ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 21 by nigel, Sat Feb 24 21:38:37 2007 UTC revision 519 by ph10, Fri May 21 16:43:17 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.10 03 May-2010
5    ------------------------
7  Version 1.09 28 Apr-98  1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8        THEN.
10    2.  (*ACCEPT) was not working when inside an atomic group.
12    3.  Inside a character class, \B is treated as a literal by default, but
13        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14        causes the error). The code is unchanged, but I tidied the documentation.
16    4.  Inside a character class, PCRE always treated \R and \X as literals,
17        whereas Perl faults them if its -w option is set. I have changed PCRE so
18        that it faults them when PCRE_EXTRA is set.
20    5.  Added support for \N, which always matches any character other than
21        newline. (It is the same as "." when PCRE_DOTALL is not set.)
23    6.  When compiling pcregrep with newer versions of gcc which may have
24        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25        declared with attribute warn_unused_result" were given. Just casting the
26        result to (void) does not stop the warnings; a more elaborate fudge is
27        needed. I've used a macro to implement this.
29    7.  Minor change to pcretest.c to avoid a compiler warning.
31    8.  Added four artifical Unicode properties to help with an option to make
32        \s etc use properties (see next item). The new properties are: Xan
33        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
35    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36        use Unicode properties. (*UCP) at the start of a pattern can be used to set
37        this option. Modified pcretest to add /W to test this facility. Added
38        REG_UCP to make it available via the POSIX interface.
40    10. Added --line-buffered to pcregrep.
43    Version 8.02 19-Mar-2010
44    ------------------------
46    1.  The Unicode data tables have been updated to Unicode 5.2.0.
48    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
49        configured.
51    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
52        original author of that file, following a query about its status.
54    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
55        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
57    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
58        quantifier applied to a forward-referencing subroutine call, could compile
59        incorrect code or give the error "internal error: previously-checked
60        referenced subpattern not found".
62    6.  Both MS Visual Studio and Symbian OS have problems with initializing
63        variables to point to external functions. For these systems, therefore,
64        pcre_malloc etc. are now initialized to local functions that call the
65        relevant global functions.
67    7.  There were two entries missing in the vectors called coptable and poptable
68        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
69        I've fixed the data, and added a kludgy way of testing at compile time that
70        the lengths are correct (equal to the number of opcodes).
72    8.  Following on from 7, I added a similar kludge to check the length of the
73        eint vector in pcreposix.c.
75    9.  Error texts for pcre_compile() are held as one long string to avoid too
76        much relocation at load time. To find a text, the string is searched,
77        counting zeros. There was no check for running off the end of the string,
78        which could happen if a new error number was added without updating the
79        string.
81    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
83    11. \K was not working if it appeared in an atomic group or in a group that
84        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
85        \K is "not well defined" if used in an assertion. PCRE now accepts it if
86        the assertion is positive, but not if it is negative.
88    12. Change 11 fortuitously reduced the size of the stack frame used in the
89        "match()" function of pcre_exec.c by one pointer. Forthcoming
90        implementation of support for (*MARK) will need an extra pointer on the
91        stack; I have reserved it now, so that the stack frame size does not
92        decrease.
94    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
95        item in branch that calls a recursion is a subroutine call - as in the
96        second branch in the above example - was incorrectly given the compile-
97        time error "recursive call could loop indefinitely" because pcre_compile()
98        was not correctly checking the subroutine for matching a non-empty string.
100    14. The checks for overrunning compiling workspace could trigger after an
101        overrun had occurred. This is a "should never occur" error, but it can be
102        triggered by pathological patterns such as hundreds of nested parentheses.
103        The checks now trigger 100 bytes before the end of the workspace.
105    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
108    Version 8.01 19-Jan-2010
109    ------------------------
111    1.  If a pattern contained a conditional subpattern with only one branch (in
112        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
113        computed the wrong minimum data length (which is of course zero for such
114        subpatterns). This could cause incorrect "no match" results.
116    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
117        the pattern is reset in the first branch, pcre_compile() failed with
118        "internal error: code overflow at offset...". This happened only when
119        the reset was to the original external option setting. (An optimization
120        abstracts leading options settings into an external setting, which was the
121        cause of this.)
123    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
124        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
125        assertion pattern did not match (meaning that the assertion was true), it
126        was incorrectly treated as false if the SKIP had been reached during the
127        matching. This also applied to assertions used as conditions.
129    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
130        assertion subpattern, including such a pattern used as a condition,
131        unpredictable results occurred, instead of the error return
134    5.  The C++ GlobalReplace function was not working like Perl for the special
135        situation when an empty string is matched. It now does the fancy magic
136        stuff that is necessary.
138    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
139        removed. (These were left over from very, very early versions of PCRE.)
141    7.  Some cosmetic changes to the code to make life easier when compiling it
142        as part of something else:
144        (a) Change DEBUG to PCRE_DEBUG.
146        (b) In pcre_compile(), rename the member of the "branch_chain" structure
147            called "current" as "current_branch", to prevent a collision with the
148            Linux macro when compiled as a kernel module.
150        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
151            prevent a collision with the Linux macro when compiled as a kernel
152            module.
154    8.  In pcre_compile() there are some checks for integer overflows that used to
155        cast potentially large values to (double). This has been changed to that
156        when building, a check for int64_t is made, and if it is found, it is used
157        instead, thus avoiding the use of floating point arithmetic. (There is no
158        other use of FP in PCRE.) If int64_t is not found, the fallback is to
159        double.
161    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
162        2005 (difference between two addresses compared to an unsigned value).
164    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
165        custom one, because of the following reported problem in Windows:
167          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
168              under Win32.
169          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
170              therefore missing the function definition.
171          - The compiler thus generates a "C" signature for the test function.
172          - The linker fails to find the "C" function.
173          - PCRE fails to configure if asked to do so against libbz2.
175    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
176        messages were output:
178          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
179          rerunning libtoolize, to keep the correct libtool macros in-tree.
180          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
182        I have done both of these things.
184    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
185        most of the time, it *can* run out if it is given a pattern that contains a
186        runaway infinite recursion. I updated the discussion in the pcrestack man
187        page.
189    13. Now that we have gone to the x.xx style of version numbers, the minor
190        version may start with zero. Using 08 or 09 is a bad idea because users
191        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
192        interpreted as invalid octal numbers. I've updated the previous comment in
193        configure.ac, and also added a check that gives an error if 08 or 09 are
194        used.
196    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
197        causing partial matching to fail when the end of the subject matched \W
198        in a UTF-8 pattern where \W was quantified with a minimum of 3.
200    15. There were some discrepancies between the declarations in pcre_internal.h
201        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
202        their definitions. The declarations used "const uschar *" and the
203        definitions used USPTR. Even though USPTR is normally defined as "const
204        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
205        reported that: "This difference in casting confuses some C++ compilers, for
206        example, SunCC recognizes above declarations as different functions and
207        generates broken code for hbpcre." I have changed the declarations to use
208        USPTR.
210    16. GNU libtool is named differently on some systems. The autogen.sh script now
211        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
212        (FreeBSD).
214    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
215        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
216        comment: "Figure out how to create a longlong from a string: strtoll and
217        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
218        instance, but it only takes 2 args instead of 3!"
220    18. A subtle bug concerned with back references has been fixed by a change of
221        specification, with a corresponding code fix. A pattern such as
222        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
223        refers, was giving matches when it shouldn't. For example, xa=xaaa would
224        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
225        same bug. Such groups have to be quantified to be useful, or contained
226        inside another quantified group. (If there's no repetition, the reference
227        can never match.) The problem arises because, having left the group and
228        moved on to the rest of the pattern, a later failure that backtracks into
229        the group uses the captured value from the final iteration of the group
230        rather than the correct earlier one. I have fixed this in PCRE by forcing
231        any group that contains a reference to itself to be an atomic group; that
232        is, there cannot be any backtracking into it once it has completed. This is
233        similar to recursive and subroutine calls.
236    Version 8.00 19-Oct-09
237    ----------------------
239    1.  The table for translating pcre_compile() error codes into POSIX error codes
240        was out-of-date, and there was no check on the pcre_compile() error code
241        being within the table. This could lead to an OK return being given in
242        error.
244    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
245        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
246        in a Windows environment.
248    3.  The pcregrep --count option prints the count for each file even when it is
249        zero, as does GNU grep. However, pcregrep was also printing all files when
250        --files-with-matches was added. Now, when both options are given, it prints
251        counts only for those files that have at least one match. (GNU grep just
252        prints the file name in this circumstance, but including the count seems
253        more useful - otherwise, why use --count?) Also ensured that the
254        combination -clh just lists non-zero counts, with no names.
256    4.  The long form of the pcregrep -F option was incorrectly implemented as
257        --fixed_strings instead of --fixed-strings. This is an incompatible change,
258        but it seems right to fix it, and I didn't think it was worth preserving
259        the old behaviour.
261    5.  The command line items --regex=pattern and --regexp=pattern were not
262        recognized by pcregrep, which required --regex pattern or --regexp pattern
263        (with a space rather than an '='). The man page documented the '=' forms,
264        which are compatible with GNU grep; these now work.
266    6.  No libpcreposix.pc file was created for pkg-config; there was just
267        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
269    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
270        when UCP support is not needed, by modifying the Python script that
271        generates it from Unicode data files. This should not matter if the module
272        is correctly used as a library, but I received one complaint about 50K of
273        unwanted data. My guess is that the person linked everything into his
274        program rather than using a library. Anyway, it does no harm.
276    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
277        was a minimum greater than 1 for a wide character in a possessive
278        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
279        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
280        character. Chaos in the form of incorrect output or a compiling loop could
281        result.
283    9.  The restrictions on what a pattern can contain when partial matching is
284        requested for pcre_exec() have been removed. All patterns can now be
285        partially matched by this function. In addition, if there are at least two
286        slots in the offset vector, the offset of the earliest inspected character
287        for the match and the offset of the end of the subject are set in them when
288        PCRE_ERROR_PARTIAL is returned.
290    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
291        synonymous with PCRE_PARTIAL, for backwards compatibility, and
292        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
293        and may be more useful for multi-segment matching.
295    11. Partial matching with pcre_exec() is now more intuitive. A partial match
296        used to be given if ever the end of the subject was reached; now it is
297        given only if matching could not proceed because another character was
298        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
299        string "Z", which now yields "no match" instead of "partial match". In the
300        case of pcre_dfa_exec(), "no match" is given if every matching path for the
301        final character ended with (*FAIL).
303    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
304        if the pattern had a "must contain" character that was already found in the
305        earlier partial match, unless partial matching was again requested. For
306        example, with the pattern /dog.(body)?/, the "must contain" character is
307        "g". If the first part-match was for the string "dog", restarting with
308        "sbody" failed. This bug has been fixed.
310    13. The string returned by pcre_dfa_exec() after a partial match has been
311        changed so that it starts at the first inspected character rather than the
312        first character of the match. This makes a difference only if the pattern
313        starts with a lookbehind assertion or \b or \B (\K is not supported by
314        pcre_dfa_exec()). It's an incompatible change, but it makes the two
315        matching functions compatible, and I think it's the right thing to do.
317    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
318        so that the demonstration program is easily available in environments where
319        PCRE has not been installed from source.
321    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
322        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
323        library.
325    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
326        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
327        is not the first non-POSIX option to be added. Clearly some people find
328        these options useful.
330    17. If a caller to the POSIX matching function regexec() passes a non-zero
331        value for nmatch with a NULL value for pmatch, the value of
332        nmatch is forced to zero.
334    18. RunGrepTest did not have a test for the availability of the -u option of
335        the diff command, as RunTest does. It now checks in the same way as
336        RunTest, and also checks for the -b option.
338    19. If an odd number of negated classes containing just a single character
339        interposed, within parentheses, between a forward reference to a named
340        subpattern and the definition of the subpattern, compilation crashed with
341        an internal error, complaining that it could not find the referenced
342        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
343        [The bug was that it was starting one character too far in when skipping
344        over the character class, thus treating the ] as data rather than
345        terminating the class. This meant it could skip too much.]
347    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
348        /g option in pcretest when the pattern contains \K, which makes it possible
349        to have an empty string match not at the start, even when the pattern is
350        anchored. Updated pcretest and pcredemo to use this option.
352    21. If the maximum number of capturing subpatterns in a recursion was greater
353        than the maximum at the outer level, the higher number was returned, but
354        with unset values at the outer level. The correct (outer level) value is
355        now given.
357    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
358        PCRE did not set those parentheses (unlike Perl). I have now found a way to
359        make it do so. The string so far is captured, making this feature
360        compatible with Perl.
362    23. The tests have been re-organized, adding tests 11 and 12, to make it
363        possible to check the Perl 5.10 features against Perl 5.10.
365    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
366        pattern matches a fixed length string. PCRE did not allow this; now it
367        does. Neither allows recursion.
369    25. I finally figured out how to implement a request to provide the minimum
370        length of subject string that was needed in order to match a given pattern.
371        (It was back references and recursion that I had previously got hung up
372        on.) This code has now been added to pcre_study(); it finds a lower bound
373        to the length of subject needed. It is not necessarily the greatest lower
374        bound, but using it to avoid searching strings that are too short does give
375        some useful speed-ups. The value is available to calling programs via
376        pcre_fullinfo().
378    26. While implementing 25, I discovered to my embarrassment that pcretest had
379        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
380        study optimizations had never been tested with that matching function.
381        Oops. What is worse, even when it was passed study data, there was a bug in
382        pcre_dfa_exec() that meant it never actually used it. Double oops. There
383        were also very few tests of studied patterns with pcre_dfa_exec().
385    27. If (?| is used to create subpatterns with duplicate numbers, they are now
386        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
387        on the other side of the coin, they are no longer allowed to have different
388        names, because these cannot be distinguished in PCRE, and this has caused
389        confusion. (This is a difference from Perl.)
391    28. When duplicate subpattern names are present (necessarily with different
392        numbers, as required by 27 above), and a test is made by name in a
393        conditional pattern, either for a subpattern having been matched, or for
394        recursion in such a pattern, all the associated numbered subpatterns are
395        tested, and the overall condition is true if the condition is true for any
396        one of them. This is the way Perl works, and is also more like the way
397        testing by number works.
400    Version 7.9 11-Apr-09
401    ---------------------
403    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
404        (pcretest), all targets were linked against these libraries. This included
405        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
406        libraries. This caused unwanted dependencies to be created. This problem
407        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
408        pcretest is linked with readline.
410    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
411        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
412        moved outside it again, because FALSE and TRUE are already defined in AIX,
413        but BOOL is not.
415    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
416        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
418    4.  The pcregrep documentation said spaces were inserted as well as colons (or
419        hyphens) following file names and line numbers when outputting matching
420        lines. This is not true; no spaces are inserted. I have also clarified the
421        wording for the --colour (or --color) option.
423    5.  In pcregrep, when --colour was used with -o, the list of matching strings
424        was not coloured; this is different to GNU grep, so I have changed it to be
425        the same.
427    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
428        each matching line was coloured. Now it goes on to look for further matches
429        of any of the test patterns, which is the same behaviour as GNU grep.
431    7.  A pattern that could match an empty string could cause pcregrep to loop; it
432        doesn't make sense to accept an empty string match in pcregrep, so I have
433        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
434        seems to be how GNU grep behaves.
436    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
437        start or after a newline", because the conditional assertion was not being
438        correctly handled. The rule now is that both the assertion and what follows
439        in the first alternative must satisfy the test.
441    9.  If auto-callout was enabled in a pattern with a conditional group whose
442        condition was an assertion, PCRE could crash during matching, both with
443        pcre_exec() and pcre_dfa_exec().
445    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
446        used for matching.
448    11. Unicode property support in character classes was not working for
449        characters (bytes) greater than 127 when not in UTF-8 mode.
451    12. Added the -M command line option to pcretest.
453    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
455    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
457    16. Added comments and documentation about mis-use of no_arg in the C++
458        wrapper.
460    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
461        from Martin Jerabek that uses macro names for all relevant character and
462        string constants.
464    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
465        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
466        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
467        these, but not everybody uses configure.
469    19. A conditional group that had only one branch was not being correctly
470        recognized as an item that could match an empty string. This meant that an
471        enclosing group might also not be so recognized, causing infinite looping
472        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
473        with the subject "ab", where knowledge that the repeated group can match
474        nothing is needed in order to break the loop.
476    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
477        exec(), but without supplying a callout function, matching went wrong.
479    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
480        leak if the size of the offset vector was greater than 30. When the vector
481        is smaller, the saved offsets during recursion go onto a local stack
482        vector, but for larger vectors malloc() is used. It was failing to free
483        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
484        error, in fact).
486    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
487        heapframe that is used only when UTF-8 support is enabled. This caused no
488        problem, but was untidy.
490    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
491        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
492        included within another project.
494    24. Steven Van Ingelgem's patches to add more options to the CMake support,
495        slightly modified by me:
497          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
498              not building pcregrep.
500          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
501              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
503    25. Forward references, both numeric and by name, in patterns that made use of
504        duplicate group numbers, could behave incorrectly or give incorrect errors,
505        because when scanning forward to find the reference group, PCRE was not
506        taking into account the duplicate group numbers. A pattern such as
507        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
509    26. Changed a few more instances of "const unsigned char *" to USPTR, making
510        the feature of a custom pointer more persuasive (as requested by a user).
512    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
513        pcretest.c, inside #ifndefs, because it seems they are sometimes already
514        pre-defined.
516    28. Added support for (*UTF8) at the start of a pattern.
518    29. Arrange for flags added by the "release type" setting in CMake to be shown
519        in the configuration summary.
522    Version 7.8 05-Sep-08
523    ---------------------
525    1.  Replaced UCP searching code with optimized version as implemented for Ad
526        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
527        stage table and inline lookup instead of a function, giving speed ups of 2
528        to 5 times on some simple patterns that I tested. Permission was given to
529        distribute the MultiStage2.py script that generates the tables (it's not in
530        the tarball, but is in the Subversion repository).
532    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
533        scripts.
535    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
536        a group with a zero qualifier. The result of the study could be incorrect,
537        or the function might crash, depending on the pattern.
539    4.  Caseless matching was not working for non-ASCII characters in back
540        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
541        It now works when Unicode Property Support is available.
543    5.  In pcretest, an escape such as \x{de} in the data was always generating
544        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
545        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
546        truncation.
548    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
550    7.  Added two (int) casts to pcregrep when printing the difference of two
551        pointers, in case they are 64-bit values.
553    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
554        test 2 if it fails.
556    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
557        and a #define of that name to empty if it is not externally set. This is to
558        allow users of MSVC to set it if necessary.
560    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
561        the convenience functions in the pcre_get.c source file.
563    11. An option change at the start of a pattern that had top-level alternatives
564        could cause overwriting and/or a crash. This command provoked a crash in
565        some environments:
567          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
569        This potential security problem was recorded as CVE-2008-2371.
571    12. For a pattern where the match had to start at the beginning or immediately
572        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
573        pcre_dfa_exec() could read past the end of the passed subject if there was
574        no match. To help with detecting such bugs (e.g. with valgrind), I modified
575        pcretest so that it places the subject at the end of its malloc-ed buffer.
577    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
578        exec() might read past the end of the data buffer in UTF-8 mode.
580    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
581        the data contained the byte 0x85 as part of a UTF-8 character within its
582        first line. This applied both to normal and DFA matching.
584    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
585        /^[^d]*?$/8 failed to match "abc".
587    16. Added a missing copyright notice to pcrecpp_internal.h.
589    17. Make it more clear in the documentation that values returned from
590        pcre_exec() in ovector are byte offsets, not character counts.
592    18. Tidied a few places to stop certain compilers from issuing warnings.
594    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
595        supplied by Stefan Weber. I made a further small update for 7.8 because
596        there is a change of source arrangements: the pcre_searchfuncs.c module is
597        replaced by pcre_ucd.c.
600    Version 7.7 07-May-08
601    ---------------------
603    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
604        a string to a long long, pretend we don't even have a long long." This is
605        done by checking for the strtoq, strtoll, and _strtoi64 functions.
607    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
608        pre-7.6 versions, which defined a global no_arg variable instead of putting
609        it in the RE class. (See also #8 below.)
611    3.  Remove a line of dead code, identified by coverity and reported by Nuno
612        Lopes.
614    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
616        (1) The include/exclude patterns were being applied to the whole pathnames
617            of files, instead of just to the final components.
619        (2) If there was more than one level of directory, the subdirectories were
620            skipped unless they satisfied the include/exclude conditions. This is
621            inconsistent with GNU grep (and could even be seen as contrary to the
622            pcregrep specification - which I improved to make it absolutely clear).
623            The action now is always to scan all levels of directory, and just
624            apply the include/exclude patterns to regular files.
626    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
627        --exclude_dir in the tests to avoid scanning .svn directories.
629    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
630        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
631        doesn't support NULs in patterns.
633    7.  Added some missing "const"s to declarations of static tables in
634        pcre_compile.c and pcre_dfa_exec.c.
636    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
637        caused by fix #2  above. (Subsequently also a second patch to fix the
638        first patch. And a third patch - this was a messy problem.)
640    9.  Applied Craig's patch to remove the use of push_back().
642    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
643        matching function regexec().
645    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
646        which, however, unlike Perl's \g{...}, are subroutine calls, not back
647        references. PCRE supports relative numbers with this syntax (I don't think
648        Oniguruma does).
650    12. Previously, a group with a zero repeat such as (...){0} was completely
651        omitted from the compiled regex. However, this means that if the group
652        was called as a subroutine from elsewhere in the pattern, things went wrong
653        (an internal error was given). Such groups are now left in the compiled
654        pattern, with a new opcode that causes them to be skipped at execution
655        time.
657    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
658        to the way PCRE behaves:
660        (a) A lone ] character is dis-allowed (Perl treats it as data).
662        (b) A back reference to an unmatched subpattern matches an empty string
663            (Perl fails the current match path).
665        (c) A data ] in a character class must be notated as \] because if the
666            first data character in a class is ], it defines an empty class. (In
667            Perl it is not possible to have an empty class.) The empty class []
668            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
669            The negative empty class [^] matches any one character, independently
670            of the DOTALL setting.
672    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
673        non-existent subpattern following a character class starting with ']' and
674        containing () gave an internal compiling error instead of "reference to
675        non-existent subpattern". Fortunately, when the pattern did exist, the
676        compiled code was correct. (When scanning forwards to check for the
677        existencd of the subpattern, it was treating the data ']' as terminating
678        the class, so got the count wrong. When actually compiling, the reference
679        was subsequently set up correctly.)
681    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
682        it was being rejected as not supported by pcre_dfa_exec(), even though
683        other assertions are supported. I have made pcre_dfa_exec() support
684        (*FAIL).
686    16. The implementation of 13c above involved the invention of a new opcode,
687        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
688        cannot be changed at match time, I realized I could make a small
689        improvement to matching performance by compiling OP_ALLANY instead of
690        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
691        on the OP_ANY path.
693    17. Compiling pcretest on Windows with readline support failed without the
694        following two fixes: (1) Make the unistd.h include conditional on
695        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
697    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
698        ncurses library to be included for pcretest when ReadLine support is
699        requested, but also to allow for it to be overridden. This patch came from
700        Daniel Bergström.
702    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
703        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
704        any errors with the current Unicode tables. Thanks to Peter Kankowski for
705        spotting this.
708    Version 7.6 28-Jan-08
709    ---------------------
711    1.  A character class containing a very large number of characters with
712        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
713        overflow.
715    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
716        HAVE_LONG_LONG is not defined.
718    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
719        bring them up to date and include new features. This patch includes:
721        - Fixed PH's badly added libz and libbz2 support.
722        - Fixed a problem with static linking.
723        - Added pcredemo. [But later removed - see 7 below.]
724        - Fixed dftables problem and added an option.
725        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
726            HAVE_LONG_LONG.
727        - Added readline support for pcretest.
728        - Added an listing of the option settings after cmake has run.
730    4.  A user submitted a patch to Makefile that makes it easy to create
731        "pcre.dll" under mingw when using Configure/Make. I added stuff to
732        Makefile.am that cause it to include this special target, without
733        affecting anything else. Note that the same mingw target plus all
734        the other distribution libraries and programs are now supported
735        when configuring with CMake (see 6 below) instead of with
736        Configure/Make.
738    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
739        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
740        exported in the Windows port". It has not yet been confirmed that the patch
741        solves the problem, but it does no harm.
743    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
744        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
745        with CMake, and also correct the comment about stack recursion.
747    7.  Remove the automatic building of pcredemo from the ./configure system and
748        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
749        of a program that users should build themselves after PCRE is installed, so
750        building it automatically is not really right. What is more, it gave
751        trouble in some build environments.
753    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
756    Version 7.5 10-Jan-08
757    ---------------------
759    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
760        values in parens when parsing an RE using the C++ wrapper."
762    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
763        Characters greater than 255 were excluded from the class instead of being
764        included.
766    3.  The same bug as (2) above applied to negated POSIX classes such as
767        [:^space:].
769    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
770        defined or documented. It seems to have been a typo for PCRE_STATIC, so
771        I have changed it.
773    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
774        first named subpattern) and a construct such as (?&a) would reference the
775        first named subpattern whose name started with "a" (in other words, the
776        length check was missing). Both these problems are fixed. "Subpattern name
777        expected" is now given for (?&) (a zero-length name), and this patch also
778        makes it give the same error for \k'' (previously it complained that that
779        was a reference to a non-existent subpattern).
781    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
782        this is right because (?- can be followed by option settings as well as by
783        digits. I have, however, made the messages clearer.
785    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
786        than the number used in the conditional) now cause a compile-time error.
787        This is actually not compatible with Perl, which accepts such patterns, but
788        treats the conditional as always being FALSE (as PCRE used to), but it
789        seems to me that giving a diagnostic is better.
791    8.  Change "alphameric" to the more common word "alphanumeric" in comments
792        and messages.
794    9.  Fix two occurrences of "backslash" in comments that should have been
795        "backspace".
797    10. Remove two redundant lines of code that can never be obeyed (their function
798        was moved elsewhere).
800    11. The program that makes PCRE's Unicode character property table had a bug
801        which caused it to generate incorrect table entries for sequences of
802        characters that have the same character type, but are in different scripts.
803        It amalgamated them into a single range, with the script of the first of
804        them. In other words, some characters were in the wrong script. There were
805        thirteen such cases, affecting characters in the following ranges:
807          U+002b0 - U+002c1
808          U+0060c - U+0060d
809          U+0061e - U+00612
810          U+0064b - U+0065e
811          U+0074d - U+0076d
812          U+01800 - U+01805
813          U+01d00 - U+01d77
814          U+01d9b - U+01dbf
815          U+0200b - U+0200f
816          U+030fc - U+030fe
817          U+03260 - U+0327f
818          U+0fb46 - U+0fbb1
819          U+10450 - U+1049d
821    12. The -o option (show only the matching part of a line) for pcregrep was not
822        compatible with GNU grep in that, if there was more than one match in a
823        line, it showed only the first of them. It now behaves in the same way as
824        GNU grep.
826    13. If the -o and -v options were combined for pcregrep, it printed a blank
827        line for every non-matching line. GNU grep prints nothing, and pcregrep now
828        does the same. The return code can be used to tell if there were any
829        non-matching lines.
831    14. Added --file-offsets and --line-offsets to pcregrep.
833    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
834        infinitely looping recursion. The bug was that positive lookaheads were not
835        being skipped when checking for a possible empty match (negative lookaheads
836        and both kinds of lookbehind were skipped).
838    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
839        inclusion of <windows.h> to before rather than after the definition of
840        INVALID_FILE_ATTRIBUTES (patch from David Byron).
842    17. Specifying a possessive quantifier with a specific limit for a Unicode
843        character property caused pcre_compile() to compile bad code, which led at
844        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
845        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
846        caused the error; without that there was no problem.
848    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
850    19. Added --enable-pcretest-libreadline.
852    20. In pcrecpp.cc, the variable 'count' was incremented twice in
853        RE::GlobalReplace(). As a result, the number of replacements returned was
854        double what it should be. I removed one of the increments, but Craig sent a
855        later patch that removed the other one (the right fix) and added unit tests
856        that check the return values (which was not done before).
858    21. Several CMake things:
860        (1) Arranged that, when cmake is used on Unix, the libraries end up with
861            the names libpcre and libpcreposix, not just pcre and pcreposix.
863        (2) The above change means that pcretest and pcregrep are now correctly
864            linked with the newly-built libraries, not previously installed ones.
868    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
869        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
870        UTF-8 newline character). The key issue is that the pattern starts .*;
871        this means that the match must be either at the beginning, or after a
872        newline. The bug was in the code for advancing after a failed match and
873        checking that the new position followed a newline. It was not taking
874        account of UTF-8 characters correctly.
876    23. PCRE was behaving differently from Perl in the way it recognized POSIX
877        character classes. PCRE was not treating the sequence [:...:] as a
878        character class unless the ... were all letters. Perl, however, seems to
879        allow any characters between [: and :], though of course it rejects as
880        unknown any "names" that contain non-letters, because all the known class
881        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
882        for example, whereas PCRE did not - it did not recognize a POSIX character
883        class. This seemed a bit dangerous, so the code has been changed to be
884        closer to Perl. The behaviour is not identical to Perl, because PCRE will
885        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
886        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
887        Perl does, and where it didn't before.
889    24. Rewrite so as to remove the single use of %n from pcregrep because in some
890        Windows environments %n is disabled by default.
893    Version 7.4 21-Sep-07
894    ---------------------
896    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
897        means that a class such as [\s] counted as "explicit reference to CR or
898        LF". That isn't really right - the whole point of the change was to try to
899        help when there was an actual mention of one of the two characters. So now
900        the change happens only if \r or \n (or a literal CR or LF) character is
901        encountered.
903    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
904        of both had grown to the point where there were only 3 bits left.
905        Fortunately, there was spare space in the data structure, and so I have
906        moved the internal flags into a new 16-bit field to free up more option
907        bits.
909    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
910        but did not set the internal JCHANGED flag - either of these is enough to
911        control the way the "get" function works - but the PCRE_INFO_JCHANGED
912        facility is supposed to tell if (?J) was ever used, so now (?J) at the
913        start sets both bits.
915    4.  Added options (at build time, compile time, exec time) to change \R from
916        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
918    5.  doc/pcresyntax.html was missing from the distribution.
920    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
921        compatibility, even though it is no longer used.
923    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
924        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
925        windows.h file is present (where different names are used). [This was
926        reversed later after testing - see 16 below.]
928    8.  Changed all #include <config.h> to #include "config.h". There were also
929        some further <pcre.h> cases that I changed to "pcre.h".
931    9.  When pcregrep was used with the --colour option, it missed the line ending
932        sequence off the lines that it output.
934    10. It was pointed out to me that arrays of string pointers cause lots of
935        relocations when a shared library is dynamically loaded. A technique of
936        using a single long string with a table of offsets can drastically reduce
937        these. I have refactored PCRE in four places to do this. The result is
938        dramatic:
940          Originally:                          290
941          After changing UCP table:            187
942          After changing error message table:   43
943          After changing table of "verbs"       36
944          After changing table of Posix names   22
946        Thanks to the folks working on Gregex for glib for this insight.
948    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
949        unicode-properties was also set.
951    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
953    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
954        checked only for CRLF.
956    14. Added casts to pcretest.c to avoid compiler warnings.
958    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
960    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
961        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
962        entirely. This removes changes made in 7 above.
964    17. The CMake files have been updated, and there is now more information about
965        building with CMake in the NON-UNIX-USE document.
968    Version 7.3 28-Aug-07
969    ---------------------
971     1. In the rejigging of the build system that eventually resulted in 7.1, the
972        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
973        brackets there is not right, since it causes compilers to look for an
974        installed pcre.h, not the version that is in the source that is being
975        compiled (which of course may be different). I have changed it back to:
977          #include "pcre.h"
979        I have a vague recollection that the change was concerned with compiling in
980        different directories, but in the new build system, that is taken care of
981        by the VPATH setting the Makefile.
983     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
984        when the subject happened to end in the byte 0x85 (e.g. if the last
985        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
986        characters but of course it shouldn't be taken as a newline when it is part
987        of another character. The bug was that, for an unlimited repeat of . in
988        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
989        characters when looking for a newline.
991     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
993     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
994        in debug output.
996     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
997        long printing in the pcrecpp unittest when running under MinGW.
999     6. ESC_K was left out of the EBCDIC table.
1001     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1002        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1003        limit also applies to "virtual nesting" when a pattern is recursive, and in
1004        this case 1000 isn't so big. I have been able to remove this limit at the
1005        expense of backing off one optimization in certain circumstances. Normally,
1006        when pcre_exec() would call its internal match() function recursively and
1007        immediately return the result unconditionally, it uses a "tail recursion"
1008        feature to save stack. However, when a subpattern that can match an empty
1009        string has an unlimited repetition quantifier, it no longer makes this
1010        optimization. That gives it a stack frame in which to save the data for
1011        checking that an empty string has been matched. Previously this was taken
1012        from the 1000-entry workspace that had been reserved. So now there is no
1013        explicit limit, but more stack is used.
1015     8. Applied Daniel's patches to solve problems with the import/export magic
1016        syntax that is required for Windows, and which was going wrong for the
1017        pcreposix and pcrecpp parts of the library. These were overlooked when this
1018        problem was solved for the main library.
1020     9. There were some crude static tests to avoid integer overflow when computing
1021        the size of patterns that contain repeated groups with explicit upper
1022        limits. As the maximum quantifier is 65535, the maximum group length was
1023        set at 30,000 so that the product of these two numbers did not overflow a
1024        32-bit integer. However, it turns out that people want to use groups that
1025        are longer than 30,000 bytes (though not repeat them that many times).
1026        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1027        made it possible to implement the integer overflow checks in a much more
1028        dynamic way, which I have now done. The artificial limitation on group
1029        length has been removed - we now have only the limit on the total length of
1030        the compiled pattern, which depends on the LINK_SIZE setting.
1032    10. Fixed a bug in the documentation for get/copy named substring when
1033        duplicate names are permitted. If none of the named substrings are set, the
1034        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1035        empty string.
1037    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1038        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1039        because the ] is interpreted as the first data character and the
1040        terminating ] is not found. PCRE has been made compatible with Perl in this
1041        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1042        cause memory overwriting.
1044    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1045        string has been matched (to stop an infinite loop). It was not recognizing
1046        a conditional subpattern that could match an empty string if that
1047        subpattern was within another subpattern. For example, it looped when
1048        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1049        condition was not nested. This bug has been fixed.
1051    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1052        past the start of the subject in the presence of bytes with the top bit
1053        set, for example "\x8aBCD".
1055    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1056        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1058    14. Optimized (?!) to (*FAIL).
1060    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1061        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1062        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1063        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1064        does: it's just the validity check that is more restrictive.
1066    16. Inserted checks for integer overflows during escape sequence (backslash)
1067        processing, and also fixed erroneous offset values for syntax errors during
1068        backslash processing.
1070    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1071        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1073    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1074        caused an overrun.
1076    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1077        something other than just ASCII characters) inside a group that had an
1078        unlimited repeat caused a loop at compile time (while checking to see
1079        whether the group could match an empty string).
1081    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1082        [\P{Any}] did so. (Error in the code for printing property names.)
1084    21. An orphan \E inside a character class could cause a crash.
1086    22. A repeated capturing bracket such as (A)? could cause a wild memory
1087        reference during compilation.
1089    23. There are several functions in pcre_compile() that scan along a compiled
1090        expression for various reasons (e.g. to see if it's fixed length for look
1091        behind). There were bugs in these functions when a repeated \p or \P was
1092        present in the pattern. These operators have additional parameters compared
1093        with \d, etc, and these were not being taken into account when moving along
1094        the compiled data. Specifically:
1096        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1097            length.
1099        (b) An item such as \pL+ within a repeated group could cause crashes or
1100            loops.
1102        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1103            "reference to non-existent subpattern" error.
1105        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1107    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1108        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1110    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1112    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1113        character were causing crashes (broken optimization).
1115    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1116        \p or \P) caused a compile-time loop.
1118    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1119        break. For example, the unstudied pattern [\r\n]A does not match the string
1120        "\r\nA" because change 7.0/46 below moves the current point on by two
1121        characters after failing to match at the start. However, the pattern \nA
1122        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1123        the same is true. There doesn't seem any very clean way out of this, but
1124        what I have chosen to do makes the common cases work: PCRE now takes note
1125        of whether there can be an explicit match for \r or \n anywhere in the
1126        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1127        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1128        pattern has explicit CR or LF references.
1130    29. Added (*CR) etc for changing newline setting at start of pattern.
1133    Version 7.2 19-Jun-07
1134    ---------------------
1136     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1137        which is apparently normally available under Windows.
1139     2. Re-jig the pcregrep tests with different newline settings in an attempt
1140        to make them independent of the local environment's newline setting.
1142     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1144     4. Some of the "internals" tests were previously cut out when the link size
1145        was not 2, because the output contained actual offsets. The recent new
1146        "Z" feature of pcretest means that these can be cut out, making the tests
1147        usable with all link sizes.
1149     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1150        stack recursion. This gives a massive performance boost under BSD, but just
1151        a small improvement under Linux. However, it saves one field in the frame
1152        in all cases.
1154     6. Added more features from the forthcoming Perl 5.10:
1156        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1157            recursion call. It refers to the nth most recently opened parentheses.
1159        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1160            to be opened parentheses.
1162        (c) Conditions that refer to capturing parentheses can be specified
1163            relatively, for example, (?(-2)... or (?(+3)...
1165        (d) \K resets the start of the current match so that everything before
1166            is not part of it.
1168        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1170        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1171            reference syntax.
1173        (g) (?| introduces a group in which the numbering of parentheses in each
1174            alternative starts with the same number.
1176        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1178     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1181     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1182        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1183        for detecting groups that can match an empty string.
1185     9. A pattern with a very large number of alternatives (more than several
1186        hundred) was running out of internal workspace during the pre-compile
1187        phase, where pcre_compile() figures out how much memory will be needed. A
1188        bit of new cunning has reduced the workspace needed for groups with
1189        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1190        workspace instead of running out of the 4096 that are available.
1192    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1194    11. Applied patch from Google to remove an optimization that didn't quite work.
1195        The report of the bug said:
1197          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1198          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1199          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1201    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1202        it matched the wrong number of bytes.
1205    Version 7.1 24-Apr-07
1206    ---------------------
1208     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1209        that is more "standard", making use of automake and other Autotools. There
1210        is some re-arrangement of the files and adjustment of comments consequent
1211        on this.
1213     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1214        for recursive directory scanning broke on some systems because the files
1215        are not scanned in any specific order and on different systems the order
1216        was different. A call to "sort" has been inserted into RunGrepTest for the
1217        approprate test as a short-term fix. In the longer term there may be an
1218        alternative.
1220     3. I had an email from Eric Raymond about problems translating some of PCRE's
1221        man pages to HTML (despite the fact that I distribute HTML pages, some
1222        people do their own conversions for various reasons). The problems
1223        concerned the use of low-level troff macros .br and .in. I have therefore
1224        removed all such uses from the man pages (some were redundant, some could
1225        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1226        HTML has been updated to handle .nf/.fi and to complain if it encounters
1227        .br or .in.
1229     4. Updated comments in configure.ac that get placed in config.h.in and also
1230        arranged for config.h to be included in the distribution, with the name
1231        config.h.generic, for the benefit of those who have to compile without
1232        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1234     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1235        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1236        makevp.bat for the current PCRE, using the additional files
1237        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1239     6. A Windows user reported a minor discrepancy with test 2, which turned out
1240        to be caused by a trailing space on an input line that had got lost in his
1241        copy. The trailing space was an accident, so I've just removed it.
1243     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1244        that is needed.
1246     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1247        as "const" (a) because they are and (b) because it helps the PHP
1248        maintainers who have recently made a script to detect big data structures
1249        in the php code that should be moved to the .rodata section. I remembered
1250        to update Builducptable as well, so it won't revert if ucptable.h is ever
1251        re-created.
1253     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1254        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1255        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1256        support is not required. This saves 1.5-2K of code, which is important in
1257        some applications.
1259        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1260        so as not to refer to the tables, even though these functions will never be
1261        called when UTF-8 support is disabled. Otherwise there are problems with a
1262        shared library.
1264    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1266        (a) It was defining its arguments as char * instead of void *.
1268        (b) It was assuming that all moves were upwards in memory; this was true
1269            a long time ago when I wrote it, but is no longer the case.
1271        The emulated memove() is provided for those environments that have neither
1272        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1273        is clearly not the case, as these two bugs were recently reported.
1275    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1276        and Detrail to create the HTML documentation, the .txt form of the man
1277        pages, and it removes trailing spaces from listed files. It also creates
1278        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1279        case, it wraps all the #defines with #ifndefs. This script should be run
1280        before "make dist".
1282    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1283        with Unicode property support.
1285        (a) For a maximizing quantifier, if the two different cases of the
1286            character were of different lengths in their UTF-8 codings (there are
1287            some cases like this - I found 11), and the matching function had to
1288            back up over a mixture of the two cases, it incorrectly assumed they
1289            were both the same length.
1291        (b) When PCRE was configured to use the heap rather than the stack for
1292            recursion during matching, it was not correctly preserving the data for
1293            the other case of a UTF-8 character when checking ahead for a match
1294            while processing a minimizing repeat. If the check also involved
1295            matching a wide character, but failed, corruption could cause an
1296            erroneous result when trying to check for a repeat of the original
1297            character.
1299    13. Some tidying changes to the testing mechanism:
1301        (a) The RunTest script now detects the internal link size and whether there
1302            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1303            values substituted by "configure". (The RunGrepTest script already did
1304            this for UTF-8.) The configure.ac script no longer substitutes the
1305            relevant variables.
1307        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1308            with length and offset values. This means that the output is different
1309            for different internal link sizes. Test 2 is skipped for link sizes
1310            other than 2 because of this, bypassing the problem. Unfortunately,
1311            there was also a test in test 3 (the locale tests) that used /B and
1312            failed for link sizes other than 2. Rather than cut the whole test out,
1313            I have added a new /Z option to pcretest that replaces the length and
1314            offset values with spaces. This is now used to make test 3 independent
1315            of link size. (Test 2 will be tidied up later.)
1317    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1318        segmentation fault instead of returning the appropriate error message.
1320    15. In multiline mode when the newline sequence was set to "any", the pattern
1321        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1322        This doesn't seem right; it now treats the CRLF combination as the line
1323        ending, and so does not match in that case. It's only a pattern such as ^$
1324        that would hit this one: something like ^ABC$ would have failed after \r
1325        and then tried again after \r\n.
1327    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1328        in an attempt to make files that differ only in their line terminators
1329        compare equal. This works on Linux.
1331    17. Under certain error circumstances pcregrep might try to free random memory
1332        as it exited. This is now fixed, thanks to valgrind.
1334    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1335        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1336        was because its rules for how to advance for /g after matching an empty
1337        string at the end of a line did not allow for this case. They now check for
1338        it specially.
1340    20. pcretest is supposed to handle patterns and data of any length, by
1341        extending its buffers when necessary. It was getting this wrong when the
1342        buffer for a data line had to be extended.
1344    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1345        CRLF as a newline sequence.
1347    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1348        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1349        I have nevertheless tidied it up.
1351    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1353    24. Added a man page for pcre-config.
1356    Version 7.0 19-Dec-06
1357    ---------------------
1359     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1360        moving to gcc 4.1.1.
1362     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1363        sys/time.h, which is documented as needed for this function. It doesn't
1364        seem to matter on Linux, but it showed up on some releases of OS X.
1366     3. It seems that there are systems where bytes whose values are greater than
1367        127 match isprint() in the "C" locale. The "C" locale should be the
1368        default when a C program starts up. In most systems, only ASCII printing
1369        characters match isprint(). This difference caused the output from pcretest
1370        to vary, making some of the tests fail. I have changed pcretest so that:
1372        (a) When it is outputting text in the compiled version of a pattern, bytes
1373            other than 32-126 are always shown as hex escapes.
1375        (b) When it is outputting text that is a matched part of a subject string,
1376            it does the same, unless a different locale has been set for the match
1377            (using the /L modifier). In this case, it uses isprint() to decide.
1379     4. Fixed a major bug that caused incorrect computation of the amount of memory
1380        required for a compiled pattern when options that changed within the
1381        pattern affected the logic of the preliminary scan that determines the
1382        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1383        that the computed length was too small. The symptoms of this bug were
1384        either the PCRE error "internal error: code overflow" from pcre_compile(),
1385        or a glibc crash with a message such as "pcretest: free(): invalid next
1386        size (fast)". Examples of patterns that provoked this bug (shown in
1387        pcretest format) are:
1389          /(?-x: )/x
1390          /(?x)(?-x: \s*#\s*)/
1391          /((?i)[\x{c0}])/8
1392          /(?i:[\x{c0}])/8
1394        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1395        is now done differently.
1397     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1398        wrapper classes; (b) implement a new function in the C++ scanner that is
1399        more efficient than the old way of doing things because it avoids levels of
1400        recursion in the regex matching; (c) add a paragraph to the documentation
1401        for the FullMatch() function.
1403     6. The escape sequence \n was being treated as whatever was defined as
1404        "newline". Not only was this contrary to the documentation, which states
1405        that \n is character 10 (hex 0A), but it also went horribly wrong when
1406        "newline" was defined as CRLF. This has been fixed.
1408     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1409        was being set to -1 for the "end of line" case (supposedly a value that no
1410        character can have). Though this value is never used (the check for end of
1411        line is "zero bytes in current character"), it caused compiler complaints.
1412        I've changed it to 0xffffffff.
1414     8. In pcre_version.c, the version string was being built by a sequence of
1415        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1416        string (as it is for production releases) called a macro with an empty
1417        argument. The C standard says the result of this is undefined. The gcc
1418        compiler treats it as an empty string (which was what was wanted) but it is
1419        reported that Visual C gives an error. The source has been hacked around to
1420        avoid this problem.
1422     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1423        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1424        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1425        of them did).
1427    10. Originally, pcretest opened its input and output without "b"; then I was
1428        told that "b" was needed in some environments, so it was added for release
1429        5.0 to both the input and output. (It makes no difference on Unix-like
1430        systems.) Later I was told that it is wrong for the input on Windows. I've
1431        now abstracted the modes into two macros, to make it easier to fiddle with
1432        them, and removed "b" from the input mode under Windows.
1434    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1436    12. Added -help and --help to pcretest as an official way of being reminded
1437        of the options.
1439    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1440        and pcrecpp.cc because they annoy compilers at high warning levels.
1442    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1444    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1445        scripts are not C programs :-) and which was not noticed because it works
1446        on Linux.
1448    16. pcretest is supposed to handle any length of pattern and data line (as one
1449        line or as a continued sequence of lines) by extending its input buffer if
1450        necessary. This feature was broken for very long pattern lines, leading to
1451        a string of junk being passed to pcre_compile() if the pattern was longer
1452        than about 50K.
1454    17. I have done a major re-factoring of the way pcre_compile() computes the
1455        amount of memory needed for a compiled pattern. Previously, there was code
1456        that made a preliminary scan of the pattern in order to do this. That was
1457        OK when PCRE was new, but as the facilities have expanded, it has become
1458        harder and harder to keep it in step with the real compile phase, and there
1459        have been a number of bugs (see for example, 4 above). I have now found a
1460        cunning way of running the real compile function in a "fake" mode that
1461        enables it to compute how much memory it would need, while actually only
1462        ever using a few hundred bytes of working memory and without too many
1463        tests of the mode. This should make future maintenance and development
1464        easier. A side effect of this work is that the limit of 200 on the nesting
1465        depth of parentheses has been removed (though this was never a serious
1466        limitation, I suspect). However, there is a downside: pcre_compile() now
1467        runs more slowly than before (30% or more, depending on the pattern). I
1468        hope this isn't a big issue. There is no effect on runtime performance.
1470    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1471        newline (only possible for the last line of a file) and it was a
1472        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1474    19. Added additional timing features to pcretest. (1) The -tm option now times
1475        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1476        separate command line item, by a number that specifies the number of
1477        repeats to use when timing. The default is 50000; this gives better
1478        precision, but takes uncomfortably long for very large patterns.
1480    20. Extended pcre_study() to be more clever in cases where a branch of a
1481        subpattern has no definite first character. For example, (a*|b*)[cd] would
1482        previously give no result from pcre_study(). Now it recognizes that the
1483        first character must be a, b, c, or d.
1485    21. There was an incorrect error "recursive call could loop indefinitely" if
1486        a subpattern (or the entire pattern) that was being tested for matching an
1487        empty string contained only one non-empty item after a nested subpattern.
1488        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1489        incorrectly, because the \d was being skipped in the check.
1491    22. The pcretest program now has a new pattern option /B and a command line
1492        option -b, which is equivalent to adding /B to every pattern. This causes
1493        it to show the compiled bytecode, without the additional information that
1494        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1495        is the same as /B/I).
1497    23. A new optimization is now able automatically to treat some sequences such
1498        as a*b as a*+b. More specifically, if something simple (such as a character
1499        or a simple class like \d) has an unlimited quantifier, and is followed by
1500        something that cannot possibly match the quantified thing, the quantifier
1501        is automatically "possessified".
1503    24. A recursive reference to a subpattern whose number was greater than 39
1504        went wrong under certain circumstances in UTF-8 mode. This bug could also
1505        have affected the operation of pcre_study().
1507    25. Realized that a little bit of performance could be had by replacing
1508        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1510    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1512    27. Possessive quantifiers such as a++ were previously implemented by turning
1513        them into atomic groups such as ($>a+). Now they have their own opcodes,
1514        which improves performance. This includes the automatically created ones
1515        from 23 above.
1517    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1518        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1519        the first matched character to be a colon. This applied both to named and
1520        numbered groups.
1522    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1524    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1525        building PCRE on Windows easier, so I have included it in the distribution.
1527    31. There is now a check in pcretest against a ridiculously large number being
1528        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1529        loop, the loop is abandoned.
1531    32. Forward references to subpatterns in conditions such as (?(2)...) where
1532        subpattern 2 is defined later cause pcre_compile() to search forwards in
1533        the pattern for the relevant set of parentheses. This search went wrong
1534        when there were unescaped parentheses in a character class, parentheses
1535        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1537    33. "Subroutine" calls and backreferences were previously restricted to
1538        referencing subpatterns earlier in the regex. This restriction has now
1539        been removed.
1541    34. Added a number of extra features that are going to be in Perl 5.10. On the
1542        whole, these are just syntactic alternatives for features that PCRE had
1543        previously implemented using the Python syntax or my own invention. The
1544        other formats are all retained for compatibility.
1546        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1547            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1548            also .NET compatible.
1550        (b) A recursion or subroutine call to a named group can now be defined as
1551            (?&name) as well as (?P>name).
1553        (c) A backreference to a named group can now be defined as \k<name> or
1554            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1555            5.10, are also .NET compatible.
1557        (d) A conditional reference to a named group can now use the syntax
1558            (?(<name>) or (?('name') as well as (?(name).
1560        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1561            groups (named and numbered) that are never evaluated inline, but can be
1562            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1563            is always false. There may be only one alternative in such a group.
1565        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1566            as the simple (?(R). The condition is true only if the most recent
1567            recursion is that of the given number or name. It does not search out
1568            through the entire recursion stack.
1570        (g) The escape \gN or \g{N} has been added, where N is a positive or
1571            negative number, specifying an absolute or relative reference.
1573    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1574        some "unreachable code" warnings.
1576    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1577        things, this adds five new scripts.
1579    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1580        There were also incompatibilities regarding the handling of \Q..\E inside
1581        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1582        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1584    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1585        matches an empty string, and forcibly breaks the loop. There were bugs in
1586        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1587        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1588        separate and independent bugs (that affected different cases) have been
1589        fixed.
1591    39. Refactored the code to abolish the use of different opcodes for small
1592        capturing bracket numbers. This is a tidy that I avoided doing when I
1593        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1594        The new approach is not only tidier, it makes it possible to reduce the
1595        memory needed to fix the previous bug (38).
1597    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1598        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1599        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1600        mode.
1602    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1603        report.
1605    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1606        copy construction and assignment in the C++ wrapper.
1608    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1609        couple of bugs that could have given wrong results in the "--newline=crlf"
1610        case.
1612    44. Added a number of casts and did some reorganization of signed/unsigned int
1613        variables following suggestions from Dair Grant. Also renamed the variable
1614        "this" as "item" because it is a C++ keyword.
1616    45. Arranged for dftables to add
1618          #include "pcre_internal.h"
1620        to pcre_chartables.c because without it, gcc 4.x may remove the array
1621        definition from the final binary if PCRE is built into a static library and
1622        dead code stripping is activated.
1624    46. For an unanchored pattern, if a match attempt fails at the start of a
1625        newline sequence, and the newline setting is CRLF or ANY, and the next two
1626        characters are CRLF, advance by two characters instead of one.
1629    Version 6.7 04-Jul-06
1630    ---------------------
1632     1. In order to handle tests when input lines are enormously long, pcretest has
1633        been re-factored so that it automatically extends its buffers when
1634        necessary. The code is crude, but this _is_ just a test program. The
1635        default size has been increased from 32K to 50K.
1637     2. The code in pcre_study() was using the value of the re argument before
1638        testing it for NULL. (Of course, in any sensible call of the function, it
1639        won't be NULL.)
1641     3. The memmove() emulation function in pcre_internal.h, which is used on
1642        systems that lack both memmove() and bcopy() - that is, hardly ever -
1643        was missing a "static" storage class specifier.
1645     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1646        containing an extended class (one that cannot be represented by a bitmap
1647        because it contains high-valued characters or Unicode property items, e.g.
1648        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1649        pattern, but PCRE should not loop if you do not (it no longer does).
1650        [Detail: two cases were found: (a) a repeated subpattern containing an
1651        extended class; (b) a recursive reference to a subpattern that followed a
1652        previous extended class. It wasn't skipping over the extended class
1653        correctly when UTF-8 mode was not set.]
1655     5. A negated single-character class was not being recognized as fixed-length
1656        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1657        compile error "lookbehind assertion is not fixed length".
1659     6. The RunPerlTest auxiliary script was showing an unexpected difference
1660        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1661        write a Perl script that can interpret lines of an input file either as
1662        byte characters or as UTF-8, which is what "perltest" was being required to
1663        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1664        can't do is switch easily at run time between having the "use utf8;" pragma
1665        or not. In the end, I fudged it by using the RunPerlTest script to insert
1666        "use utf8;" explicitly for the UTF-8 tests.
1668     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1669        the end of the subject string, contrary to the documentation and to what
1670        Perl does. This was true of both matching functions. Now it matches only at
1671        the start of the subject and immediately after *internal* newlines.
1673     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1674        a pointer to an int instead of a pointer to an unsigned long int. This
1675        caused problems on 64-bit systems.
1677     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1678        instance of the 'standard' template library not being so standard".
1680    10. There was no check on the number of named subpatterns nor the maximum
1681        length of a subpattern name. The product of these values is used to compute
1682        the size of the memory block for a compiled pattern. By supplying a very
1683        long subpattern name and a large number of named subpatterns, the size
1684        computation could be caused to overflow. This is now prevented by limiting
1685        the length of names to 32 characters, and the number of named subpatterns
1686        to 10,000.
1688    11. Subpatterns that are repeated with specific counts have to be replicated in
1689        the compiled pattern. The size of memory for this was computed from the
1690        length of the subpattern and the repeat count. The latter is limited to
1691        65535, but there was no limit on the former, meaning that integer overflow
1692        could in principle occur. The compiled length of a repeated subpattern is
1693        now limited to 30,000 bytes in order to prevent this.
1695    12. Added the optional facility to have named substrings with the same name.
1697    13. Added the ability to use a named substring as a condition, using the
1698        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1699        are numbers (not recommended). Forward references are permitted.
1701    14. Added forward references in named backreferences (if you see what I mean).
1703    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1704        pattern could run off the end of the subject. For example, the pattern
1705        "(?s)(.{1,5})"8 did this with the subject "ab".
1707    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1708        PCRE_CASELESS was set when matching characters that were quantified with ?
1709        or *.
1711    17. A character class other than a single negated character that had a minimum
1712        but no maximum quantifier - for example [ab]{6,} - was not handled
1713        correctly by pce_dfa_exec(). It would match only one character.
1715    18. A valid (though odd) pattern that looked like a POSIX character
1716        class but used an invalid character after [ (for example [[,abc,]]) caused
1717        pcre_compile() to give the error "Failed: internal error: code overflow" or
1718        in some cases to crash with a glibc free() error. This could even happen if
1719        the pattern terminated after [[ but there just happened to be a sequence of
1720        letters, a binary zero, and a closing ] in the memory that followed.
1722    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1723        over the years. Originally (before any Unicode support), just the bottom 8
1724        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1725        output from "man perlunicode" includes this:
1727          The regular expression compiler produces polymorphic opcodes.  That
1728          is, the pattern adapts to the data and automatically switches to
1729          the Unicode character scheme when presented with Unicode data--or
1730          instead uses a traditional byte scheme when presented with byte
1731          data.
1733        Sadly, a wide octal escape does not cause a switch, and in a string with
1734        no other multibyte characters, these octal escapes are treated as before.
1735        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1736        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1737        Unicode string.
1739        I have not perpetrated such confusion in PCRE. Up till now, it took just
1740        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1741        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1742        translate to the appropriate multibyte character.
1744    29. Applied some refactoring to reduce the number of warnings from Microsoft
1745        and Borland compilers. This has included removing the fudge introduced
1746        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1747        a warning about an unused variable.
1749    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1750        characters since release 4.0, because Perl (from release 5.004) does not.
1751        [Or at least, is documented not to: some releases seem to be in conflict
1752        with the documentation.] However, when a pattern was studied with
1753        pcre_study() and all its branches started with \s, PCRE still included VT
1754        as a possible starting character. Of course, this did no harm; it just
1755        caused an unnecessary match attempt.
1757    22. Removed a now-redundant internal flag bit that recorded the fact that case
1758        dependency changed within the pattern. This was once needed for "required
1759        byte" processing, but is no longer used. This recovers a now-scarce options
1760        bit. Also moved the least significant internal flag bit to the most-
1761        significant bit of the word, which was not previously used (hangover from
1762        the days when it was an int rather than a uint) to free up another bit for
1763        the future.
1765    23. Added support for CRLF line endings as well as CR and LF. As well as the
1766        default being selectable at build time, it can now be changed at runtime
1767        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1768        specify that it is scanning data with non-default line endings.
1770    24. Changed the definition of CXXLINK to make it agree with the definition of
1771        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1773    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1774        recursions. This makes a big different to stack usage for some patterns.
1776    26. If a subpattern containing a named recursion or subroutine reference such
1777        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1778        the space required for the compiled pattern went wrong and gave too small a
1779        value. Depending on the environment, this could lead to "Failed: internal
1780        error: code overflow at offset 49" or "glibc detected double free or
1781        corruption" errors.
1783    27. Applied patches from Google (a) to support the new newline modes and (b) to
1784        advance over multibyte UTF-8 characters in GlobalReplace.
1786    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1787        difference for some implementation of PCRE in some Windows version.
1789    29. Added some extra testing facilities to pcretest:
1791        \q<number>   in a data line sets the "match limit" value
1792        \Q<number>   in a data line sets the "match recursion limt" value
1793        -S <number>  sets the stack size, where <number> is in megabytes
1795        The -S option isn't available for Windows.
1798    Version 6.6 06-Feb-06
1799    ---------------------
1801     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1802        in pcreposix.h. I have copied the definition from pcre.h.
1804     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1805        because pcre.h is no longer a built file.
1807     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1808        not normally included in the compiled code.
1811    Version 6.5 01-Feb-06
1812    ---------------------
1814     1. When using the partial match feature with pcre_dfa_exec(), it was not
1815        anchoring the second and subsequent partial matches at the new starting
1816        point. This could lead to incorrect results. For example, with the pattern
1817        /1234/, partially matching against "123" and then "a4" gave a match.
1819     2. Changes to pcregrep:
1821        (a) All non-match returns from pcre_exec() were being treated as failures
1822            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1823            error message is output. Some extra information is given for the
1824            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1825            probably the only errors that are likely to be caused by users (by
1826            specifying a regex that has nested indefinite repeats, for instance).
1827            If there are more than 20 of these errors, pcregrep is abandoned.
1829        (b) A binary zero was treated as data while matching, but terminated the
1830            output line if it was written out. This has been fixed: binary zeroes
1831            are now no different to any other data bytes.
1833        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1834            used to set a locale for matching. The --locale=xxxx long option has
1835            been added (no short equivalent) to specify a locale explicitly on the
1836            pcregrep command, overriding the environment variables.
1838        (d) When -B was used with -n, some line numbers in the output were one less
1839            than they should have been.
1841        (e) Added the -o (--only-matching) option.
1843        (f) If -A or -C was used with -c (count only), some lines of context were
1844            accidentally printed for the final match.
1846        (g) Added the -H (--with-filename) option.
1848        (h) The combination of options -rh failed to suppress file names for files
1849            that were found from directory arguments.
1851        (i) Added the -D (--devices) and -d (--directories) options.
1853        (j) Added the -F (--fixed-strings) option.
1855        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1857        (l) Added the --colo(u)r option.
1859        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1860            is not present by default.
1862     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1863        items such as (?R) or (?1), when the recursion could match a number of
1864        alternatives. If it matched one of the alternatives, but subsequently,
1865        outside the recursion, there was a failure, the code tried to back up into
1866        the recursion. However, because of the way PCRE is implemented, this is not
1867        possible, and the result was an incorrect result from the match.
1869        In order to prevent this happening, the specification of recursion has
1870        been changed so that all such subpatterns are automatically treated as
1871        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1873     4. I had overlooked the fact that, in some locales, there are characters for
1874        which isalpha() is true but neither isupper() nor islower() are true. In
1875        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1876        and ordfeminine) are like this. This affected the treatment of \w and \W
1877        when they appeared in character classes, but not when they appeared outside
1878        a character class. The bit map for "word" characters is now created
1879        separately from the results of isalnum() instead of just taking it from the
1880        upper, lower, and digit maps. (Plus the underscore character, of course.)
1882     5. The above bug also affected the handling of POSIX character classes such as
1883        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1884        permanent tables. Instead, the bit maps for such a class were previously
1885        created as the appropriate unions of the upper, lower, and digit bitmaps.
1886        Now they are created by subtraction from the [[:word:]] class, which has
1887        its own bitmap.
1889     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1890        It is created by subtracting the vertical space characters (\x09, \x0a,
1891        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1892        subtraction was done in the overall bitmap for a character class, meaning
1893        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1894        be recognized. This bug has been fixed.
1896     7. Patches from the folks at Google:
1898          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1899          real life, but is still worth protecting against".
1901          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1902          regular expressions".
1904          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1905          have it.
1907          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1908          "configure" and the latter not, in order to fix a problem somebody had
1909          with compiling the Arg class on HP-UX.
1911          (e) Improve the error-handling of the C++ wrapper a little bit.
1913          (f) New tests for checking recursion limiting.
1915     8. The pcre_memmove() function, which is used only if the environment does not
1916        have a standard memmove() function (and is therefore rarely compiled),
1917        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1918        returning a result (though PCRE never actually uses the result).
1920     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1921        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1922        returned instead of calling malloc() with an overflowing number that would
1923        most likely cause subsequent chaos.
1925    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1927    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1928        with this option is matched, the nmatch and pmatch options of regexec() are
1929        ignored.
1931    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1932        provided in case anyone wants to the the POSIX interface with UTF-8
1933        strings.
1935    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1936        C++ linking (needed for some HP-UX environments).
1938    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1939        (unused parameter) and in the pcre_printint() function (omitted "default"
1940        switch label when the default is to do nothing).
1942    15. Added some code to make it possible, when PCRE is compiled as a C++
1943        library, to replace subject pointers for pcre_exec() with a smart pointer
1944        class, thus making it possible to process discontinuous strings.
1946    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1947        much the same function. They were added by different people who were trying
1948        to make PCRE easy to compile on non-Unix systems. It has been suggested
1949        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1950        for compiling on Windows systems. I have therefore replaced it with
1951        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1952        defaults to "extern" for C or "extern C" for C++, which works fine on
1953        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1954        SCOPE with something explicit in config.h. In addition:
1956        (a) pcreposix.h still had just "extern" instead of either of these macros;
1957            I have replaced it with PCRE_DATA_SCOPE.
1959        (b) Functions such as _pcre_xclass(), which are internal to the library,
1960            but external in the C sense, all had PCRE_EXPORT in their definitions.
1961            This is apparently wrong for the Windows case, so I have removed it.
1962            (It makes no difference on Unix-like systems.)
1964    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1965        of recursive calls to match(). This is different to MATCH_LIMIT because
1966        that limits the total number of calls to match(), not all of which increase
1967        the depth of recursion. Limiting the recursion depth limits the amount of
1968        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1969        when PCRE is compiled, and changed at run time. A patch from Google adds
1970        this functionality to the C++ interface.
1972    18. Changes to the handling of Unicode character properties:
1974        (a) Updated the table to Unicode 4.1.0.
1976        (b) Recognize characters that are not in the table as "Cn" (undefined).
1978        (c) I revised the way the table is implemented to a much improved format
1979            which includes recognition of ranges. It now supports the ranges that
1980            are defined in UnicodeData.txt, and it also amalgamates other
1981            characters into ranges. This has reduced the number of entries in the
1982            table from around 16,000 to around 3,000, thus reducing its size
1983            considerably. I realized I did not need to use a tree structure after
1984            all - a binary chop search is just as efficient. Having reduced the
1985            number of entries, I extended their size from 6 bytes to 8 bytes to
1986            allow for more data.
1988        (d) Added support for Unicode script names via properties such as \p{Han}.
1990    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1991        matching that character.
1993    20. When matching a repeated Unicode property with a minimum greater than zero,
1994        (for example \pL{2,}), PCRE could look past the end of the subject if it
1995        reached it while seeking the minimum number of characters. This could
1996        happen only if some of the characters were more than one byte long, because
1997        there is a check for at least the minimum number of bytes.
1999    21. Refactored the implementation of \p and \P so as to be more general, to
2000        allow for more different types of property in future. This has changed the
2001        compiled form incompatibly. Anybody with saved compiled patterns that use
2002        \p or \P will have to recompile them.
2004    22. Added "Any" and "L&" to the supported property types.
2006    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2007        but give a compile time error if the value is greater than 0xff.
2009    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2010        accidentally not being installed or uninstalled.
2012    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2013        made were to insert the current release number. This seemed silly, because
2014        it made things harder for people building PCRE on systems that don't run
2015        "configure". I have turned pcre.h into a distributed file, no longer built
2016        by "configure", with the version identification directly included. There is
2017        no longer a pcre.h.in file.
2019        However, this change necessitated a change to the pcre-config script as
2020        well. It is built from pcre-config.in, and one of the substitutions was the
2021        release number. I have updated configure.ac so that ./configure now finds
2022        the release number by grepping pcre.h.
2024    26. Added the ability to run the tests under valgrind.
2027    Version 6.4 05-Sep-05
2028    ---------------------
2030     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2031        "--" to be printed when multiple files were scanned, even when none of the
2032        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2033        consider it to be a bug, and have restored the previous behaviour.
2035     2. A couple of code tidies to get rid of compiler warnings.
2037     3. The pcretest program used to cheat by referring to symbols in the library
2038        whose names begin with _pcre_. These are internal symbols that are not
2039        really supposed to be visible externally, and in some environments it is
2040        possible to suppress them. The cheating is now confined to including
2041        certain files from the library's source, which is a bit cleaner.
2043     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2044        file's purpose clearer.
2046     5. Reorganized pcre_ucp_findchar().
2049    Version 6.3 15-Aug-05
2050    ---------------------
2052     1. The file libpcre.pc.in did not have general read permission in the tarball.
2054     2. There were some problems when building without C++ support:
2056        (a) If C++ support was not built, "make install" and "make test" still
2057            tried to test it.
2059        (b) There were problems when the value of CXX was explicitly set. Some
2060            changes have been made to try to fix these, and ...
2062        (c) --disable-cpp can now be used to explicitly disable C++ support.
2064        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2065            backslash in a target when C++ was disabled. This confuses some
2066            versions of "make", apparently. Using an intermediate variable solves
2067            this. (Same for CPP_LOBJ.)
2069     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2070        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2071        necessary on certain architectures.
2073     4. Added a setting of -export-symbols-regex to the link command to remove
2074        those symbols that are exported in the C sense, but actually are local
2075        within the library, and not documented. Their names all begin with
2076        "_pcre_". This is not a perfect job, because (a) we have to except some
2077        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2078        available (and never for static libraries). I have made a note to try to
2079        find a way round (a) in the future.
2082    Version 6.2 01-Aug-05
2083    ---------------------
2085     1. There was no test for integer overflow of quantifier values. A construction
2086        such as {1111111111111111} would give undefined results. What is worse, if
2087        a minimum quantifier for a parenthesized subpattern overflowed and became
2088        negative, the calculation of the memory size went wrong. This could have
2089        led to memory overwriting.
2091     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2093     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2094        operating environments where this matters.
2096     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2097        PCRE options from within the C++ wrapper.
2099     5. Named capturing subpatterns were not being correctly counted when a pattern
2100        was compiled. This caused two problems: (a) If there were more than 100
2101        such subpatterns, the calculation of the memory needed for the whole
2102        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2103        back references of the form \12, where the number was greater than 9, were
2104        not recognized as back references, even though there were sufficient
2105        previous subpatterns.
2107     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2108        versions of gcc, e.g. 2.95.4.
2111    Version 6.1 21-Jun-05
2112    ---------------------
2114     1. There was one reference to the variable "posix" in pcretest.c that was not
2115        surrounded by "#if !defined NOPOSIX".
2117     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2118        the cross-check on the old pcre_info() function, for the benefit of the
2119        cut-down version of PCRE that is currently imported into Exim.
2121     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2122        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2123        patterns that start (?i) but I don't think that matters. The use of (?i) is
2124        just an example; this all applies to the other options as well.
2126     4. Since libtool seems to echo the compile commands it is issuing, the output
2127        from "make" can be reduced a bit by putting "@" in front of each libtool
2128        compile command.
2130     5. Patch from the folks at Google for configure.in to be a bit more thorough
2131        in checking for a suitable C++ installation before trying to compile the
2132        C++ stuff. This should fix a reported problem when a compiler was present,
2133        but no suitable headers.
2135     6. The man pages all had just "PCRE" as their title. I have changed them to
2136        be the relevant file name. I have also arranged that these names are
2137        retained in the file doc/pcre.txt, which is a concatenation in text format
2138        of all the man pages except the little individual ones for each function.
2140     7. The NON-UNIX-USE file had not been updated for the different set of source
2141        files that come with release 6. I also added a few comments about the C++
2142        wrapper.
2145    Version 6.0 07-Jun-05
2146    ---------------------
2148     1. Some minor internal re-organization to help with my DFA experiments.
2150     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2151        didn't matter for the library itself when fully configured, but did matter
2152        when compiling without UCP support, or within Exim, where the ucp files are
2153        not imported.
2155     3. Refactoring of the library code to split up the various functions into
2156        different source modules. The addition of the new DFA matching code (see
2157        below) to a single monolithic source would have made it really too
2158        unwieldy, quite apart from causing all the code to be include in a
2159        statically linked application, when only some functions are used. This is
2160        relevant even without the DFA addition now that patterns can be compiled in
2161        one application and matched in another.
2163        The downside of splitting up is that there have to be some external
2164        functions and data tables that are used internally in different modules of
2165        the library but which are not part of the API. These have all had their
2166        names changed to start with "_pcre_" so that they are unlikely to clash
2167        with other external names.
2169     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2170        a different (DFA) algorithm. Although it is slower than the original
2171        function, it does have some advantages for certain types of matching
2172        problem.
2174     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2175        including restarting after a partial match.
2177     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2178        defined when compiling for Windows was sent to me. I have put it into the
2179        code, though I have no means of testing or verifying it.
2181     7. Added the pcre_refcount() auxiliary function.
2183     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2184        match before or at the first newline in the subject string. In pcretest,
2185        the /f option on a pattern can be used to set this.
2187     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2188        would behave wrongly. This has been present in PCRE since release 4.0.
2190    10. A number of changes to the pcregrep command:
2192        (a) Refactored how -x works; insert ^(...)$ instead of setting
2193            PCRE_ANCHORED and checking the length, in preparation for adding
2194            something similar for -w.
2196        (b) Added the -w (match as a word) option.
2198        (c) Refactored the way lines are read and buffered so as to have more
2199            than one at a time available.
2201        (d) Implemented a pcregrep test script.
2203        (e) Added the -M (multiline match) option. This allows patterns to match
2204            over several lines of the subject. The buffering ensures that at least
2205            8K, or the rest of the document (whichever is the shorter) is available
2206            for matching (and similarly the previous 8K for lookbehind assertions).
2208        (f) Changed the --help output so that it now says
2210              -w, --word-regex(p)
2212            instead of two lines, one with "regex" and the other with "regexp"
2213            because that confused at least one person since the short forms are the
2214            same. (This required a bit of code, as the output is generated
2215            automatically from a table. It wasn't just a text change.)
2217        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2218            option but starts with a hyphen. Could be a pattern or a path name
2219            starting with a hyphen, for instance.
2221        (h) "-" can be given as a file name to represent stdin.
2223        (i) When file names are being printed, "(standard input)" is used for
2224            the standard input, for compatibility with GNU grep. Previously
2225            "<stdin>" was used.
2227        (j) The option --label=xxx can be used to supply a name to be used for
2228            stdin when file names are being printed. There is no short form.
2230        (k) Re-factored the options decoding logic because we are going to add
2231            two more options that take data. Such options can now be given in four
2232            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2234        (l) Added the -A, -B, and -C options for requesting that lines of context
2235            around matches be printed.
2237        (m) Added the -L option to print the names of files that do not contain
2238            any matching lines, that is, the complement of -l.
2240        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2241            continue to scan other files.
2243        (o) The -s option was incorrectly implemented. For compatibility with other
2244            greps, it now suppresses the error message for a non-existent or non-
2245            accessible file (but not the return code). There is a new option called
2246            -q that suppresses the output of matching lines, which was what -s was
2247            previously doing.
2249        (p) Added --include and --exclude options to specify files for inclusion
2250            and exclusion when recursing.
2252    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2253        Hopefully, it now does.
2255    12. Missing cast in pcre_study().
2257    13. Added an "uninstall" target to the makefile.
2259    14. Replaced "extern" in the function prototypes in Makefile.in with
2260        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2261        world, but is set differently for Windows.
2263    15. Added a second compiling function called pcre_compile2(). The only
2264        difference is that it has an extra argument, which is a pointer to an
2265        integer error code. When there is a compile-time failure, this is set
2266        non-zero, in addition to the error test pointer being set to point to an
2267        error message. The new argument may be NULL if no error number is required
2268        (but then you may as well call pcre_compile(), which is now just a
2269        wrapper). This facility is provided because some applications need a
2270        numeric error indication, but it has also enabled me to tidy up the way
2271        compile-time errors are handled in the POSIX wrapper.
2273    16. Added VPATH=.libs to the makefile; this should help when building with one
2274        prefix path and installing with another. (Or so I'm told by someone who
2275        knows more about this stuff than I do.)
2277    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2278        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2279        match everything, including newlines. This is not POSIX-compatible, but
2280        somebody wanted the feature. From pcretest it can be activated by using
2281        both the P and the s flags.
2283    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2285    19. libpcre.pc was being incorrectly installed as executable.
2287    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2288        it now also looks for '\r' so that it will work unmodified on Windows.
2290    21. Added Google's contributed C++ wrapper to the distribution.
2292    22. Added some untidy missing memory free() calls in pcretest, to keep
2293        Electric Fence happy when testing.
2297    Version 5.0 13-Sep-04
2298    ---------------------
2300     1. Internal change: literal characters are no longer packed up into items
2301        containing multiple characters in a single byte-string. Each character
2302        is now matched using a separate opcode. However, there may be more than one
2303        byte in the character in UTF-8 mode.
2305     2. The pcre_callout_block structure has two new fields: pattern_position and
2306        next_item_length. These contain the offset in the pattern to the next match
2307        item, and its length, respectively.
2309     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2310        insertion of callouts before each pattern item. Added the /C option to
2311        pcretest to make use of this.
2313     4. On the advice of a Windows user, the lines
2315          #if defined(_WIN32) || defined(WIN32)
2316          _setmode( _fileno( stdout ), 0x8000 );
2317          #endif  /* defined(_WIN32) || defined(WIN32) */
2319        have been added to the source of pcretest. This apparently does useful
2320        magic in relation to line terminators.
2322     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2323        for the benefit of those environments where the "b" makes a difference.
2325     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2326        to know about it. I have put a hack into configure.in that adds in code
2327        to set GCC=yes if CC=icc. This seems to end up at a point in the
2328        generated configure script that is early enough to affect the setting of
2329        compiler options, which is what is needed, but I have no means of testing
2330        whether it really works. (The user who reported this had patched the
2331        generated configure script, which of course I cannot do.)
2333        LATER: After change 22 below (new libtool files), the configure script
2334        seems to know about icc (and also ecc). Therefore, I have commented out
2335        this hack in configure.in.
2337     7. Added support for pkg-config (2 patches were sent in).
2339     8. Negated POSIX character classes that used a combination of internal tables
2340        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2341        [[:^ascii]]. Typically, they would match almost any characters. The other
2342        POSIX classes were not broken in this way.
2344     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2345        to find the match, as PCRE was deluded into thinking that the match had to
2346        start at the start point or following a newline. The same bug applied to
2347        patterns with negative forward assertions or any backward assertions
2348        preceding ".*" at the start, unless the pattern required a fixed first
2349        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2351    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2352        starting at the last subject character, bytes beyond the end of the subject
2353        string were read.
2355    11. Renamed the variable "class" as "classbits" to make life easier for C++
2356        users. (Previously there was a macro definition, but it apparently wasn't
2357        enough.)
2359    12. Added the new field "tables" to the extra data so that tables can be passed
2360        in at exec time, or the internal tables can be re-selected. This allows
2361        a compiled regex to be saved and re-used at a later time by a different
2362        program that might have everything at different addresses.
2364    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2365        -R library as well as a -L library.
2367    14. The debugging options of pcretest (-d on the command line or D on a
2368        pattern) showed incorrect output for anything following an extended class
2369        that contained multibyte characters and which was followed by a quantifier.
2371    15. Added optional support for general category Unicode character properties
2372        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2373        support. It adds about 90K to the size of the library. The meanings of the
2374        inbuilt class escapes such as \d and \s have NOT been changed.
2376    16. Updated pcredemo.c to include calls to free() to release the memory for the
2377        compiled pattern.
2379    17. The generated file chartables.c was being created in the source directory
2380        instead of in the building directory. This caused the build to fail if the
2381        source directory was different from the building directory, and was
2382        read-only.
2384    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2385        file. No doubt somebody will tell me if they don't make sense... Also added
2386        Dan Mooney's comments about building on OpenVMS.
2388    19. Added support for partial matching via the PCRE_PARTIAL option for
2389        pcre_exec() and the \P data escape in pcretest.
2391    20. Extended pcretest with 3 new pattern features:
2393        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2394              write the compiled pattern to the file whose name is "rest-of-line".
2395              This is a straight binary dump of the data, with the saved pointer to
2396              the character tables forced to be NULL. The study data, if any, is
2397              written too. After writing, pcretest reads a new pattern.
2399        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2400              compiled pattern from the given file. There must not be any
2401              occurrences of "<" in the file name (pretty unlikely); if there are,
2402              pcretest will instead treat the initial "<" as a pattern delimiter.
2403              After reading in the pattern, pcretest goes on to read data lines as
2404              usual.
2406        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2407              and 16-bit fields in a compiled pattern, to simulate a pattern that
2408              was compiled on a host of opposite endianness.
2410    21. The pcre-exec() function can now cope with patterns that were compiled on
2411        hosts of opposite endianness, with this restriction:
2413          As for any compiled expression that is saved and used later, the tables
2414          pointer field cannot be preserved; the extra_data field in the arguments
2415          to pcre_exec() should be used to pass in a tables address if a value
2416          other than the default internal tables were used at compile time.
2418    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2419        now diagnosed as an error. Previously, most of the time, a negative number
2420        would have been treated as zero, but if in addition "ovector" was passed as
2421        NULL, a crash could occur.
2423    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2424        new versions from the libtool 1.5 distribution (the last one is a copy of
2425        a file called libtool.m4). This seems to have fixed the need to patch
2426        "configure" to support Darwin 1.3 (which I used to do). However, I still
2427        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2428        workstation).
2430    24. Changed the PCRE licence to be the more standard "BSD" licence.
2433    Version 4.5 01-Dec-03
2434    ---------------------
2436     1. There has been some re-arrangement of the code for the match() function so
2437        that it can be compiled in a version that does not call itself recursively.
2438        Instead, it keeps those local variables that need separate instances for
2439        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2440        needs to "recurse". Keeping track of where control must go is done by means
2441        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2442        hide most of the details from the main code, and operates only if
2443        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2444        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2445        operating.
2447        To make it easier for callers to provide specially tailored get/free
2448        functions for this usage, two new functions, pcre_stack_malloc, and
2449        pcre_stack_free, are used. They are always called in strict stacking order,
2450        and the size of block requested is always the same.
2452        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2453        PCRE has been compiled to use the stack or the heap for recursion. The
2454        -C option of pcretest uses this to show which version is compiled.
2456        A new data escape \S, is added to pcretest; it causes the amounts of store
2457        obtained and freed by both kinds of malloc/free at match time to be added
2458        to the output.
2460     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2461        what's available on my current Linux desktop machine.
2463     3. When matching a UTF-8 string, the test for a valid string at the start has
2464        been extended. If start_offset is not zero, PCRE now checks that it points
2465        to a byte that is the start of a UTF-8 character. If not, it returns
2466        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2467        this is necessary because there may be backward assertions in the pattern.
2468        When matching the same subject several times, it may save resources to use
2469        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2471     4. The code for checking the validity of UTF-8 strings has been tightened so
2472        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2473        containing "overlong sequences".
2475     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2476        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2477        should have been "&", but it just so happened that all the cases this let
2478        through by mistake were picked up later in the function.
2480     6. I had used a variable called "isblank" - this is a C99 function, causing
2481        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2483     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2484        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2485        so the version is shown; (c) stop "make test" echoing "./RunTest".
2487     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2489     9. The prototype for memmove() for systems that don't have it was using
2490        size_t, but the inclusion of the header that defines size_t was later. I've
2491        moved the #includes for the C headers earlier to avoid this.
2493    10. Added some adjustments to the code to make it easier to compiler on certain
2494        special systems:
2496          (a) Some "const" qualifiers were missing.
2497          (b) Added the macro EXPORT before all exported functions; by default this
2498              is defined to be empty.
2499          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2500              that it reads its output file name as an argument instead of writing
2501              to the standard output and assuming this can be redirected.
2503    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2504        class containing characters with values greater than 255, PCRE compilation
2505        went into a loop.
2507    12. A recursive reference to a subpattern that was within another subpattern
2508        that had a minimum quantifier of zero caused PCRE to crash. For example,
2509        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2510        recursion. If the recursively-called subpattern itself had a zero repeat,
2511        that was OK.
2513    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2514        buffer into which it was copied (for escape processing) was still set at
2515        1024, so long lines caused crashes.
2517    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2518        "internal error: code overflow...". This applied to any character class
2519        that was followed by a possessive quantifier.
2521    15. Modified the Makefile to add libpcre.la as a prerequisite for
2522        libpcreposix.la because I was told this is needed for a parallel build to
2523        work.
2525    16. If a pattern that contained .* following optional items at the start was
2526        studied, the wrong optimizing data was generated, leading to matching
2527        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2528        matching string must start with a or b or c. The correct conclusion for
2529        this pattern is that a match can start with any character.
2532    Version 4.4 13-Aug-03
2533    ---------------------
2535     1. In UTF-8 mode, a character class containing characters with values between
2536        127 and 255 was not handled correctly if the compiled pattern was studied.
2537        In fixing this, I have also improved the studying algorithm for such
2538        classes (slightly).
2540     2. Three internal functions had redundant arguments passed to them. Removal
2541        might give a very teeny performance improvement.
2543     3. Documentation bug: the value of the capture_top field in a callout is *one
2544        more than* the number of the hightest numbered captured substring.
2546     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2547        in incorrectly linking with a previously installed version. They now link
2548        explicitly with libpcre.la.
2550     5. configure.in no longer needs to recognize Cygwin specially.
2552     6. A problem in pcre.in for Windows platforms is fixed.
2554     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2555        pcretest, it used to include the size of the study block as part of its
2556        output. Unfortunately, the structure contains a field that has a different
2557        size on different hardware architectures. This meant that the tests that
2558        showed this size failed. As the block is currently always of a fixed size,
2559        this information isn't actually particularly useful in pcretest output, so
2560        I have just removed it.
2562     8. Three pre-processor statements accidentally did not start in column 1.
2563        Sadly, there are *still* compilers around that complain, even though
2564        standard C has not required this for well over a decade. Sigh.
2566     9. In pcretest, the code for checking callouts passed small integers in the
2567        callout_data field, which is a void * field. However, some picky compilers
2568        complained about the casts involved for this on 64-bit systems. Now
2569        pcretest passes the address of the small integer instead, which should get
2570        rid of the warnings.
2572    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2573        both compile and run time, and gives an error if an invalid UTF-8 sequence
2574        is found. There is a option for disabling this check in cases where the
2575        string is known to be correct and/or the maximum performance is wanted.
2577    11. In response to a bug report, I changed one line in Makefile.in from
2579            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2580        to
2581            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2583        to look similar to other lines, but I have no way of telling whether this
2584        is the right thing to do, as I do not use Windows. No doubt I'll get told
2585        if it's wrong...
2588    Version 4.3 21-May-03
2589    ---------------------
2591    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2592       Makefile.
2594    2. Some refactoring to improve the quality of the code:
2596       (i)   The utf8_table... variables are now declared "const".
2598       (ii)  The code for \cx, which used the "case flipping" table to upper case
2599             lower case letters, now just substracts 32. This is ASCII-specific,
2600             but the whole concept of \cx is ASCII-specific, so it seems
2601             reasonable.
2603       (iii) PCRE was using its character types table to recognize decimal and
2604             hexadecimal digits in the pattern. This is silly, because it handles
2605             only 0-9, a-f, and A-F, but the character types table is locale-
2606             specific, which means strange things might happen. A private
2607             table is now used for this - though it costs 256 bytes, a table is
2608             much faster than multiple explicit tests. Of course, the standard
2609             character types table is still used for matching digits in subject
2610             strings against \d.
2612       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2613             ending in _t are). So I've renamed it as ESC_tee.
2615    3. The first argument for regexec() in the POSIX wrapper should have been
2616       defined as "const".
2618    4. Changed pcretest to use malloc() for its buffers so that they can be
2619       Electric Fenced for debugging.
2621    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2622       to read one or more bytes before the start of the subject string. Often this
2623       had no effect on PCRE's behaviour, but in some circumstances it could
2624       provoke a segmentation fault.
2626    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2627       to try to read one or more bytes before the start of the subject string.
2629    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2630       UTF-8 support could misbehave in various ways if the subject string
2631       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2632       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2633       back over UTF-8 characters.)
2636    Version 4.2 14-Apr-03
2637    ---------------------
2639    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2641    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2642         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2643         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2644         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2645         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2646           and BUILD_EXEEXT
2647         Note: automatic setting of the BUILD variables is not yet working
2648         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2649           compile-time but not at link-time
2650         [LINK]: use for linking executables only
2651         make different versions for Windows and non-Windows
2652         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2653           libraries
2654         [LINK_FOR_BUILD]: new variable
2655         [OBJEXT]: use throughout
2656         [EXEEXT]: use throughout
2657         <winshared>: new target
2658         <wininstall>: new target
2659         <dftables.o>: use native compiler
2660         <dftables>: use native linker
2661         <install>: handle Windows platform correctly
2662         <clean>: ditto
2663         <check>: ditto
2664         copy DLL to top builddir before testing
2666       As part of these changes, -no-undefined was removed again. This was reported
2667       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2668       in any case.
2670    3. Some tidies to get rid of compiler warnings:
2672       . In the match_data structure, match_limit was an unsigned long int, whereas
2673         match_call_count was an int. I've made them both unsigned long ints.
2675       . In pcretest the fact that a const uschar * doesn't automatically cast to
2676         a void * provoked a warning.
2678       . Turning on some more compiler warnings threw up some "shadow" variables
2679         and a few more missing casts.
2681    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2682       option, a class that contained a single character with a value between 128
2683       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2685    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2686       option, a class that contained several characters, but with at least one
2687       whose value was between 128 and 255 caused PCRE to crash.
2690    Version 4.1 12-Mar-03
2691    ---------------------
2693    1. Compiling with gcc -pedantic found a couple of places where casts were
2694    needed, and a string in dftables.c that was longer than standard compilers are
2695    required to support.
2697    2. Compiling with Sun's compiler found a few more places where the code could
2698    be tidied up in order to avoid warnings.
2700    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2701    first of these names is deprecated in the latest Autoconf in favour of the name
2702    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2703    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2704    analogy I have changed it to CFLAGS_FOR_BUILD.
2706    4. Added -no-undefined to the linking command in the Makefile, because this is
2707    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2708    linking step for the pcreposix library.
2710    5. PCRE was failing to diagnose the case of two named groups with the same
2711    name.
2713    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2714    literal character that is needed in the subject for a match, and scans along to
2715    ensure that it is present before embarking on the full matching process. This
2716    saves time in cases of nested unlimited repeats that are never going to match.
2717    Problem: the scan can take a lot of time if the subject is very long (e.g.
2718    megabytes), thus penalizing straightforward matches. It is now done only if the
2719    amount of subject to be scanned is less than 1000 bytes.
2721    7. A lesser problem with the same optimization is that it was recording the
2722    first character of an anchored pattern as "needed", thus provoking a search
2723    right along the subject, even when the first match of the pattern was going to
2724    fail. The "needed" character is now not set for anchored patterns, unless it
2725    follows something in the pattern that is of non-fixed length. Thus, it still
2726    fulfils its original purpose of finding quick non-matches in cases of nested
2727    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2730    Version 4.0 17-Feb-03
2731    ---------------------
2733    1. If a comment in an extended regex that started immediately after a meta-item
2734    extended to the end of string, PCRE compiled incorrect data. This could lead to
2735    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2737    2. Moved to autoconf 2.53 and libtool 1.4.2.
2739    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2740    the special perltest8 script is no longer needed - all the tests can be run
2741    from a single perltest script.
2743    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2744    by \s. It has now been removed in PCRE. This means it isn't recognized as
2745    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2746    class [:space:] *does* include VT, thereby creating a mess.
2748    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2749    space and tab.
2751    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2752    its new features into the main test script, reducing the number of scripts.
2754    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2755    were backward compatible, and made the (?i) apply to the whole pattern, as if
2756    /i were given. Now it behaves more logically, and applies the option setting
2757    only to what follows. PCRE has been changed to follow suit. However, if it
2758    finds options settings right at the start of the pattern, it extracts them into
2759    the global options, as before. Thus, they show up in the info data.
2761    8. Added support for the \Q...\E escape sequence. Characters in between are
2762    treated as literals. This is slightly different from Perl in that $ and @ are
2763    also handled as literals inside the quotes. In Perl, they will cause variable
2764    interpolation. Note the following examples:
2766        Pattern            PCRE matches      Perl matches
2768        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2769        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2770        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2772    For compatibility with Perl, \Q...\E sequences are recognized inside character
2773    classes as well as outside them.
2775    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2776    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2777    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2778    signed/unsigned warnings.
2780    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2781    option for pcretest, so I've replaced it by a simple function that does just
2782    that job.
2784    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2785    "pcregrep -".
2787    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2788    Java package. This provides some syntactic sugar for simple cases of what my
2789    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2790    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2791    item, you can use this simplified notation. Note that only makes sense with
2792    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2793    greediness, whatever the setting of the PCRE_UNGREEDY option.
2795    13. A change of greediness default within a pattern was not taking effect at
2796    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2797    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2798    was abstracted outside.
2800    14. PCRE now supports the \G assertion. It is true when the current matching
2801    position is at the start point of the match. This differs from \A when the
2802    starting offset is non-zero. Used with the /g option of pcretest (or similar
2803    code), it works in the same way as it does for Perl's /g option. If all
2804    alternatives of a regex begin with \G, the expression is anchored to the start
2805    match position, and the "anchored" flag is set in the compiled expression.
2807    15. Some bugs concerning the handling of certain option changes within patterns
2808    have been fixed. These applied to options other than (?ims). For example,
2809    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2810    been the other way round. Some of this was related to change 7 above.
2812    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2813    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2814    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2815    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2817    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2818    mode. Unlike ".", it always matches newline, whatever the setting of
2819    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2820    assertions. Perl allows it, but it doesn't (in general) work because it can't
2821    calculate the length of the lookbehind. At least, that's the case for Perl
2822    5.8.0 - I've been told they are going to document that it doesn't work in
2823    future.
2825    18. Added an error diagnosis for escapes that PCRE does not support: these are
2826    \L, \l, \N, \P, \p, \U, \u, and \X.
2828    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2829    reading past the end of the pattern in cases such as /[abcd/.
2831    20. PCRE was getting more memory than necessary for patterns with classes that
2832    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2834    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2835    compiling PCRE for use with Virtual Pascal.
2837    22. Small fix to the Makefile to make it work properly if the build is done
2838    outside the source tree.
2840    23. Added a new extension: a condition to go with recursion. If a conditional
2841    subpattern starts with (?(R) the "true" branch is used if recursion has
2842    happened, whereas the "false" branch is used only at the top level.
2844    24. When there was a very long string of literal characters (over 255 bytes
2845    without UTF support, over 250 bytes with UTF support), the computation of how
2846    much memory was required could be incorrect, leading to segfaults or other
2847    strange effects.
2849    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2850    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2851    there was a subsequent back reference to those brackets. This meant that, for
2852    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2853    possible to check for precisely this case. All we can do is abandon the
2854    optimization if .* occurs inside capturing brackets when there are any back
2855    references whatsoever. (See below for a better fix that came later.)
2857    26. The handling of the optimization for finding the first character of a
2858    non-anchored pattern, and for finding a character that is required later in the
2859    match were failing in some cases. This didn't break the matching; it just
2860    failed to optimize when it could. The way this is done has been re-implemented.
2862    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2864    28. Added a new feature that provides some of the functionality that Perl
2865    provides with (?{...}). The facility is termed a "callout". The way it is done
2866    in PCRE is for the caller to provide an optional function, by setting
2867    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2868    global variable. By default it is unset, which disables all calling out. To get
2869    the function called, the regex must include (?C) at appropriate points. This
2870    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2871    This provides a means of identifying different callout points. When PCRE
2872    reaches such a point in the regex, if pcre_callout has been set, the external
2873    function is called. It is provided with data in a structure called
2874    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2875    matching continues; if it returns a non-zero value, the match at the current
2876    point fails. However, backtracking will occur if possible. [This was changed
2877    later and other features added - see item 49 below.]
2879    29. pcretest is upgraded to test the callout functionality. It provides a
2880    callout function that displays information. By default, it shows the start of
2881    the match and the current position in the text. There are some new data escapes
2882    to vary what happens:
2884        \C+         in addition, show current contents of captured substrings
2885        \C-         do not supply a callout function
2886        \C!n        return 1 when callout number n is reached
2887        \C!n!m      return 1 when callout number n is reached for the mth time
2889    30. If pcregrep was called with the -l option and just a single file name, it
2890    output "<stdin>" if a match was found, instead of the file name.
2892    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2893    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2894    pcre_exec(). This saves a malloc/free per call. The default value of
2895    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2896    when configuring.
2898    32. The default maximum size of a compiled pattern is 64K. There have been a
2899    few cases of people hitting this limit. The code now uses macros to handle the
2900    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2901    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2902    configuring. Tests 2 and 5 work only with 2-byte links because they output
2903    debugging information about compiled patterns.
2905    33. Internal code re-arrangements:
2907    (a) Moved the debugging function for printing out a compiled regex into
2908        its own source file (printint.c) and used #include to pull it into
2909        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2910        separate copies.
2912    (b) Defined the list of op-code names for debugging as a macro in
2913        internal.h so that it is next to the definition of the opcodes.
2915    (c) Defined a table of op-code lengths for simpler skipping along compiled
2916        code. This is again a macro in internal.h so that it is next to the
2917        definition of the opcodes.
2919    34. Added support for recursive calls to individual subpatterns, along the
2920    lines of Robin Houston's patch (but implemented somewhat differently).
2922    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2923    allow it to read and process whole directories in Win32. This code was
2924    contributed by Lionel Fourquaux; it has not been tested by me.
2926    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2927    used to name a group. Names consist of alphanumerics and underscores, and must
2928    be unique. Back references use the syntax (?P=name) and recursive calls use
2929    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2930    numbers. The function pcre_fullinfo() can be used after compilation to extract
2931    a name/number map. There are three relevant calls:
2933      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2934      PCRE_INFO_NAMECOUNT            yields the number of entries
2935      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2937    The map is a vector of fixed-size entries. The size of each entry depends on
2938    the length of the longest name used. The first two bytes of each entry are the
2939    group number, most significant byte first. There follows the corresponding
2940    name, zero terminated. The names are in alphabetical order.
2942    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2943    case instead of 255. Making it the same both with and without UTF-8 support
2944    means that the same test output works with both.
2946    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2947    calling malloc() with a zero argument.
2949    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2950    optimization. I've improved things by keeping a bitmap of backreferences with
2951    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2952    fact referenced, the optimization can be applied. It is unlikely that a
2953    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2954    the match to follow \n) will appear inside brackets with a number greater than
2955    31, but if it does, any back reference > 31 suppresses the optimization.
2957    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2958    of disabling numbered capturing parentheses. Any opening parenthesis that is
2959    not followed by ? behaves as if it were followed by ?: but named parentheses
2960    can still be used for capturing (and they will acquire numbers in the usual
2961    way).
2963    41. Redesigned the return codes from the match() function into yes/no/error so
2964    that errors can be passed back from deep inside the nested calls. A malloc
2965    failure while inside a recursive subpattern call now causes the
2966    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2968    42. It is now possible to set a limit on the number of times the match()
2969    function is called in a call to pcre_exec(). This facility makes it possible to
2970    limit the amount of recursion and backtracking, though not in a directly
2971    obvious way, because the match() function is used in a number of different
2972    circumstances. The count starts from zero for each position in the subject
2973    string (for non-anchored patterns). The default limit is, for compatibility, a
2974    large number, namely 10 000 000. You can change this in two ways:
2976    (a) When configuring PCRE before making, you can use --with-match-limit=n
2977        to set a default value for the compiled library.
2979    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2980        a different value is set. See 45 below.
2982    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2984    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2985    of things that can be changed at compile time. The first argument specifies
2986    what is wanted and the second points to where the information is to be placed.
2987    The current list of available information is:
2989      PCRE_CONFIG_UTF8
2991    The output is an integer that is set to one if UTF-8 support is available;
2992    otherwise it is set to zero.
2996    The output is an integer that it set to the value of the code that is used for
2997    newline. It is either LF (10) or CR (13).
3001    The output is an integer that contains the number of bytes used for internal
3002    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3006    The output is an integer that contains the threshold above which the POSIX
3007    interface uses malloc() for output vectors. See item 31 above.
3011    The output is an unsigned integer that contains the default limit of the number
3012    of match() calls in a pcre_exec() execution. See 42 above.
3014    44. pcretest has been upgraded by the addition of the -C option. This causes it
3015    to extract all the available output from the new pcre_config() function, and to
3016    output it. The program then exits immediately.
3018    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3019    order to support additional features. One way would have been to define
3020    pcre_exec2() (for example) with extra arguments, but this would not have been
3021    extensible, and would also have required all calls to the original function to
3022    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3023    is used for passing in "extra" data from pcre_study().
3025    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3026    contains the following fields:
3028      flags         a bitmap indicating which of the following fields are set
3029      study_data    opaque data from pcre_study()
3030      match_limit   a way of specifying a limit on match() calls for a specific
3031                      call to pcre_exec()
3032      callout_data  data for callouts (see 49 below)
3034    The flag bits are also defined in pcre.h, and are
3040    The pcre_study() function now returns one of these new pcre_extra blocks, with
3041    the actual study data pointed to by the study_data field, and the
3042    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3043    before. That is, this change is entirely upwards-compatible and requires no
3044    change to existing code.
3046    If you want to pass in additional data to pcre_exec(), you can either place it
3047    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3048    block.
3050    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3051    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3052    times with different match limits, until it finds the minimum value needed for
3053    pcre_exec() to complete. The value is then output. This can be instructive; for
3054    most simple matches the number is quite small, but for pathological cases it
3055    gets very large very quickly.
3057    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3058    returns the size of the data block pointed to by the study_data field in a
3059    pcre_extra block, that is, the value that was passed as the argument to
3060    pcre_malloc() when PCRE was getting memory in which to place the information
3061    created by pcre_study(). The fourth argument should point to a size_t variable.
3062    pcretest has been extended so that this information is shown after a successful
3063    pcre_study() call when information about the compiled regex is being displayed.
3065    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3066    because what follows is always an absolute path. (Later: it turns out that this
3067    is more than cosmetic for MinGW, because it doesn't like empty path
3068    components.)
3070    49. Some changes have been made to the callout feature (see 28 above):
3072    (i)  A callout function now has three choices for what it returns:
3074           0  =>  success, carry on matching
3075         > 0  =>  failure at this point, but backtrack if possible
3076         < 0  =>  serious error, return this value from pcre_exec()
3078         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3079         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3080         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3081         use by callout functions. It will never be used by PCRE itself.
3083    (ii) The pcre_extra structure (see 45 above) has a void * field called
3084         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3085         pcre_callout_block structure has a field of the same name. The contents of
3086         the field passed in the pcre_extra structure are passed to the callout
3087         function in the corresponding field in the callout block. This makes it
3088         easier to use the same callout-containing regex from multiple threads. For
3089         testing, the pcretest program has a new data escape
3091           \C*n        pass the number n (may be negative) as callout_data
3093         If the callout function in pcretest receives a non-zero value as
3094         callout_data, it returns that value.
3096    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3097    there were some redundant $(CFLAGS) in commands that are now specified as
3098    $(LINK), which already includes $(CFLAGS).
3100    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3101    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3102    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3103    one-byte characters throughout. Note that case-insensitive matching applies
3104    only to characters whose values are less than 256. PCRE doesn't support the
3105    notion of cases for higher-valued characters.
3107    (i)   A character class whose characters are all within 0-255 is handled as
3108          a bit map, and the map is inverted for negative classes. Previously, a
3109          character > 255 always failed to match such a class; however it should
3110          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3112    (ii)  A negated character class with a single character < 255 is coded as
3113          "not this character" (OP_NOT). This wasn't working properly when the test
3114          character was multibyte, either singly or repeated.
3116    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3117          mode, for example: \x{100}{2,3}.
3119    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3120          singly or repeated) now correctly test multibyte characters. However,
3121          PCRE doesn't recognize any characters with values greater than 255 as
3122          digits, spaces, or word characters. Such characters always match \D, \S,
3123          and \W, and never match \d, \s, or \w.
3125    (v)   Classes may now contain characters and character ranges with values
3126          greater than 255. For example: [ab\x{100}-\x{400}].
3128    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3129          PCRE in UTF-8 mode.
3131    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3132    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3133    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3134    value.)
3136    53. The single man page has become too large. I have therefore split it up into
3137    a number of separate man pages. These also give rise to individual HTML pages;
3138    these are now put in a separate directory, and there is an index.html page that
3139    lists them all. Some hyperlinking between the pages has been installed.
3141    54. Added convenience functions for handling named capturing parentheses.
3143    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3144    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3145    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3146    are faulted.
3148    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3149    calling configure. These values are used when compiling the dftables.c program
3150    which is run to generate the source of the default character tables. They
3151    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3152    you will need to set these values.
3154    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3157    Version 3.9 02-Jan-02
3158    ---------------------
3160    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3162    2. If --disable-static was given, the building process failed when trying to
3163    build pcretest and pcregrep. (For some reason it was using libtool to compile
3164    them, which is not right, as they aren't part of the library.)
3167    Version 3.8 18-Dec-01
3168    ---------------------
3170    1. The experimental UTF-8 code was completely screwed up. It was packing the
3171    bytes in the wrong order. How dumb can you get?
3174    Version 3.7 29-Oct-01
3175    ---------------------
3177    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3178    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3179    this didn't happen under Solaris 8, where I normally test things.
3181    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3182    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3183    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3184    this fix an hour or so after the initial 3.7 release.)
3187    Version 3.6 23-Oct-01
3188    ---------------------
3190    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3191    offsets passed as NULL with zero offset count.
3193    2. The config.guess and config.sub files had not been updated when I moved to
3194    the latest autoconf.
3197    Version 3.5 15-Aug-01
3198    ---------------------
3200    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3201    had been forgotten.
3203    2. By using declared but undefined structures, we can avoid using "void"
3204    definitions in pcre.h while keeping the internal definitions of the structures
3205    private.
3207    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3208    user point of view, this means that both static and shared libraries are built
3209    by default, but this can be individually controlled. More of the work of
3210    handling this static/shared cases is now inside libtool instead of PCRE's make
3211    file.
3213    4. The pcretest utility is now installed along with pcregrep because it is
3214    useful for users (to test regexs) and by doing this, it automatically gets
3215    relinked by libtool. The documentation has been turned into a man page, so
3216    there are now .1, .txt, and .html versions in /doc.
3218    5. Upgrades to pcregrep:
3219       (i)   Added long-form option names like gnu grep.
3220       (ii)  Added --help to list all options with an explanatory phrase.
3221       (iii) Added -r, --recursive to recurse into sub-directories.
3222       (iv)  Added -f, --file to read patterns from a file.
3224    6. pcre_exec() was referring to its "code" argument before testing that
3225    argument for NULL (and giving an error if it was NULL).
3227    7. Upgraded Makefile.in to allow for compiling in a different directory from
3228    the source directory.
3230    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3231    options bits, the pointer it was passed was to an int instead of to an unsigned
3232    long int. This mattered only on 64-bit systems.
3234    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3235    generated) instead of pcre.in, which it its source. Also made the same change
3236    in several of the .c files.
3238    10. A new release of gcc defines printf() as a macro, which broke pcretest
3239    because it had an ifdef in the middle of a string argument for printf(). Fixed
3240    by using separate calls to printf().
3242    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3243    script, to force use of CR or LF instead of \n in the source. On non-Unix
3244    systems, the value can be set in config.h.
3246    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3247    absolute limit. Changed the text of the error message to make this clear, and
3248    likewise updated the man page.
3250    13. The limit of 99 on the number of capturing subpatterns has been removed.
3251    The new limit is 65535, which I hope will not be a "real" limit.
3254    Version 3.4 22-Aug-00
3255    ---------------------
3257    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3259    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3262    Version 3.3 01-Aug-00
3263    ---------------------
3265    1. If an octal character was given, but the value was greater than \377, it
3266    was not getting masked to the least significant bits, as documented. This could
3267    lead to crashes in some systems.
3269    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3270    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3272    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3273    These just pass their arguments on to (pcre_free)(), but they are provided
3274    because some uses of PCRE bind it to non-C systems that can call its functions,
3275    but cannot call free() or pcre_free() directly.
3277    4. Add "make test" as a synonym for "make check". Corrected some comments in
3278    the Makefile.
3280    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3281    Makefile.
3283    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3284    command called pgrep for grepping around the active processes.
3286    7. Added the beginnings of support for UTF-8 character strings.
3288    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3289    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3290    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3291    out for the ar command.)
3294    Version 3.2 12-May-00
3295    ---------------------
3297    This is purely a bug fixing release.
3299    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3300    of ZA. This was just one example of several cases that could provoke this bug,
3301    which was introduced by change 9 of version 2.00. The code for breaking
3302    infinite loops after an iteration that matches an empty string was't working
3303    correctly.
3305    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3306    when matched against abbab (for example). After matching an empty string, it
3307    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3308    caused it to match further down the string than it should.
3310    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3311    was there because it doesn't seem to be needed, and it causes trouble on some
3312    systems, as it is not a Standard C header. It has been removed.
3314    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3315    were reported on the Macintosh. The changes were from
3317      while ((c = *(++ptr)) != 0 && c != '\n');
3318    to
3319      while ((c = *(++ptr)) != 0 && c != '\n') ;
3321    Totally extraordinary, but if that's what it takes...
3323    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3324    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3325    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3326    assumes the way PCRE uses memmove() (always moving upwards).
3328    6. PCRE is being used in one environment where strchr() is not available. There
3329    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3330    faster code anyway.
3333    Version 3.1 09-Feb-00
3334    ---------------------
3336    The only change in this release is the fixing of some bugs in Makefile.in for
3337    the "install" target:
3339    (1) It was failing to install pcreposix.h.
3341    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3344    Version 3.0 01-Feb-00
3345    ---------------------
3347    1. Add support for the /+ modifier to perltest (to output $` like it does in
3348    pcretest).
3350    2. Add support for the /g modifier to perltest.
3352    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3353    matches null strings.
3355    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3356    pattern. Perl treats empty patterns specially - it reuses the most recent
3357    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3358    effect.
3360    5. The POSIX interface was broken in that it was just handing over the POSIX
3361    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3362    required a bigger vector, with some working space on the end. This means that
3363    the POSIX wrapper now has to get and free some memory, and copy the results.
3365    6. Added some simple autoconf support, placing the test data and the
3366    documentation in separate directories, re-organizing some of the
3367    information files, and making it build pcre-config (a GNU standard). Also added
3368    libtool support for building PCRE as a shared library, which is now the
3369    default.
3371    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3372    09 are not valid octal constants. Single digits will be used for minor values
3373    less than 10.
3375    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3376    existing programs that set these in the