added a second compiling function called pcre_compile2(). The only
2377        difference is that it has an extra argument, which is a pointer to an
2378        integer error code. When there is a compile-time failure, this is set
2379        non-zero, in addition to the error test pointer being set to point to an
2380        error message. The new argument may be NULL if no error number is required
2381        (but then you may as well call pcre_compile(), which is now just a
2382        wrapper). This facility is provided because some applications need a
2383        numeric error indication, but it has also enabled me to tidy up the way
2384        compile-time errors are handled in the POSIX wrapper.
2386    16. Added VPATH=.libs to the makefile; this should help when building with one
2387        prefix path and installing with another. (Or so I'm told by someone who
2388        knows more about this stuff than I do.)
2390    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2391        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2392        match everything, including newlines. This is not POSIX-compatible, but
2393        somebody wanted the feature. From pcretest it can be activated by using
2394        both the P and the s flags.
2396    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2398    19. libpcre.pc was being incorrectly installed as executable.
2400    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2401        it now also looks for '\r' so that it will work unmodified on Windows.
2403    21. Added Google's contributed C++ wrapper to the distribution.
2405    22. Added some untidy missing memory free() calls in pcretest, to keep
2406        Electric Fence happy when testing.
2410    Version 5.0 13-Sep-04
2411    ---------------------
2413     1. Internal change: literal characters are no longer packed up into items
2414        containing multiple characters in a single byte-string. Each character
2415        is now matched using a separate opcode. However, there may be more than one
2416        byte in the character in UTF-8 mode.
2418     2. The pcre_callout_block structure has two new fields: pattern_position and
2419        next_item_length. These contain the offset in the pattern to the next match
2420        item, and its length, respectively.
2422     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2423        insertion of callouts before each pattern item. Added the /C option to
2424        pcretest to make use of this.
2426     4. On the advice of a Windows user, the lines
2428          #if defined(_WIN32) || defined(WIN32)
2429          _setmode( _fileno( stdout ), 0x8000 );
2430          #endif  /* defined(_WIN32) || defined(WIN32) */
2432        have been added to the source of pcretest. This apparently does useful
2433        magic in relation to line terminators.
2435     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2436        for the benefit of those environments where the "b" makes a difference.
2438     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2439        to know about it. I have put a hack into configure.in that adds in code
2440        to set GCC=yes if CC=icc. This seems to end up at a point in the
2441        generated configure script that is early enough to affect the setting of
2442        compiler options, which is what is needed, but I have no means of testing
2443        whether it really works. (The user who reported this had patched the
2444        generated configure script, which of course I cannot do.)
2446        LATER: After change 22 below (new libtool files), the configure script
2447        seems to know about icc (and also ecc). Therefore, I have commented out
2448        this hack in configure.in.
2450     7. Added support for pkg-config (2 patches were sent in).
2452     8. Negated POSIX character classes that used a combination of internal tables
2453        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2454        [[:^ascii]]. Typically, they would match almost any characters. The other
2455        POSIX classes were not broken in this way.
2457     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2458        to find the match, as PCRE was deluded into thinking that the match had to
2459        start at the start point or following a newline. The same bug applied to
2460        patterns with negative forward assertions or any backward assertions
2461        preceding ".*" at the start, unless the pattern required a fixed first
2462        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2464    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2465        starting at the last subject character, bytes beyond the end of the subject
2466        string were read.
2468    11. Renamed the variable "class" as "classbits" to make life easier for C++
2469        users. (Previously there was a macro definition, but it apparently wasn't
2470        enough.)
2472    12. Added the new field "tables" to the extra data so that tables can be passed
2473        in at exec time, or the internal tables can be re-selected. This allows
2474        a compiled regex to be saved and re-used at a later time by a different
2475        program that might have everything at different addresses.
2477    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2478        -R library as well as a -L library.
2480    14. The debugging options of pcretest (-d on the command line or D on a
2481        pattern) showed incorrect output for anything following an extended class
2482        that contained multibyte characters and which was followed by a quantifier.
2484    15. Added optional support for general category Unicode character properties
2485        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2486        support. It adds about 90K to the size of the library. The meanings of the
2487        inbuilt class escapes such as \d and \s have NOT been changed.
2489    16. Updated pcredemo.c to include calls to free() to release the memory for the
2490        compiled pattern.
2492    17. The generated file chartables.c was being created in the source directory
2493        instead of in the building directory. This caused the build to fail if the
2494        source directory was different from the building directory, and was
2495        read-only.
2497    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2498        file. No doubt somebody will tell me if they don't make sense... Also added
2499        Dan Mooney's comments about building on OpenVMS.
2501    19. Added support for partial matching via the PCRE_PARTIAL option for
2502        pcre_exec() and the \P data escape in pcretest.
2504    20. Extended pcretest with 3 new pattern features:
2506        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2507              write the compiled pattern to the file whose name is "rest-of-line".
2508              This is a straight binary dump of the data, with the saved pointer to
2509              the character tables forced to be NULL. The study data, if any, is
2510              written too. After writing, pcretest reads a new pattern.
2512        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2513              compiled pattern from the given file. There must not be any
2514              occurrences of "<" in the file name (pretty unlikely); if there are,
2515              pcretest will instead treat the initial "<" as a pattern delimiter.
2516              After reading in the pattern, pcretest goes on to read data lines as
2517              usual.
2519        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2520              and 16-bit fields in a compiled pattern, to simulate a pattern that
2521              was compiled on a host of opposite endianness.
2523    21. The pcre-exec() function can now cope with patterns that were compiled on
2524        hosts of opposite endianness, with this restriction:
2526          As for any compiled expression that is saved and used later, the tables
2527          pointer field cannot be preserved; the extra_data field in the arguments
2528          to pcre_exec() should be used to pass in a tables address if a value
2529          other than the default internal tables were used at compile time.
2531    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2532        now diagnosed as an error. Previously, most of the time, a negative number
2533        would have been treated as zero, but if in addition "ovector" was passed as
2534        NULL, a crash could occur.
2536    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2537        new versions from the libtool 1.5 distribution (the last one is a copy of
2538        a file called libtool.m4). This seems to have fixed the need to patch
2539        "configure" to support Darwin 1.3 (which I used to do). However, I still
2540        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2541        workstation).
2543    24. Changed the PCRE licence to be the more standard "BSD" licence.
2546    Version 4.5 01-Dec-03
2547    ---------------------
2549     1. There has been some re-arrangement of the code for the match() function so
2550        that it can be compiled in a version that does not call itself recursively.
2551        Instead, it keeps those local variables that need separate instances for
2552        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2553        needs to "recurse". Keeping track of where control must go is done by means
2554        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2555        hide most of the details from the main code, and operates only if
2556        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2557        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2558        operating.
2560        To make it easier for callers to provide specially tailored get/free
2561        functions for this usage, two new functions, pcre_stack_malloc, and
2562        pcre_stack_free, are used. They are always called in strict stacking order,
2563        and the size of block requested is always the same.
2565        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2566        PCRE has been compiled to use the stack or the heap for recursion. The
2567        -C option of pcretest uses this to show which version is compiled.
2569        A new data escape \S, is added to pcretest; it causes the amounts of store
2570        obtained and freed by both kinds of malloc/free at match time to be added
2571        to the output.
2573     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2574        what's available on my current Linux desktop machine.
2576     3. When matching a UTF-8 string, the test for a valid string at the start has
2577        been extended. If start_offset is not zero, PCRE now checks that it points
2578        to a byte that is the start of a UTF-8 character. If not, it returns
2579        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2580        this is necessary because there may be backward assertions in the pattern.
2581        When matching the same subject several times, it may save resources to use
2582        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2584     4. The code for checking the validity of UTF-8 strings has been tightened so
2585        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2586        containing "overlong sequences".
2588     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2589        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2590        should have been "&", but it just so happened that all the cases this let
2591        through by mistake were picked up later in the function.
2593     6. I had used a variable called "isblank" - this is a C99 function, causing
2594        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2596     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2597        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2598        so the version is shown; (c) stop "make test" echoing "./RunTest".
2600     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2602     9. The prototype for memmove() for systems that don't have it was using
2603        size_t, but the inclusion of the header that defines size_t was later. I've
2604        moved the #includes for the C headers earlier to avoid this.
2606    10. Added some adjustments to the code to make it easier to compiler on certain
2607        special systems:
2609          (a) Some "const" qualifiers were missing.
2610          (b) Added the macro EXPORT before all exported functions; by default this
2611              is defined to be empty.
2612          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2613              that it reads its output file name as an argument instead of writing
2614              to the standard output and assuming this can be redirected.
2616    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2617        class containing characters with values greater than 255, PCRE compilation
2618        went into a loop.
2620    12. A recursive reference to a subpattern that was within another subpattern
2621        that had a minimum quantifier of zero caused PCRE to crash. For example,
2622        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2623        recursion. If the recursively-called subpattern itself had a zero repeat,
2624        that was OK.
2626    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2627        buffer into which it was copied (for escape processing) was still set at
2628        1024, so long lines caused crashes.
2630    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2631        "internal error: code overflow...". This applied to any character class
2632        that was followed by a possessive quantifier.
2634    15. Modified the Makefile to add libpcre.la as a prerequisite for
2635        libpcreposix.la because I was told this is needed for a parallel build to
2636        work.
2638    16. If a pattern that contained .* following optional items at the start was
2639        studied, the wrong optimizing data was generated, leading to matching
2640        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2641        matching string must start with a or b or c. The correct conclusion for
2642        this pattern is that a match can start with any character.
2645    Version 4.4 13-Aug-03
2646    ---------------------
2648     1. In UTF-8 mode, a character class containing characters with values between
2649        127 and 255 was not handled correctly if the compiled pattern was studied.
2650        In fixing this, I have also improved the studying algorithm for such
2651        classes (slightly).
2653     2. Three internal functions had redundant arguments passed to them. Removal
2654        might give a very teeny performance improvement.
2656     3. Documentation bug: the value of the capture_top field in a callout is *one
2657        more than* the number of the hightest numbered captured substring.
2659     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2660        in incorrectly linking with a previously installed version. They now link
2661        explicitly with libpcre.la.
2663     5. configure.in no longer needs to recognize Cygwin specially.
2665     6. A problem in pcre.in for Windows platforms is fixed.
2667     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2668        pcretest, it used to include the size of the study block as part of its
2669        output. Unfortunately, the structure contains a field that has a different
2670        size on different hardware architectures. This meant that the tests that
2671        showed this size failed. As the block is currently always of a fixed size,
2672        this information isn't actually particularly useful in pcretest output, so
2673        I have just removed it.
2675     8. Three pre-processor statements accidentally did not start in column 1.
2676        Sadly, there are *still* compilers around that complain, even though
2677        standard C has not required this for well over a decade. Sigh.
2679     9. In pcretest, the code for checking callouts passed small integers in the
2680        callout_data field, which is a void * field. However, some picky compilers
2681        complained about the casts involved for this on 64-bit systems. Now
2682        pcretest passes the address of the small integer instead, which should get
2683        rid of the warnings.
2685    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2686        both compile and run time, and gives an error if an invalid UTF-8 sequence
2687        is found. There is a option for disabling this check in cases where the
2688        string is known to be correct and/or the maximum performance is wanted.
2690    11. In response to a bug report, I changed one line in Makefile.in from
2692            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2693        to
2694            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2696        to look similar to other lines, but I have no way of telling whether this
2697        is the right thing to do, as I do not use Windows. No doubt I'll get told
2698        if it's wrong...
2701    Version 4.3 21-May-03
2702    ---------------------
2704    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2705       Makefile.
2707    2. Some refactoring to improve the quality of the code:
2709       (i)   The utf8_table... variables are now declared "const".
2711       (ii)  The code for \cx, which used the "case flipping" table to upper case
2712             lower case letters, now just substracts 32. This is ASCII-specific,
2713             but the whole concept of \cx is ASCII-specific, so it seems
2714             reasonable.
2716       (iii) PCRE was using its character types table to recognize decimal and
2717             hexadecimal digits in the pattern. This is silly, because it handles
2718             only 0-9, a-f, and A-F, but the character types table is locale-
2719             specific, which means strange things might happen. A private
2720             table is now used for this - though it costs 256 bytes, a table is
2721             much faster than multiple explicit tests. Of course, the standard
2722             character types table is still used for matching digits in subject
2723             strings against \d.
2725       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2726             ending in _t are). So I've renamed it as ESC_tee.
2728    3. The first argument for regexec() in the POSIX wrapper should have been
2729       defined as "const".
2731    4. Changed pcretest to use malloc() for its buffers so that they can be
2732       Electric Fenced for debugging.
2734    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2735       to read one or more bytes before the start of the subject string. Often this
2736       had no effect on PCRE's behaviour, but in some circumstances it could
2737       provoke a segmentation fault.
2739    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2740       to try to read one or more bytes before the start of the subject string.
2742    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2743       UTF-8 support could misbehave in various ways if the subject string
2744       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2745       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2746       back over UTF-8 characters.)
2749    Version 4.2 14-Apr-03
2750    ---------------------
2752    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2754    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2755         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2756         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2757         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2758         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2759           and BUILD_EXEEXT
2760         Note: automatic setting of the BUILD variables is not yet working
2761         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2762           compile-time but not at link-time
2763         [LINK]: use for linking executables only
2764         make different versions for Windows and non-Windows
2765         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2766           libraries
2767         [LINK_FOR_BUILD]: new variable
2768         [OBJEXT]: use throughout
2769         [EXEEXT]: use throughout
2770         <winshared>: new target
2771         <wininstall>: new target
2772         <dftables.o>: use native compiler
2773         <dftables>: use native linker
2774         <install>: handle Windows platform correctly
2775         <clean>: ditto
2776         <check>: ditto
2777         copy DLL to top builddir before testing
2779       As part of these changes, -no-undefined was removed again. This was reported
2780       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2781       in any case.
2783    3. Some tidies to get rid of compiler warnings:
2785       . In the match_data structure, match_limit was an unsigned long int, whereas
2786         match_call_count was an int. I've made them both unsigned long ints.
2788       . In pcretest the fact that a const uschar * doesn't automatically cast to
2789         a void * provoked a warning.
2791       . Turning on some more compiler warnings threw up some "shadow" variables
2792         and a few more missing casts.
2794    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2795       option, a class that contained a single character with a value between 128
2796       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2798    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2799       option, a class that contained several characters, but with at least one
2800       whose value was between 128 and 255 caused PCRE to crash.
2803    Version 4.1 12-Mar-03
2804    ---------------------
2806    1. Compiling with gcc -pedantic found a couple of places where casts were
2807    needed, and a string in dftables.c that was longer than standard compilers are
2808    required to support.
2810    2. Compiling with Sun's compiler found a few more places where the code could
2811    be tidied up in order to avoid warnings.
2813    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2814    first of these names is deprecated in the latest Autoconf in favour of the name
2815    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2816    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2817    analogy I have changed it to CFLAGS_FOR_BUILD.
2819    4. Added -no-undefined to the linking command in the Makefile, because this is
2820    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2821    linking step for the pcreposix library.
2823    5. PCRE was failing to diagnose the case of two named groups with the same
2824    name.
2826    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2827    literal character that is needed in the subject for a match, and scans along to
2828    ensure that it is present before embarking on the full matching process. This
2829    saves time in cases of nested unlimited repeats that are never going to match.
2830    Problem: the scan can take a lot of time if the subject is very long (e.g.
2831    megabytes), thus penalizing straightforward matches. It is now done only if the
2832    amount of subject to be scanned is less than 1000 bytes.
2834    7. A lesser problem with the same optimization is that it was recording the
2835    first character of an anchored pattern as "needed", thus provoking a search
2836    right along the subject, even when the first match of the pattern was going to
2837    fail. The "needed" character is now not set for anchored patterns, unless it
2838    follows something in the pattern that is of non-fixed length. Thus, it still
2839    fulfils its original purpose of finding quick non-matches in cases of nested
2840    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2843    Version 4.0 17-Feb-03
2844    ---------------------
2846    1. If a comment in an extended regex that started immediately after a meta-item
2847    extended to the end of string, PCRE compiled incorrect data. This could lead to
2848    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2850    2. Moved to autoconf 2.53 and libtool 1.4.2.
2852    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2853    the special perltest8 script is no longer needed - all the tests can be run
2854    from a single perltest script.
2856    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2857    by \s. It has now been removed in PCRE. This means it isn't recognized as
2858    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2859    class [:space:] *does* include VT, thereby creating a mess.
2861    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2862    space and tab.
2864    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2865    its new features into the main test script, reducing the number of scripts.
2867    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2868    were backward compatible, and made the (?i) apply to the whole pattern, as if
2869    /i were given. Now it behaves more logically, and applies the option setting
2870    only to what follows. PCRE has been changed to follow suit. However, if it
2871    finds options settings right at the start of the pattern, it extracts them into
2872    the global options, as before. Thus, they show up in the info data.
2874    8. Added support for the \Q...\E escape sequence. Characters in between are
2875    treated as literals. This is slightly different from Perl in that $ and @ are
2876    also handled as literals inside the quotes. In Perl, they will cause variable
2877    interpolation. Note the following examples:
2879        Pattern            PCRE matches      Perl matches
2881        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2882        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2883        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2885    For compatibility with Perl, \Q...\E sequences are recognized inside character
2886    classes as well as outside them.
2888    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2889    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2890    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2891    signed/unsigned warnings.
2893    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2894    option for pcretest, so I've replaced it by a simple function that does just
2895    that job.
2897    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2898    "pcregrep -".
2900    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2901    Java package. This provides some syntactic sugar for simple cases of what my
2902    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2903    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2904    item, you can use this simplified notation. Note that only makes sense with
2905    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2906    greediness, whatever the setting of the PCRE_UNGREEDY option.
2908    13. A change of greediness default within a pattern was not taking effect at
2909    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2910    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2911    was abstracted outside.
2913    14. PCRE now supports the \G assertion. It is true when the current matching
2914    position is at the start point of the match. This differs from \A when the
2915    starting offset is non-zero. Used with the /g option of pcretest (or similar
2916    code), it works in the same way as it does for Perl's /g option. If all
2917    alternatives of a regex begin with \G, the expression is anchored to the start
2918    match position, and the "anchored" flag is set in the compiled expression.
2920    15. Some bugs concerning the handling of certain option changes within patterns
2921    have been fixed. These applied to options other than (?ims). For example,
2922    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2923    been the other way round. Some of this was related to change 7 above.
2925    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2926    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2927    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2928    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2930    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2931    mode. Unlike ".", it always matches newline, whatever the setting of
2932    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2933    assertions. Perl allows it, but it doesn't (in general) work because it can't
2934    calculate the length of the lookbehind. At least, that's the case for Perl
2935    5.8.0 - I've been told they are going to document that it doesn't work in
2936    future.
2938    18. Added an error diagnosis for escapes that PCRE does not support: these are
2939    \L, \l, \N, \P, \p, \U, \u, and \X.
2941    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2942    reading past the end of the pattern in cases such as /[abcd/.
2944    20. PCRE was getting more memory than necessary for patterns with classes that
2945    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2947    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2948    compiling PCRE for use with Virtual Pascal.
2950    22. Small fix to the Makefile to make it work properly if the build is done
2951    outside the source tree.
2953    23. Added a new extension: a condition to go with recursion. If a conditional
2954    subpattern starts with (?(R) the "true" branch is used if recursion has
2955    happened, whereas the "false" branch is used only at the top level.
2957    24. When there was a very long string of literal characters (over 255 bytes
2958    without UTF support, over 250 bytes with UTF support), the computation of how
2959    much memory was required could be incorrect, leading to segfaults or other
2960    strange effects.
2962    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2963    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2964    there was a subsequent back reference to those brackets. This meant that, for
2965    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2966    possible to check for precisely this case. All we can do is abandon the
2967    optimization if .* occurs inside capturing brackets when there are any back
2968    references whatsoever. (See below for a better fix that came later.)
2970    26. The handling of the optimization for finding the first character of a
2971    non-anchored pattern, and for finding a character that is required later in the
2972    match were failing in some cases. This didn't break the matching; it just
2973    failed to optimize when it could. The way this is done has been re-implemented.
2975    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2977    28. Added a new feature that provides some of the functionality that Perl
2978    provides with (?{...}). The facility is termed a "callout". The way it is done
2979    in PCRE is for the caller to provide an optional function, by setting
2980    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2981    global variable. By default it is unset, which disables all calling out. To get
2982    the function called, the regex must include (?C) at appropriate points. This
2983    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2984    This provides a means of identifying different callout points. When PCRE
2985    reaches such a point in the regex, if pcre_callout has been set, the external
2986    function is called. It is provided with data in a structure called
2987    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2988    matching continues; if it returns a non-zero value, the match at the current
2989    point fails. However, backtracking will occur if possible. [This was changed
2990    later and other features added - see item 49 below.]
2992    29. pcretest is upgraded to test the callout functionality. It provides a
2993    callout function that displays information. By default, it shows the start of
2994    the match and the current position in the text. There are some new data escapes
2995    to vary what happens:
2997        \C+         in addition, show current contents of captured substrings
2998        \C-         do not supply a callout function
2999        \C!n        return 1 when callout number n is reached
3000        \C!n!m      return 1 when callout number n is reached for the mth time
3002    30. If pcregrep was called with the -l option and just a single file name, it
3003    output "<stdin>" if a match was found, instead of the file name.
3005    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3006    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3007    pcre_exec(). This saves a malloc/free per call. The default value of
3008    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3009    when configuring.
3011    32. The default maximum size of a compiled pattern is 64K. There have been a
3012    few cases of people hitting this limit. The code now uses macros to handle the
3013    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3014    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3015    configuring. Tests 2 and 5 work only with 2-byte links because they output
3016    debugging information about compiled patterns.
3018    33. Internal code re-arrangements:
3020    (a) Moved the debugging function for printing out a compiled regex into
3021        its own source file (printint.c) and used #include to pull it into
3022        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3023        separate copies.
3025    (b) Defined the list of op-code names for debugging as a macro in
3026        internal.h so that it is next to the definition of the opcodes.
3028    (c) Defined a table of op-code lengths for simpler skipping along compiled
3029        code. This is again a macro in internal.h so that it is next to the
3030        definition of the opcodes.
3032    34. Added support for recursive calls to individual subpatterns, along the
3033    lines of Robin Houston's patch (but implemented somewhat differently).
3035    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3036    allow it to read and process whole directories in Win32. This code was
3037    contributed by Lionel Fourquaux; it has not been tested by me.
3039    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3040    used to name a group. Names consist of alphanumerics and underscores, and must
3041    be unique. Back references use the syntax (?P=name) and recursive calls use
3042    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3043    numbers. The function pcre_fullinfo() can be used after compilation to extract
3044    a name/number map. There are three relevant calls:
3046      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3047      PCRE_INFO_NAMECOUNT            yields the number of entries
3048      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3050    The map is a vector of fixed-size entries. The size of each entry depends on
3051    the length of the longest name used. The first two bytes of each entry are the
3052    group number, most significant byte first. There follows the corresponding
3053    name, zero terminated. The names are in alphabetical order.
3055    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3056    case instead of 255. Making it the same both with and without UTF-8 support
3057    means that the same test output works with both.
3059    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3060    calling malloc() with a zero argument.
3062    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3063    optimization. I've improved things by keeping a bitmap of backreferences with
3064    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3065    fact referenced, the optimization can be applied. It is unlikely that a
3066    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3067    the match to follow \n) will appear inside brackets with a number greater than
3068    31, but if it does, any back reference > 31 suppresses the optimization.
3070    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3071    of disabling numbered capturing parentheses. Any opening parenthesis that is
3072    not followed by ? behaves as if it were followed by ?: but named parentheses
3073    can still be used for capturing (and they will acquire numbers in the usual
3074    way).
3076    41. Redesigned the return codes from the match() function into yes/no/error so
3077    that errors can be passed back from deep inside the nested calls. A malloc
3078    failure while inside a recursive subpattern call now causes the
3079    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3081    42. It is now possible to set a limit on the number of times the match()
3082    function is called in a call to pcre_exec(). This facility makes it possible to
3083    limit the amount of recursion and backtracking, though not in a directly
3084    obvious way, because the match() function is used in a number of different
3085    circumstances. The count starts from zero for each position in the subject
3086    string (for non-anchored patterns). The default limit is, for compatibility, a
3087    large number, namely 10 000 000. You can change this in two ways:
3089    (a) When configuring PCRE before making, you can use --with-match-limit=n
3090        to set a default value for the compiled library.
3092    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3093        a different value is set. See 45 below.
3095    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3097    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3098    of things that can be changed at compile time. The first argument specifies
3099    what is wanted and the second points to where the information is to be placed.
3100    The current list of available information is:
3102      PCRE_CONFIG_UTF8
3104    The output is an integer that is set to one if UTF-8 support is available;
3105    otherwise it is set to zero.
3109    The output is an integer that it set to the value of the code that is used for
3110    newline. It is either LF (10) or CR (13).
3114    The output is an integer that contains the number of bytes used for internal
3115    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3119    The output is an integer that contains the threshold above which the POSIX
3120    interface uses malloc() for output vectors. See item 31 above.
3124    The output is an unsigned integer that contains the default limit of the number
3125    of match() calls in a pcre_exec() execution. See 42 above.
3127    44. pcretest has been upgraded by the addition of the -C option. This causes it
3128    to extract all the available output from the new pcre_config() function, and to
3129    output it. The program then exits immediately.
3131    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3132    order to support additional features. One way would have been to define
3133    pcre_exec2() (for example) with extra arguments, but this would not have been
3134    extensible, and would also have required all calls to the original function to
3135    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3136    is used for passing in "extra" data from pcre_study().
3138    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3139    contains the following fields:
3141      flags         a bitmap indicating which of the following fields are set
3142      study_data    opaque data from pcre_study()
3143      match_limit   a way of specifying a limit on match() calls for a specific
3144                      call to pcre_exec()
3145      callout_data  data for callouts (see 49 below)
3147    The flag bits are also defined in pcre.h, and are
3153    The pcre_study() function now returns one of these new pcre_extra blocks, with
3154    the actual study data pointed to by the study_data field, and the
3155    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3156    before. That is, this change is entirely upwards-compatible and requires no
3157    change to existing code.
3159    If you want to pass in additional data to pcre_exec(), you can either place it
3160    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3161    block.
3163    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3164    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3165    times with different match limits, until it finds the minimum value needed for
3166    pcre_exec() to complete. The value is then output. This can be instructive; for
3167    most simple matches the number is quite small, but for pathological cases it
3168    gets very large very quickly.
3170    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3171    returns the size of the data block pointed to by the study_data field in a
3172    pcre_extra block, that is, the value that was passed as the argument to
3173    pcre_malloc() when PCRE was getting memory in which to place the information
3174    created by pcre_study(). The fourth argument should point to a size_t variable.
3175    pcretest has been extended so that this information is shown after a successful
3176    pcre_study() call when information about the compiled regex is being displayed.
3178    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3179    because what follows is always an absolute path. (Later: it turns out that this
3180    is more than cosmetic for MinGW, because it doesn't like empty path
3181    components.)
3183    49. Some changes have been made to the callout feature (see 28 above):
3185    (i)  A callout function now has three choices for what it returns:
3187           0  =>  success, carry on matching
3188         > 0  =>  failure at this point, but backtrack if possible
3189         < 0  =>  serious error, return this value from pcre_exec()
3191         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3192         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3193         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3194         use by callout functions. It will never be used by PCRE itself.
3196    (ii) The pcre_extra structure (see 45 above) has a void * field called
3197         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3198         pcre_callout_block structure has a field of the same name. The contents of
3199         the field passed in the pcre_extra structure are passed to the callout
3200         function in the corresponding field in the callout block. This makes it
3201         easier to use the same callout-containing regex from multiple threads. For
3202         testing, the pcretest program has a new data escape
3204           \C*n        pass the number n (may be negative) as callout_data
3206         If the callout function in pcretest receives a non-zero value as
3207         callout_data, it returns that value.
3209    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3210    there were some redundant $(CFLAGS) in commands that are now specified as
3211    $(LINK), which already includes $(CFLAGS).
3213    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3214    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3215    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3216    one-byte characters throughout. Note that case-insensitive matching applies
3217    only to characters whose values are less than 256. PCRE doesn't support the
3218    notion of cases for higher-valued characters.
3220    (i)   A character class whose characters are all within 0-255 is handled as
3221          a bit map, and the map is inverted for negative classes. Previously, a
3222          character > 255 always failed to match such a class; however it should
3223          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3225    (ii)  A negated character class with a single character < 255 is coded as
3226          "not this character" (OP_NOT). This wasn't working properly when the test
3227          character was multibyte, either singly or repeated.
3229    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3230          mode, for example: \x{100}{2,3}.
3232    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3233          singly or repeated) now correctly test multibyte characters. However,
3234          PCRE doesn't recognize any characters with values greater than 255 as
3235          digits, spaces, or word characters. Such characters always match \D, \S,
3236          and \W, and never match \d, \s, or \w.
3238    (v)   Classes may now contain characters and character ranges with values
3239          greater than 255. For example: [ab\x{100}-\x{400}].
3241    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3242          PCRE in UTF-8 mode.
3244    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3245    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3246    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3247    value.)
3249    53. The single man page has become too large. I have therefore split it up into
3250    a number of separate man pages. These also give rise to individual HTML pages;
3251    these are now put in a separate directory, and there is an index.html page that
3252    lists them all. Some hyperlinking between the pages has been installed.
3254    54. Added convenience functions for handling named capturing parentheses.
3256    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3257    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3258    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3259    are faulted.
3261    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3262    calling configure. These values are used when compiling the dftables.c program
3263    which is run to generate the source of the default character tables. They
3264    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3265    you will need to set these values.
3267    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3270    Version 3.9 02-Jan-02
3271    ---------------------
3273    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3275    2. If --disable-static was given, the building process failed when trying to
3276    build pcretest and pcregrep. (For some reason it was using libtool to compile
3277    them, which is not right, as they aren't part of the library.)
3280    Version 3.8 18-Dec-01
3281    ---------------------
3283    1. The experimental UTF-8 code was completely screwed up. It was packing the
3284    bytes in the wrong order. How dumb can you get?
3287    Version 3.7 29-Oct-01
3288    ---------------------
3290    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3291    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3292    this didn't happen under Solaris 8, where I normally test things.
3294    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3295    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3296    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3297    this fix an hour or so after the initial 3.7 release.)
3300    Version 3.6 23-Oct-01
3301    ---------------------
3303    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3304    offsets passed as NULL with zero offset count.
3306    2. The config.guess and config.sub files had not been updated when I moved to
3307    the latest autoconf.
3310    Version 3.5 15-Aug-01
3311    ---------------------
3313    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3314    had been forgotten.
3316    2. By using declared but undefined structures, we can avoid using "void"
3317    definitions in pcre.h while keeping the internal definitions of the structures
3318    private.
3320    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3321    user point of view, this means that both static and shared libraries are built
3322    by default, but this can be individually controlled. More of the work of
3323    handling this static/shared cases is now inside libtool instead of PCRE's make
3324    file.
3326    4. The pcretest utility is now installed along with pcregrep because it is
3327    useful for users (to test regexs) and by doing this, it automatically gets
3328    relinked by libtool. The documentation has been turned into a man page, so
3329    there are now .1, .txt, and .html versions in /doc.
3331    5. Upgrades to pcregrep:
3332       (i)   Added long-form option names like gnu grep.
3333       (ii)  Added --help to list all options with an explanatory phrase.
3334       (iii) Added -r, --recursive to recurse into sub-directories.
3335       (iv)  Added -f, --file to read patterns from a file.
3337    6. pcre_exec() was referring to its "code" argument before testing that
3338    argument for NULL (and giving an error if it was NULL).
3340    7. Upgraded Makefile.in to allow for compiling in a different directory from
3341    the source directory.
3343    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3344    options bits, the pointer it was passed was to an int instead of to an unsigned
3345    long int. This mattered only on 64-bit systems.
3347    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3348    generated) instead of pcre.in, which it its source. Also made the same change
3349    in several of the .c files.
3351    10. A new release of gcc defines printf() as a macro, which broke pcretest
3352    because it had an ifdef in the middle of a string argument for printf(). Fixed
3353    by using separate calls to printf().
3355    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3356    script, to force use of CR or LF instead of \n in the source. On non-Unix
3357    systems, the value can be set in config.h.
3359    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3360    absolute limit. Changed the text of the error message to make this clear, and
3361    likewise updated the man page.
3363    13. The limit of 99 on the number of capturing subpatterns has been removed.
3364    The new limit is 65535, which I hope will not be a "real" limit.
3367    Version 3.4 22-Aug-00
3368    ---------------------
3370    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3372    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3375    Version 3.3 01-Aug-00
3376    ---------------------
3378    1. If an octal character was given, but the value was greater than \377, it
3379    was not getting masked to the least significant bits, as documented. This could
3380    lead to crashes in some systems.
3382    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3383    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3385    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3386    These just pass their arguments on to (pcre_free)(), but they are provided
3387    because some uses of PCRE bind it to non-C systems that can call its functions,
3388    but cannot call free() or pcre_free() directly.
3390    4. Add "make test" as a synonym for "make check". Corrected some comments in
3391    the Makefile.
3393    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3394    Makefile.
3396    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3397    command called pgrep for grepping around the active processes.
3399    7. Added the beginnings of support for UTF-8 character strings.
3401    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3402    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3403    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3404    out for the ar command.)
3407    Version 3.2 12-May-00
3408    ---------------------
3410    This is purely a bug fixing release.
3412    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3413    of ZA. This was just one example of several cases that could provoke this bug,
3414    which was introduced by change 9 of version 2.00. The code for breaking
3415    infinite loops after an iteration that matches an empty string was't working
3416    correctly.
3418    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3419    when matched against abbab (for example). After matching an empty string, it
3420    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3421    caused it to match further down the string than it should.
3423    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3424    was there because it doesn't seem to be needed, and it causes trouble on some
3425    systems, as it is not a Standard C header. It has been removed.
3427    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3428    were reported on the Macintosh. The changes were from
3430      while ((c = *(++ptr)) != 0 && c != '\n');
3431    to
3432      while ((c = *(++ptr)) != 0 && c != '\n') ;
3434    Totally extraordinary, but if that's what it takes...
3436    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3437    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3438    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3439    assumes the way PCRE uses memmove() (always moving upwards).
3441    6. PCRE is being used in one environment where strchr() is not available. There
3442    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3443    faster code anyway.
3446    Version 3.1 09-Feb-00
3447    ---------------------
3449    The only change in this release is the fixing of some bugs in Makefile.in for
3450    the "install" target:
3452    (1) It was failing to install pcreposix.h.
3454    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3457    Version 3.0 01-Feb-00
3458    ---------------------
3460    1. Add support for the /+ modifier to perltest (to output $` like it does in
3461    pcretest).
3463    2. Add support for the /g modifier to perltest.
3465    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3466    matches null strings.
3468    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3469    pattern. Perl treats empty patterns specially - it reuses the most recent
3470    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3471    effect.
3473    5. The POSIX interface was broken in that it was just handing over the POSIX
3474    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3475    required a bigger vector, with some working space on the end. This means that
3476    the POSIX wrapper now has to get and free some memory, and copy the results.
3478    6. Added some simple autoconf support, placing the test data and the
3479    documentation in separate directories, re-organizing some of the
3480    information files, and making it build pcre-config (a GNU standard). Also added
3481    libtool support for building PCRE as a shared library, which is now the
3482    default.
3484    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3485    09 are not valid octal constants. Single digits will be used for minor values
3486    less than 10.
3488    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3489    existing programs that set these in the POSIX interface can use PCRE without
3490    modification.
3492    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3493    return all that pcre_info() returns, plus additional data. The pcre_info()
3494    function is retained for compatibility, but is considered to be obsolete.
3496    10. Added experimental recursion feature (?R) to handle one common case that
3497    Perl 5.6 will be able to do with (?p{...}).
3499    11. Added support for POSIX character classes like [:alpha:], which Perl is
3500    adopting.
3503    Version 2.08 31-Aug-99
3504    ----------------------
3506    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3507    trying to match at the startoffset position, but instead was moving forward to
3508    the next newline as if a previous match had failed.
3510    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3511    and could get into a loop if a null string was matched other than at the start
3512    of the subject.
3514    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3515    be distinguished at compile time, and for completeness also added PCRE_DATE.
3517    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3518    in GnuWin32 environments.
3521    Version 2.07 29-Jul-99
3522    ----------------------
3524    1. The documentation is now supplied in plain text form and HTML as well as in
3525    the form of man page sources.
3527    2. C++ compilers don't like assigning (void *) values to other pointer types.
3528    In particular this affects malloc(). Although there is no problem in Standard
3529    C, I've put in casts to keep C++ compilers happy.
3531    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3532    should be (const char *).
3534    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3535    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3536    However, I haven't made this a standard facility. The documentation doesn't
3537    mention it, and the Makefile doesn't support it.
3539    5. The Makefile now contains an "install" target, with editable destinations at
3540    the top of the file. The pcretest program is not installed.
3542    6. pgrep -V now gives the PCRE version number and date.
3544    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3545    causing the entire string to be ignored, instead of just the last character.
3547    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3548    non-matching string, it can take a very, very long time, even for strings of
3549    quite modest length, because of the nested recursion. PCRE now does better in
3550    some of these cases. It does this by remembering the last required literal
3551    character in the pattern, and pre-searching the subject to ensure it is present
3552    before running the real match. In other words, it applies a heuristic to detect
3553    some types of certain failure quickly, and in the above example, if presented
3554    with a string that has no trailing " it gives "no match" very quickly.
3556    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3557    other alternatives are tried instead.
3560    Version 2.06 09-Jun-99
3561    ----------------------
3563    1. Change pcretest's output for amount of store used to show just the code
3564    space, because the remainder (the data block) varies in size between 32-bit and
3565    64-bit systems.
3567    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3568    start matching at. This allows lookbehinds to work when searching for multiple
3569    occurrences in a string.
3571    3. Added additional options to pcretest for testing multiple occurrences:
3573       /+   outputs the rest of the string that follows a match
3574       /g   loops for multiple occurrences, using the new startoffset argument
3575       /G   loops for multiple occurrences by passing an incremented pointer
3577    4. PCRE wasn't doing the "first character" optimization for patterns starting
3578    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3579    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3580    the letter 'x'. On long subject strings, this gives a significant speed-up.
3583    Version 2.05 21-Apr-99
3584    ----------------------
3586    1. Changed the type of magic_number from int to long int so that it works
3587    properly on 16-bit systems.
3589    2. Fixed a bug which caused patterns starting with .* not to work correctly
3590    when the subject string contained newline characters. PCRE was assuming
3591    anchoring for such patterns in all cases, which is not correct because .* will
3592    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3593    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3594    must be retried after every newline in the subject.
3597    Version 2.04 18-Feb-99
3598    ----------------------
3600    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3601    computation of the store needed to hold the pattern was incorrect (too large).
3602    If such patterns were nested a few deep, this could multiply and become a real
3603    problem.
3605    2. Added /M option to pcretest to show the memory requirement of a specific
3606    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3608    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3609    compiled in such a way that the backtracking after subsequent failure was