ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 19 by nigel, Sat Feb 24 21:38:33 2007 UTC revision 588 by ph10, Sat Jan 15 11:22:47 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.15 12-Jan-2011
5    ------------------------
7    1.  Fixed some typos in the markup of the man pages, and wrote a script that
8        checks for such things as part of the documentation building process.
10    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
11        --match-limit and --recursion-limit options (added for 8.11). In
12        particular, this made one of the standard tests fail. (The integer value
13        went into the wrong half of a long int.)
15    3.  If the --colour option was given to pcregrep with -v (invert match), it
16        did strange things, either producing crazy output, or crashing. It should,
17        of course, ignore a request for colour when reporting lines that do not
18        match.
20    4.  Another pcregrep bug caused similar problems if --colour was specified with
21        -M (multiline) and the pattern match finished with a line ending.
23    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
24        matched in multiline mode, the following line was shown as part of the
25        match. This seems wrong, so I have changed it.
27    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
28        the check for further matches in the same line (so they could be coloured)
29        to overrun the end of the current line. If another match was found, it was
30        incorrectly shown (and then shown again when found in the next line).
32    7.  If pcregrep was compiled under Windows, there was a reference to the
33        function pcregrep_exit() before it was defined. I am assuming this was
34        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
35        reported by a user. I've moved the definition above the reference.
38    Version 8.11 10-Dec-2010
39    ------------------------
41    1.  (*THEN) was not working properly if there were untried alternatives prior
42        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
43        backtracked to try for "b" instead of moving to the next alternative branch
44        at the same level (in this case, to look for "c"). The Perl documentation
45        is clear that when (*THEN) is backtracked onto, it goes to the "next
46        alternative in the innermost enclosing group".
48    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
49        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
50        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
51        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
52        (*THEN).
54    3.  If \s appeared in a character class, it removed the VT character from
55        the class, even if it had been included by some previous item, for example
56        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
57        of \s, but is part of the POSIX "space" class.)
59    4.  A partial match never returns an empty string (because you can always
60        match an empty string at the end of the subject); however the checking for
61        an empty string was starting at the "start of match" point. This has been
62        changed to the "earliest inspected character" point, because the returned
63        data for a partial match starts at this character. This means that, for
64        example, /(?<=abc)def/ gives a partial match for the subject "abc"
65        (previously it gave "no match").
67    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
68        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
69        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
70        has an implication that the given string is incomplete (because a partial
71        match is preferred over a full match). For this reason, these items now
72        give a partial match in this situation. [Aside: previously, the one case
73        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
74        match rather than a full match, which was wrong by the old rules, but is
75        now correct.]
77    6.  There was a bug in the handling of #-introduced comments, recognized when
78        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
79        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
80        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
81        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
82        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
83        places in pcre_compile().
85    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
86        comments when looking ahead for named forward references to subpatterns,
87        the only newline sequence it recognized was NL. It now handles newlines
88        according to the set newline convention.
90    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
91        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
92        cater for a lack of strerror(). These oversights have been fixed.
94    9.  Added --match-limit and --recursion-limit to pcregrep.
96    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
98    11. When the -o option was used, pcregrep was setting a return code of 1, even
99        when matches were found, and --line-buffered was not being honoured.
101    12. Added an optional parentheses number to the -o and --only-matching options
102        of pcregrep.
104    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
105        can match an empty string. The code to do it in pcretest and pcredemo
106        needed fixing:
108        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
109            only one byte after an empty string match just before CRLF (this case
110            just got forgotten; "any" and "anycrlf" were OK).
112        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
113            mode when an empty string match preceded an ASCII character followed by
114            a non-ASCII character. (The code for advancing by one character rather
115            than one byte was nonsense.)
117        (c) The pcredemo.c sample program did not have any code at all to handle
118            the cases when CRLF is a valid newline sequence.
120    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
121        as a starting offset was within the subject string. There is now a new
122        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
123        negative or greater than the length of the string. In order to test this,
124        pcretest is extended to allow the setting of negative starting offsets.
126    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
127        starting offset points to the beginning of a UTF-8 character was
128        unnecessarily clumsy. I tidied it up.
130    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
131        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
133    17. Nobody had reported that the --include_dir option, which was added in
134        release 7.7 should have been called --include-dir (hyphen, not underscore)
135        for compatibility with GNU grep. I have changed it to --include-dir, but
136        left --include_dir as an undocumented synonym, and the same for
137        --exclude-dir, though that is not available in GNU grep, at least as of
138        release 2.5.4.
140    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
141        characters from a string of bytes) have been redefined so as not to use
142        loops, in order to improve performance in some environments. At the same
143        time, I abstracted some of the common code into auxiliary macros to save
144        repetition (this should not affect the compiled code).
146    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
147        compile-time error is now given if \c is not followed by an ASCII
148        character, that is, a byte less than 128. (In EBCDIC mode, the code is
149        different, and any byte value is allowed.)
151    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
152        START_OPTIMIZE option, which is now allowed at compile time - but just
153        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
154        to pcregrep and other applications that have no direct access to PCRE
155        options. The new /Y option in pcretest sets this option when calling
156        pcre_compile().
158    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
159        back references. Groups containing recursive back references were forced to
160        be atomic by that change, but in the case of named groups, the amount of
161        memory required was incorrectly computed, leading to "Failed: internal
162        error: code overflow". This has been fixed.
164    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
165        pcretest.c, to avoid build problems in some Borland environments.
168    Version 8.10 25-Jun-2010
169    ------------------------
171    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
172        THEN.
174    2.  (*ACCEPT) was not working when inside an atomic group.
176    3.  Inside a character class, \B is treated as a literal by default, but
177        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
178        causes the error). The code is unchanged, but I tidied the documentation.
180    4.  Inside a character class, PCRE always treated \R and \X as literals,
181        whereas Perl faults them if its -w option is set. I have changed PCRE so
182        that it faults them when PCRE_EXTRA is set.
184    5.  Added support for \N, which always matches any character other than
185        newline. (It is the same as "." when PCRE_DOTALL is not set.)
187    6.  When compiling pcregrep with newer versions of gcc which may have
188        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
189        declared with attribute warn_unused_result" were given. Just casting the
190        result to (void) does not stop the warnings; a more elaborate fudge is
191        needed. I've used a macro to implement this.
193    7.  Minor change to pcretest.c to avoid a compiler warning.
195    8.  Added four artifical Unicode properties to help with an option to make
196        \s etc use properties (see next item). The new properties are: Xan
197        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
199    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
200        use Unicode properties. (*UCP) at the start of a pattern can be used to set
201        this option. Modified pcretest to add /W to test this facility. Added
202        REG_UCP to make it available via the POSIX interface.
204    10. Added --line-buffered to pcregrep.
206    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
207        studied, and the match started with a letter with a code point greater than
208        127 whose first byte was different to the first byte of the other case of
209        the letter, the other case of this starting letter was not recognized
210        (#976).
212    12. If a pattern that was studied started with a repeated Unicode property
213        test, for example, \p{Nd}+, there was the theoretical possibility of
214        setting up an incorrect bitmap of starting bytes, but fortunately it could
215        not have actually happened in practice until change 8 above was made (it
216        added property types that matched character-matching opcodes).
218    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
219        possible starting bytes for non-anchored patterns.
221    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
222        \R, and also a number of cases that involve Unicode properties, both
223        explicit and implicit when PCRE_UCP is set.
225    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
226        input, it could crash or give wrong results if characters with values
227        greater than 0xc0 were present in the subject string. (Detail: it assumed
228        UTF-8 input when processing these items.)
230    16. Added a lot of (int) casts to avoid compiler warnings in systems where
231        size_t is 64-bit (#991).
233    17. Added a check for running out of memory when PCRE is compiled with
234        --disable-stack-for-recursion (#990).
236    18. If the last data line in a file for pcretest does not have a newline on
237        the end, a newline was missing in the output.
239    19. The default pcre_chartables.c file recognizes only ASCII characters (values
240        less than 128) in its various bitmaps. However, there is a facility for
241        generating tables according to the current locale when PCRE is compiled. It
242        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
243        characters, are recognized by isspace() and therefore were getting set in
244        these tables, and indeed these tables seem to approximate to ISO 8859. This
245        caused a problem in UTF-8 mode when pcre_study() was used to create a list
246        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
247        which of course cannot start UTF-8 characters. I have changed the code so
248        that only real ASCII characters (less than 128) and the correct starting
249        bytes for UTF-8 encodings are set for characters greater than 127 when in
250        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
251        altogether.)
253    20. Added the /T option to pcretest so as to be able to run tests with non-
254        standard character tables, thus making it possible to include the tests
255        used for 19 above in the standard set of tests.
257    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
258        reference to a subpattern the other side of a comment that contains an
259        opening parenthesis caused either an internal compiling error, or a
260        reference to the wrong subpattern.
263    Version 8.02 19-Mar-2010
264    ------------------------
266    1.  The Unicode data tables have been updated to Unicode 5.2.0.
268    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
269        configured.
271    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
272        original author of that file, following a query about its status.
274    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
275        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
277    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
278        quantifier applied to a forward-referencing subroutine call, could compile
279        incorrect code or give the error "internal error: previously-checked
280        referenced subpattern not found".
282    6.  Both MS Visual Studio and Symbian OS have problems with initializing
283        variables to point to external functions. For these systems, therefore,
284        pcre_malloc etc. are now initialized to local functions that call the
285        relevant global functions.
287    7.  There were two entries missing in the vectors called coptable and poptable
288        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
289        I've fixed the data, and added a kludgy way of testing at compile time that
290        the lengths are correct (equal to the number of opcodes).
292    8.  Following on from 7, I added a similar kludge to check the length of the
293        eint vector in pcreposix.c.
295    9.  Error texts for pcre_compile() are held as one long string to avoid too
296        much relocation at load time. To find a text, the string is searched,
297        counting zeros. There was no check for running off the end of the string,
298        which could happen if a new error number was added without updating the
299        string.
301    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
303    11. \K was not working if it appeared in an atomic group or in a group that
304        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
305        \K is "not well defined" if used in an assertion. PCRE now accepts it if
306        the assertion is positive, but not if it is negative.
308    12. Change 11 fortuitously reduced the size of the stack frame used in the
309        "match()" function of pcre_exec.c by one pointer. Forthcoming
310        implementation of support for (*MARK) will need an extra pointer on the
311        stack; I have reserved it now, so that the stack frame size does not
312        decrease.
314    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
315        item in branch that calls a recursion is a subroutine call - as in the
316        second branch in the above example - was incorrectly given the compile-
317        time error "recursive call could loop indefinitely" because pcre_compile()
318        was not correctly checking the subroutine for matching a non-empty string.
320    14. The checks for overrunning compiling workspace could trigger after an
321        overrun had occurred. This is a "should never occur" error, but it can be
322        triggered by pathological patterns such as hundreds of nested parentheses.
323        The checks now trigger 100 bytes before the end of the workspace.
325    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
328    Version 8.01 19-Jan-2010
329    ------------------------
331    1.  If a pattern contained a conditional subpattern with only one branch (in
332        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
333        computed the wrong minimum data length (which is of course zero for such
334        subpatterns). This could cause incorrect "no match" results.
336    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
337        the pattern is reset in the first branch, pcre_compile() failed with
338        "internal error: code overflow at offset...". This happened only when
339        the reset was to the original external option setting. (An optimization
340        abstracts leading options settings into an external setting, which was the
341        cause of this.)
343    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
344        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
345        assertion pattern did not match (meaning that the assertion was true), it
346        was incorrectly treated as false if the SKIP had been reached during the
347        matching. This also applied to assertions used as conditions.
349    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
350        assertion subpattern, including such a pattern used as a condition,
351        unpredictable results occurred, instead of the error return
354    5.  The C++ GlobalReplace function was not working like Perl for the special
355        situation when an empty string is matched. It now does the fancy magic
356        stuff that is necessary.
358    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
359        removed. (These were left over from very, very early versions of PCRE.)
361    7.  Some cosmetic changes to the code to make life easier when compiling it
362        as part of something else:
364        (a) Change DEBUG to PCRE_DEBUG.
366        (b) In pcre_compile(), rename the member of the "branch_chain" structure
367            called "current" as "current_branch", to prevent a collision with the
368            Linux macro when compiled as a kernel module.
370        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
371            prevent a collision with the Linux macro when compiled as a kernel
372            module.
374    8.  In pcre_compile() there are some checks for integer overflows that used to
375        cast potentially large values to (double). This has been changed to that
376        when building, a check for int64_t is made, and if it is found, it is used
377        instead, thus avoiding the use of floating point arithmetic. (There is no
378        other use of FP in PCRE.) If int64_t is not found, the fallback is to
379        double.
381    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
382        2005 (difference between two addresses compared to an unsigned value).
384    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
385        custom one, because of the following reported problem in Windows:
387          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
388              under Win32.
389          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
390              therefore missing the function definition.
391          - The compiler thus generates a "C" signature for the test function.
392          - The linker fails to find the "C" function.
393          - PCRE fails to configure if asked to do so against libbz2.
395    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
396        messages were output:
398          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
399          rerunning libtoolize, to keep the correct libtool macros in-tree.
400          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
402        I have done both of these things.
404    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
405        most of the time, it *can* run out if it is given a pattern that contains a
406        runaway infinite recursion. I updated the discussion in the pcrestack man
407        page.
409    13. Now that we have gone to the x.xx style of version numbers, the minor
410        version may start with zero. Using 08 or 09 is a bad idea because users
411        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
412        interpreted as invalid octal numbers. I've updated the previous comment in
413        configure.ac, and also added a check that gives an error if 08 or 09 are
414        used.
416    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
417        causing partial matching to fail when the end of the subject matched \W
418        in a UTF-8 pattern where \W was quantified with a minimum of 3.
420    15. There were some discrepancies between the declarations in pcre_internal.h
421        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
422        their definitions. The declarations used "const uschar *" and the
423        definitions used USPTR. Even though USPTR is normally defined as "const
424        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
425        reported that: "This difference in casting confuses some C++ compilers, for
426        example, SunCC recognizes above declarations as different functions and
427        generates broken code for hbpcre." I have changed the declarations to use
428        USPTR.
430    16. GNU libtool is named differently on some systems. The autogen.sh script now
431        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
432        (FreeBSD).
434    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
435        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
436        comment: "Figure out how to create a longlong from a string: strtoll and
437        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
438        instance, but it only takes 2 args instead of 3!"
440    18. A subtle bug concerned with back references has been fixed by a change of
441        specification, with a corresponding code fix. A pattern such as
442        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
443        refers, was giving matches when it shouldn't. For example, xa=xaaa would
444        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
445        same bug. Such groups have to be quantified to be useful, or contained
446        inside another quantified group. (If there's no repetition, the reference
447        can never match.) The problem arises because, having left the group and
448        moved on to the rest of the pattern, a later failure that backtracks into
449        the group uses the captured value from the final iteration of the group
450        rather than the correct earlier one. I have fixed this in PCRE by forcing
451        any group that contains a reference to itself to be an atomic group; that
452        is, there cannot be any backtracking into it once it has completed. This is
453        similar to recursive and subroutine calls.
456    Version 8.00 19-Oct-09
457    ----------------------
459    1.  The table for translating pcre_compile() error codes into POSIX error codes
460        was out-of-date, and there was no check on the pcre_compile() error code
461        being within the table. This could lead to an OK return being given in
462        error.
464    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
465        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
466        in a Windows environment.
468    3.  The pcregrep --count option prints the count for each file even when it is
469        zero, as does GNU grep. However, pcregrep was also printing all files when
470        --files-with-matches was added. Now, when both options are given, it prints
471        counts only for those files that have at least one match. (GNU grep just
472        prints the file name in this circumstance, but including the count seems
473        more useful - otherwise, why use --count?) Also ensured that the
474        combination -clh just lists non-zero counts, with no names.
476    4.  The long form of the pcregrep -F option was incorrectly implemented as
477        --fixed_strings instead of --fixed-strings. This is an incompatible change,
478        but it seems right to fix it, and I didn't think it was worth preserving
479        the old behaviour.
481    5.  The command line items --regex=pattern and --regexp=pattern were not
482        recognized by pcregrep, which required --regex pattern or --regexp pattern
483        (with a space rather than an '='). The man page documented the '=' forms,
484        which are compatible with GNU grep; these now work.
486    6.  No libpcreposix.pc file was created for pkg-config; there was just
487        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
489    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
490        when UCP support is not needed, by modifying the Python script that
491        generates it from Unicode data files. This should not matter if the module
492        is correctly used as a library, but I received one complaint about 50K of
493        unwanted data. My guess is that the person linked everything into his
494        program rather than using a library. Anyway, it does no harm.
496    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
497        was a minimum greater than 1 for a wide character in a possessive
498        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
499        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
500        character. Chaos in the form of incorrect output or a compiling loop could
501        result.
503    9.  The restrictions on what a pattern can contain when partial matching is
504        requested for pcre_exec() have been removed. All patterns can now be
505        partially matched by this function. In addition, if there are at least two
506        slots in the offset vector, the offset of the earliest inspected character
507        for the match and the offset of the end of the subject are set in them when
508        PCRE_ERROR_PARTIAL is returned.
510    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
511        synonymous with PCRE_PARTIAL, for backwards compatibility, and
512        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
513        and may be more useful for multi-segment matching.
515    11. Partial matching with pcre_exec() is now more intuitive. A partial match
516        used to be given if ever the end of the subject was reached; now it is
517        given only if matching could not proceed because another character was
518        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
519        string "Z", which now yields "no match" instead of "partial match". In the
520        case of pcre_dfa_exec(), "no match" is given if every matching path for the
521        final character ended with (*FAIL).
523    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
524        if the pattern had a "must contain" character that was already found in the
525        earlier partial match, unless partial matching was again requested. For
526        example, with the pattern /dog.(body)?/, the "must contain" character is
527        "g". If the first part-match was for the string "dog", restarting with
528        "sbody" failed. This bug has been fixed.
530    13. The string returned by pcre_dfa_exec() after a partial match has been
531        changed so that it starts at the first inspected character rather than the
532        first character of the match. This makes a difference only if the pattern
533        starts with a lookbehind assertion or \b or \B (\K is not supported by
534        pcre_dfa_exec()). It's an incompatible change, but it makes the two
535        matching functions compatible, and I think it's the right thing to do.
537    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
538        so that the demonstration program is easily available in environments where
539        PCRE has not been installed from source.
541    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
542        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
543        library.
545    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
546        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
547        is not the first non-POSIX option to be added. Clearly some people find
548        these options useful.
550    17. If a caller to the POSIX matching function regexec() passes a non-zero
551        value for nmatch with a NULL value for pmatch, the value of
552        nmatch is forced to zero.
554    18. RunGrepTest did not have a test for the availability of the -u option of
555        the diff command, as RunTest does. It now checks in the same way as
556        RunTest, and also checks for the -b option.
558    19. If an odd number of negated classes containing just a single character
559        interposed, within parentheses, between a forward reference to a named
560        subpattern and the definition of the subpattern, compilation crashed with
561        an internal error, complaining that it could not find the referenced
562        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
563        [The bug was that it was starting one character too far in when skipping
564        over the character class, thus treating the ] as data rather than
565        terminating the class. This meant it could skip too much.]
567    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
568        /g option in pcretest when the pattern contains \K, which makes it possible
569        to have an empty string match not at the start, even when the pattern is
570        anchored. Updated pcretest and pcredemo to use this option.
572    21. If the maximum number of capturing subpatterns in a recursion was greater
573        than the maximum at the outer level, the higher number was returned, but
574        with unset values at the outer level. The correct (outer level) value is
575        now given.
577    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
578        PCRE did not set those parentheses (unlike Perl). I have now found a way to
579        make it do so. The string so far is captured, making this feature
580        compatible with Perl.
582    23. The tests have been re-organized, adding tests 11 and 12, to make it
583        possible to check the Perl 5.10 features against Perl 5.10.
585    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
586        pattern matches a fixed length string. PCRE did not allow this; now it
587        does. Neither allows recursion.
589    25. I finally figured out how to implement a request to provide the minimum
590        length of subject string that was needed in order to match a given pattern.
591        (It was back references and recursion that I had previously got hung up
592        on.) This code has now been added to pcre_study(); it finds a lower bound
593        to the length of subject needed. It is not necessarily the greatest lower
594        bound, but using it to avoid searching strings that are too short does give
595        some useful speed-ups. The value is available to calling programs via
596        pcre_fullinfo().
598    26. While implementing 25, I discovered to my embarrassment that pcretest had
599        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
600        study optimizations had never been tested with that matching function.
601        Oops. What is worse, even when it was passed study data, there was a bug in
602        pcre_dfa_exec() that meant it never actually used it. Double oops. There
603        were also very few tests of studied patterns with pcre_dfa_exec().
605    27. If (?| is used to create subpatterns with duplicate numbers, they are now
606        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
607        on the other side of the coin, they are no longer allowed to have different
608        names, because these cannot be distinguished in PCRE, and this has caused
609        confusion. (This is a difference from Perl.)
611    28. When duplicate subpattern names are present (necessarily with different
612        numbers, as required by 27 above), and a test is made by name in a
613        conditional pattern, either for a subpattern having been matched, or for
614        recursion in such a pattern, all the associated numbered subpatterns are
615        tested, and the overall condition is true if the condition is true for any
616        one of them. This is the way Perl works, and is also more like the way
617        testing by number works.
620    Version 7.9 11-Apr-09
621    ---------------------
623    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
624        (pcretest), all targets were linked against these libraries. This included
625        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
626        libraries. This caused unwanted dependencies to be created. This problem
627        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
628        pcretest is linked with readline.
630    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
631        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
632        moved outside it again, because FALSE and TRUE are already defined in AIX,
633        but BOOL is not.
635    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
636        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
638    4.  The pcregrep documentation said spaces were inserted as well as colons (or
639        hyphens) following file names and line numbers when outputting matching
640        lines. This is not true; no spaces are inserted. I have also clarified the
641        wording for the --colour (or --color) option.
643    5.  In pcregrep, when --colour was used with -o, the list of matching strings
644        was not coloured; this is different to GNU grep, so I have changed it to be
645        the same.
647    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
648        each matching line was coloured. Now it goes on to look for further matches
649        of any of the test patterns, which is the same behaviour as GNU grep.
651    7.  A pattern that could match an empty string could cause pcregrep to loop; it
652        doesn't make sense to accept an empty string match in pcregrep, so I have
653        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
654        seems to be how GNU grep behaves.
656    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
657        start or after a newline", because the conditional assertion was not being
658        correctly handled. The rule now is that both the assertion and what follows
659        in the first alternative must satisfy the test.
661    9.  If auto-callout was enabled in a pattern with a conditional group whose
662        condition was an assertion, PCRE could crash during matching, both with
663        pcre_exec() and pcre_dfa_exec().
665    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
666        used for matching.
668    11. Unicode property support in character classes was not working for
669        characters (bytes) greater than 127 when not in UTF-8 mode.
671    12. Added the -M command line option to pcretest.
673    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
675    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
677    16. Added comments and documentation about mis-use of no_arg in the C++
678        wrapper.
680    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
681        from Martin Jerabek that uses macro names for all relevant character and
682        string constants.
684    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
685        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
686        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
687        these, but not everybody uses configure.
689    19. A conditional group that had only one branch was not being correctly
690        recognized as an item that could match an empty string. This meant that an
691        enclosing group might also not be so recognized, causing infinite looping
692        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
693        with the subject "ab", where knowledge that the repeated group can match
694        nothing is needed in order to break the loop.
696    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
697        exec(), but without supplying a callout function, matching went wrong.
699    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
700        leak if the size of the offset vector was greater than 30. When the vector
701        is smaller, the saved offsets during recursion go onto a local stack
702        vector, but for larger vectors malloc() is used. It was failing to free
703        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
704        error, in fact).
706    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
707        heapframe that is used only when UTF-8 support is enabled. This caused no
708        problem, but was untidy.
710    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
711        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
712        included within another project.
714    24. Steven Van Ingelgem's patches to add more options to the CMake support,
715        slightly modified by me:
717          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
718              not building pcregrep.
720          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
721              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
723    25. Forward references, both numeric and by name, in patterns that made use of
724        duplicate group numbers, could behave incorrectly or give incorrect errors,
725        because when scanning forward to find the reference group, PCRE was not
726        taking into account the duplicate group numbers. A pattern such as
727        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
729    26. Changed a few more instances of "const unsigned char *" to USPTR, making
730        the feature of a custom pointer more persuasive (as requested by a user).
732    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
733        pcretest.c, inside #ifndefs, because it seems they are sometimes already
734        pre-defined.
736    28. Added support for (*UTF8) at the start of a pattern.
738    29. Arrange for flags added by the "release type" setting in CMake to be shown
739        in the configuration summary.
742    Version 7.8 05-Sep-08
743    ---------------------
745    1.  Replaced UCP searching code with optimized version as implemented for Ad
746        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
747        stage table and inline lookup instead of a function, giving speed ups of 2
748        to 5 times on some simple patterns that I tested. Permission was given to
749        distribute the MultiStage2.py script that generates the tables (it's not in
750        the tarball, but is in the Subversion repository).
752    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
753        scripts.
755    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
756        a group with a zero qualifier. The result of the study could be incorrect,
757        or the function might crash, depending on the pattern.
759    4.  Caseless matching was not working for non-ASCII characters in back
760        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
761        It now works when Unicode Property Support is available.
763    5.  In pcretest, an escape such as \x{de} in the data was always generating
764        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
765        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
766        truncation.
768    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
770    7.  Added two (int) casts to pcregrep when printing the difference of two
771        pointers, in case they are 64-bit values.
773    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
774        test 2 if it fails.
776    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
777        and a #define of that name to empty if it is not externally set. This is to
778        allow users of MSVC to set it if necessary.
780    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
781        the convenience functions in the pcre_get.c source file.
783    11. An option change at the start of a pattern that had top-level alternatives
784        could cause overwriting and/or a crash. This command provoked a crash in
785        some environments:
787          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
789        This potential security problem was recorded as CVE-2008-2371.
791    12. For a pattern where the match had to start at the beginning or immediately
792        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
793        pcre_dfa_exec() could read past the end of the passed subject if there was
794        no match. To help with detecting such bugs (e.g. with valgrind), I modified
795        pcretest so that it places the subject at the end of its malloc-ed buffer.
797    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
798        exec() might read past the end of the data buffer in UTF-8 mode.
800    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
801        the data contained the byte 0x85 as part of a UTF-8 character within its
802        first line. This applied both to normal and DFA matching.
804    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
805        /^[^d]*?$/8 failed to match "abc".
807    16. Added a missing copyright notice to pcrecpp_internal.h.
809    17. Make it more clear in the documentation that values returned from
810        pcre_exec() in ovector are byte offsets, not character counts.
812    18. Tidied a few places to stop certain compilers from issuing warnings.
814    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
815        supplied by Stefan Weber. I made a further small update for 7.8 because
816        there is a change of source arrangements: the pcre_searchfuncs.c module is
817        replaced by pcre_ucd.c.
820    Version 7.7 07-May-08
821    ---------------------
823    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
824        a string to a long long, pretend we don't even have a long long." This is
825        done by checking for the strtoq, strtoll, and _strtoi64 functions.
827    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
828        pre-7.6 versions, which defined a global no_arg variable instead of putting
829        it in the RE class. (See also #8 below.)
831    3.  Remove a line of dead code, identified by coverity and reported by Nuno
832        Lopes.
834    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
836        (1) The include/exclude patterns were being applied to the whole pathnames
837            of files, instead of just to the final components.
839        (2) If there was more than one level of directory, the subdirectories were
840            skipped unless they satisfied the include/exclude conditions. This is
841            inconsistent with GNU grep (and could even be seen as contrary to the
842            pcregrep specification - which I improved to make it absolutely clear).
843            The action now is always to scan all levels of directory, and just
844            apply the include/exclude patterns to regular files.
846    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
847        --exclude_dir in the tests to avoid scanning .svn directories.
849    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
850        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
851        doesn't support NULs in patterns.
853    7.  Added some missing "const"s to declarations of static tables in
854        pcre_compile.c and pcre_dfa_exec.c.
856    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
857        caused by fix #2  above. (Subsequently also a second patch to fix the
858        first patch. And a third patch - this was a messy problem.)
860    9.  Applied Craig's patch to remove the use of push_back().
862    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
863        matching function regexec().
865    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
866        which, however, unlike Perl's \g{...}, are subroutine calls, not back
867        references. PCRE supports relative numbers with this syntax (I don't think
868        Oniguruma does).
870    12. Previously, a group with a zero repeat such as (...){0} was completely
871        omitted from the compiled regex. However, this means that if the group
872        was called as a subroutine from elsewhere in the pattern, things went wrong
873        (an internal error was given). Such groups are now left in the compiled
874        pattern, with a new opcode that causes them to be skipped at execution
875        time.
877    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
878        to the way PCRE behaves:
880        (a) A lone ] character is dis-allowed (Perl treats it as data).
882        (b) A back reference to an unmatched subpattern matches an empty string
883            (Perl fails the current match path).
885        (c) A data ] in a character class must be notated as \] because if the
886            first data character in a class is ], it defines an empty class. (In
887            Perl it is not possible to have an empty class.) The empty class []
888            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
889            The negative empty class [^] matches any one character, independently
890            of the DOTALL setting.
892    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
893        non-existent subpattern following a character class starting with ']' and
894        containing () gave an internal compiling error instead of "reference to
895        non-existent subpattern". Fortunately, when the pattern did exist, the
896        compiled code was correct. (When scanning forwards to check for the
897        existencd of the subpattern, it was treating the data ']' as terminating
898        the class, so got the count wrong. When actually compiling, the reference
899        was subsequently set up correctly.)
901    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
902        it was being rejected as not supported by pcre_dfa_exec(), even though
903        other assertions are supported. I have made pcre_dfa_exec() support
904        (*FAIL).
906    16. The implementation of 13c above involved the invention of a new opcode,
907        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
908        cannot be changed at match time, I realized I could make a small
909        improvement to matching performance by compiling OP_ALLANY instead of
910        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
911        on the OP_ANY path.
913    17. Compiling pcretest on Windows with readline support failed without the
914        following two fixes: (1) Make the unistd.h include conditional on
915        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
917    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
918        ncurses library to be included for pcretest when ReadLine support is
919        requested, but also to allow for it to be overridden. This patch came from
920        Daniel Bergström.
922    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
923        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
924        any errors with the current Unicode tables. Thanks to Peter Kankowski for
925        spotting this.
928    Version 7.6 28-Jan-08
929    ---------------------
931    1.  A character class containing a very large number of characters with
932        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
933        overflow.
935    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
936        HAVE_LONG_LONG is not defined.
938    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
939        bring them up to date and include new features. This patch includes:
941        - Fixed PH's badly added libz and libbz2 support.
942        - Fixed a problem with static linking.
943        - Added pcredemo. [But later removed - see 7 below.]
944        - Fixed dftables problem and added an option.
945        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
946            HAVE_LONG_LONG.
947        - Added readline support for pcretest.
948        - Added an listing of the option settings after cmake has run.
950    4.  A user submitted a patch to Makefile that makes it easy to create
951        "pcre.dll" under mingw when using Configure/Make. I added stuff to
952        Makefile.am that cause it to include this special target, without
953        affecting anything else. Note that the same mingw target plus all
954        the other distribution libraries and programs are now supported
955        when configuring with CMake (see 6 below) instead of with
956        Configure/Make.
958    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
959        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
960        exported in the Windows port". It has not yet been confirmed that the patch
961        solves the problem, but it does no harm.
963    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
964        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
965        with CMake, and also correct the comment about stack recursion.
967    7.  Remove the automatic building of pcredemo from the ./configure system and
968        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
969        of a program that users should build themselves after PCRE is installed, so
970        building it automatically is not really right. What is more, it gave
971        trouble in some build environments.
973    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
976    Version 7.5 10-Jan-08
977    ---------------------
979    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
980        values in parens when parsing an RE using the C++ wrapper."
982    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
983        Characters greater than 255 were excluded from the class instead of being
984        included.
986    3.  The same bug as (2) above applied to negated POSIX classes such as
987        [:^space:].
989    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
990        defined or documented. It seems to have been a typo for PCRE_STATIC, so
991        I have changed it.
993    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
994        first named subpattern) and a construct such as (?&a) would reference the
995        first named subpattern whose name started with "a" (in other words, the
996        length check was missing). Both these problems are fixed. "Subpattern name
997        expected" is now given for (?&) (a zero-length name), and this patch also
998        makes it give the same error for \k'' (previously it complained that that
999        was a reference to a non-existent subpattern).
1001    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1002        this is right because (?- can be followed by option settings as well as by
1003        digits. I have, however, made the messages clearer.
1005    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1006        than the number used in the conditional) now cause a compile-time error.
1007        This is actually not compatible with Perl, which accepts such patterns, but
1008        treats the conditional as always being FALSE (as PCRE used to), but it
1009        seems to me that giving a diagnostic is better.
1011    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1012        and messages.
1014    9.  Fix two occurrences of "backslash" in comments that should have been
1015        "backspace".
1017    10. Remove two redundant lines of code that can never be obeyed (their function
1018        was moved elsewhere).
1020    11. The program that makes PCRE's Unicode character property table had a bug
1021        which caused it to generate incorrect table entries for sequences of
1022        characters that have the same character type, but are in different scripts.
1023        It amalgamated them into a single range, with the script of the first of
1024        them. In other words, some characters were in the wrong script. There were
1025        thirteen such cases, affecting characters in the following ranges:
1027          U+002b0 - U+002c1
1028          U+0060c - U+0060d
1029          U+0061e - U+00612
1030          U+0064b - U+0065e
1031          U+0074d - U+0076d
1032          U+01800 - U+01805
1033          U+01d00 - U+01d77
1034          U+01d9b - U+01dbf
1035          U+0200b - U+0200f
1036          U+030fc - U+030fe
1037          U+03260 - U+0327f
1038          U+0fb46 - U+0fbb1
1039          U+10450 - U+1049d
1041    12. The -o option (show only the matching part of a line) for pcregrep was not
1042        compatible with GNU grep in that, if there was more than one match in a
1043        line, it showed only the first of them. It now behaves in the same way as
1044        GNU grep.
1046    13. If the -o and -v options were combined for pcregrep, it printed a blank
1047        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1048        does the same. The return code can be used to tell if there were any
1049        non-matching lines.
1051    14. Added --file-offsets and --line-offsets to pcregrep.
1053    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1054        infinitely looping recursion. The bug was that positive lookaheads were not
1055        being skipped when checking for a possible empty match (negative lookaheads
1056        and both kinds of lookbehind were skipped).
1058    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1059        inclusion of <windows.h> to before rather than after the definition of
1060        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1062    17. Specifying a possessive quantifier with a specific limit for a Unicode
1063        character property caused pcre_compile() to compile bad code, which led at
1064        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1065        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1066        caused the error; without that there was no problem.
1068    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1070    19. Added --enable-pcretest-libreadline.
1072    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1073        RE::GlobalReplace(). As a result, the number of replacements returned was
1074        double what it should be. I removed one of the increments, but Craig sent a
1075        later patch that removed the other one (the right fix) and added unit tests
1076        that check the return values (which was not done before).
1078    21. Several CMake things:
1080        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1081            the names libpcre and libpcreposix, not just pcre and pcreposix.
1083        (2) The above change means that pcretest and pcregrep are now correctly
1084            linked with the newly-built libraries, not previously installed ones.
1088    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1089        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1090        UTF-8 newline character). The key issue is that the pattern starts .*;
1091        this means that the match must be either at the beginning, or after a
1092        newline. The bug was in the code for advancing after a failed match and
1093        checking that the new position followed a newline. It was not taking
1094        account of UTF-8 characters correctly.
1096    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1097        character classes. PCRE was not treating the sequence [:...:] as a
1098        character class unless the ... were all letters. Perl, however, seems to
1099        allow any characters between [: and :], though of course it rejects as
1100        unknown any "names" that contain non-letters, because all the known class
1101        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1102        for example, whereas PCRE did not - it did not recognize a POSIX character
1103        class. This seemed a bit dangerous, so the code has been changed to be
1104        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1105        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1106        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1107        Perl does, and where it didn't before.
1109    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1110        Windows environments %n is disabled by default.
1113    Version 7.4 21-Sep-07
1114    ---------------------
1116    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1117        means that a class such as [\s] counted as "explicit reference to CR or
1118        LF". That isn't really right - the whole point of the change was to try to
1119        help when there was an actual mention of one of the two characters. So now
1120        the change happens only if \r or \n (or a literal CR or LF) character is
1121        encountered.
1123    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1124        of both had grown to the point where there were only 3 bits left.
1125        Fortunately, there was spare space in the data structure, and so I have
1126        moved the internal flags into a new 16-bit field to free up more option
1127        bits.
1129    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1130        but did not set the internal JCHANGED flag - either of these is enough to
1131        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1132        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1133        start sets both bits.
1135    4.  Added options (at build time, compile time, exec time) to change \R from
1136        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1138    5.  doc/pcresyntax.html was missing from the distribution.
1140    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1141        compatibility, even though it is no longer used.
1143    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1144        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1145        windows.h file is present (where different names are used). [This was
1146        reversed later after testing - see 16 below.]
1148    8.  Changed all #include <config.h> to #include "config.h". There were also
1149        some further <pcre.h> cases that I changed to "pcre.h".
1151    9.  When pcregrep was used with the --colour option, it missed the line ending
1152        sequence off the lines that it output.
1154    10. It was pointed out to me that arrays of string pointers cause lots of
1155        relocations when a shared library is dynamically loaded. A technique of
1156        using a single long string with a table of offsets can drastically reduce
1157        these. I have refactored PCRE in four places to do this. The result is
1158        dramatic:
1160          Originally:                          290
1161          After changing UCP table:            187
1162          After changing error message table:   43
1163          After changing table of "verbs"       36
1164          After changing table of Posix names   22
1166        Thanks to the folks working on Gregex for glib for this insight.
1168    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1169        unicode-properties was also set.
1171    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1173    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1174        checked only for CRLF.
1176    14. Added casts to pcretest.c to avoid compiler warnings.
1178    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1180    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1181        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1182        entirely. This removes changes made in 7 above.
1184    17. The CMake files have been updated, and there is now more information about
1185        building with CMake in the NON-UNIX-USE document.
1188    Version 7.3 28-Aug-07
1189    ---------------------
1191     1. In the rejigging of the build system that eventually resulted in 7.1, the
1192        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1193        brackets there is not right, since it causes compilers to look for an
1194        installed pcre.h, not the version that is in the source that is being
1195        compiled (which of course may be different). I have changed it back to:
1197          #include "pcre.h"
1199        I have a vague recollection that the change was concerned with compiling in
1200        different directories, but in the new build system, that is taken care of
1201        by the VPATH setting the Makefile.
1203     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1204        when the subject happened to end in the byte 0x85 (e.g. if the last
1205        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1206        characters but of course it shouldn't be taken as a newline when it is part
1207        of another character. The bug was that, for an unlimited repeat of . in
1208        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1209        characters when looking for a newline.
1211     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1213     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1214        in debug output.
1216     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1217        long printing in the pcrecpp unittest when running under MinGW.
1219     6. ESC_K was left out of the EBCDIC table.
1221     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1222        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1223        limit also applies to "virtual nesting" when a pattern is recursive, and in
1224        this case 1000 isn't so big. I have been able to remove this limit at the
1225        expense of backing off one optimization in certain circumstances. Normally,
1226        when pcre_exec() would call its internal match() function recursively and
1227        immediately return the result unconditionally, it uses a "tail recursion"
1228        feature to save stack. However, when a subpattern that can match an empty
1229        string has an unlimited repetition quantifier, it no longer makes this
1230        optimization. That gives it a stack frame in which to save the data for
1231        checking that an empty string has been matched. Previously this was taken
1232        from the 1000-entry workspace that had been reserved. So now there is no
1233        explicit limit, but more stack is used.
1235     8. Applied Daniel's patches to solve problems with the import/export magic
1236        syntax that is required for Windows, and which was going wrong for the
1237        pcreposix and pcrecpp parts of the library. These were overlooked when this
1238        problem was solved for the main library.
1240     9. There were some crude static tests to avoid integer overflow when computing
1241        the size of patterns that contain repeated groups with explicit upper
1242        limits. As the maximum quantifier is 65535, the maximum group length was
1243        set at 30,000 so that the product of these two numbers did not overflow a
1244        32-bit integer. However, it turns out that people want to use groups that
1245        are longer than 30,000 bytes (though not repeat them that many times).
1246        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1247        made it possible to implement the integer overflow checks in a much more
1248        dynamic way, which I have now done. The artificial limitation on group
1249        length has been removed - we now have only the limit on the total length of
1250        the compiled pattern, which depends on the LINK_SIZE setting.
1252    10. Fixed a bug in the documentation for get/copy named substring when
1253        duplicate names are permitted. If none of the named substrings are set, the
1254        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1255        empty string.
1257    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1258        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1259        because the ] is interpreted as the first data character and the
1260        terminating ] is not found. PCRE has been made compatible with Perl in this
1261        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1262        cause memory overwriting.
1264    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1265        string has been matched (to stop an infinite loop). It was not recognizing
1266        a conditional subpattern that could match an empty string if that
1267        subpattern was within another subpattern. For example, it looped when
1268        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1269        condition was not nested. This bug has been fixed.
1271    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1272        past the start of the subject in the presence of bytes with the top bit
1273        set, for example "\x8aBCD".
1275    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1276        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1278    14. Optimized (?!) to (*FAIL).
1280    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1281        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1282        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1283        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1284        does: it's just the validity check that is more restrictive.
1286    16. Inserted checks for integer overflows during escape sequence (backslash)
1287        processing, and also fixed erroneous offset values for syntax errors during
1288        backslash processing.
1290    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1291        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1293    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1294        caused an overrun.
1296    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1297        something other than just ASCII characters) inside a group that had an
1298        unlimited repeat caused a loop at compile time (while checking to see
1299        whether the group could match an empty string).
1301    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1302        [\P{Any}] did so. (Error in the code for printing property names.)
1304    21. An orphan \E inside a character class could cause a crash.
1306    22. A repeated capturing bracket such as (A)? could cause a wild memory
1307        reference during compilation.
1309    23. There are several functions in pcre_compile() that scan along a compiled
1310        expression for various reasons (e.g. to see if it's fixed length for look
1311        behind). There were bugs in these functions when a repeated \p or \P was
1312        present in the pattern. These operators have additional parameters compared
1313        with \d, etc, and these were not being taken into account when moving along
1314        the compiled data. Specifically:
1316        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1317            length.
1319        (b) An item such as \pL+ within a repeated group could cause crashes or
1320            loops.
1322        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1323            "reference to non-existent subpattern" error.
1325        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1327    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1328        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1330    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1332    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1333        character were causing crashes (broken optimization).
1335    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1336        \p or \P) caused a compile-time loop.
1338    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1339        break. For example, the unstudied pattern [\r\n]A does not match the string
1340        "\r\nA" because change 7.0/46 below moves the current point on by two
1341        characters after failing to match at the start. However, the pattern \nA
1342        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1343        the same is true. There doesn't seem any very clean way out of this, but
1344        what I have chosen to do makes the common cases work: PCRE now takes note
1345        of whether there can be an explicit match for \r or \n anywhere in the
1346        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1347        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1348        pattern has explicit CR or LF references.
1350    29. Added (*CR) etc for changing newline setting at start of pattern.
1353    Version 7.2 19-Jun-07
1354    ---------------------
1356     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1357        which is apparently normally available under Windows.
1359     2. Re-jig the pcregrep tests with different newline settings in an attempt
1360        to make them independent of the local environment's newline setting.
1362     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1364     4. Some of the "internals" tests were previously cut out when the link size
1365        was not 2, because the output contained actual offsets. The recent new
1366        "Z" feature of pcretest means that these can be cut out, making the tests
1367        usable with all link sizes.
1369     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1370        stack recursion. This gives a massive performance boost under BSD, but just
1371        a small improvement under Linux. However, it saves one field in the frame
1372        in all cases.
1374     6. Added more features from the forthcoming Perl 5.10:
1376        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1377            recursion call. It refers to the nth most recently opened parentheses.
1379        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1380            to be opened parentheses.
1382        (c) Conditions that refer to capturing parentheses can be specified
1383            relatively, for example, (?(-2)... or (?(+3)...
1385        (d) \K resets the start of the current match so that everything before
1386            is not part of it.
1388        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1390        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1391            reference syntax.
1393        (g) (?| introduces a group in which the numbering of parentheses in each
1394            alternative starts with the same number.
1396        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1398     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1401     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1402        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1403        for detecting groups that can match an empty string.
1405     9. A pattern with a very large number of alternatives (more than several
1406        hundred) was running out of internal workspace during the pre-compile
1407        phase, where pcre_compile() figures out how much memory will be needed. A
1408        bit of new cunning has reduced the workspace needed for groups with
1409        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1410        workspace instead of running out of the 4096 that are available.
1412    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1414    11. Applied patch from Google to remove an optimization that didn't quite work.
1415        The report of the bug said:
1417          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1418          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1419          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1421    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1422        it matched the wrong number of bytes.
1425    Version 7.1 24-Apr-07
1426    ---------------------
1428     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1429        that is more "standard", making use of automake and other Autotools. There
1430        is some re-arrangement of the files and adjustment of comments consequent
1431        on this.
1433     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1434        for recursive directory scanning broke on some systems because the files
1435        are not scanned in any specific order and on different systems the order
1436        was different. A call to "sort" has been inserted into RunGrepTest for the
1437        approprate test as a short-term fix. In the longer term there may be an
1438        alternative.
1440     3. I had an email from Eric Raymond about problems translating some of PCRE's
1441        man pages to HTML (despite the fact that I distribute HTML pages, some
1442        people do their own conversions for various reasons). The problems
1443        concerned the use of low-level troff macros .br and .in. I have therefore
1444        removed all such uses from the man pages (some were redundant, some could
1445        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1446        HTML has been updated to handle .nf/.fi and to complain if it encounters
1447        .br or .in.
1449     4. Updated comments in configure.ac that get placed in config.h.in and also
1450        arranged for config.h to be included in the distribution, with the name
1451        config.h.generic, for the benefit of those who have to compile without
1452        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1454     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1455        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1456        makevp.bat for the current PCRE, using the additional files
1457        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1459     6. A Windows user reported a minor discrepancy with test 2, which turned out
1460        to be caused by a trailing space on an input line that had got lost in his
1461        copy. The trailing space was an accident, so I've just removed it.
1463     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1464        that is needed.
1466     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1467        as "const" (a) because they are and (b) because it helps the PHP
1468        maintainers who have recently made a script to detect big data structures
1469        in the php code that should be moved to the .rodata section. I remembered
1470        to update Builducptable as well, so it won't revert if ucptable.h is ever
1471        re-created.
1473     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1474        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1475        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1476        support is not required. This saves 1.5-2K of code, which is important in
1477        some applications.
1479        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1480        so as not to refer to the tables, even though these functions will never be
1481        called when UTF-8 support is disabled. Otherwise there are problems with a
1482        shared library.
1484    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1486        (a) It was defining its arguments as char * instead of void *.
1488        (b) It was assuming that all moves were upwards in memory; this was true
1489            a long time ago when I wrote it, but is no longer the case.
1491        The emulated memove() is provided for those environments that have neither
1492        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1493        is clearly not the case, as these two bugs were recently reported.
1495    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1496        and Detrail to create the HTML documentation, the .txt form of the man
1497        pages, and it removes trailing spaces from listed files. It also creates
1498        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1499        case, it wraps all the #defines with #ifndefs. This script should be run
1500        before "make dist".
1502    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1503        with Unicode property support.
1505        (a) For a maximizing quantifier, if the two different cases of the
1506            character were of different lengths in their UTF-8 codings (there are
1507            some cases like this - I found 11), and the matching function had to
1508            back up over a mixture of the two cases, it incorrectly assumed they
1509            were both the same length.
1511        (b) When PCRE was configured to use the heap rather than the stack for
1512            recursion during matching, it was not correctly preserving the data for
1513            the other case of a UTF-8 character when checking ahead for a match
1514            while processing a minimizing repeat. If the check also involved
1515            matching a wide character, but failed, corruption could cause an
1516            erroneous result when trying to check for a repeat of the original
1517            character.
1519    13. Some tidying changes to the testing mechanism:
1521        (a) The RunTest script now detects the internal link size and whether there
1522            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1523            values substituted by "configure". (The RunGrepTest script already did
1524            this for UTF-8.) The configure.ac script no longer substitutes the
1525            relevant variables.
1527        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1528            with length and offset values. This means that the output is different
1529            for different internal link sizes. Test 2 is skipped for link sizes
1530            other than 2 because of this, bypassing the problem. Unfortunately,
1531            there was also a test in test 3 (the locale tests) that used /B and
1532            failed for link sizes other than 2. Rather than cut the whole test out,
1533            I have added a new /Z option to pcretest that replaces the length and
1534            offset values with spaces. This is now used to make test 3 independent
1535            of link size. (Test 2 will be tidied up later.)
1537    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1538        segmentation fault instead of returning the appropriate error message.
1540    15. In multiline mode when the newline sequence was set to "any", the pattern
1541        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1542        This doesn't seem right; it now treats the CRLF combination as the line
1543        ending, and so does not match in that case. It's only a pattern such as ^$
1544        that would hit this one: something like ^ABC$ would have failed after \r
1545        and then tried again after \r\n.
1547    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1548        in an attempt to make files that differ only in their line terminators
1549        compare equal. This works on Linux.
1551    17. Under certain error circumstances pcregrep might try to free random memory
1552        as it exited. This is now fixed, thanks to valgrind.
1554    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1555        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1556        was because its rules for how to advance for /g after matching an empty
1557        string at the end of a line did not allow for this case. They now check for
1558        it specially.
1560    20. pcretest is supposed to handle patterns and data of any length, by
1561        extending its buffers when necessary. It was getting this wrong when the
1562        buffer for a data line had to be extended.
1564    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1565        CRLF as a newline sequence.
1567    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1568        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1569        I have nevertheless tidied it up.
1571    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1573    24. Added a man page for pcre-config.
1576    Version 7.0 19-Dec-06
1577    ---------------------
1579     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1580        moving to gcc 4.1.1.
1582     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1583        sys/time.h, which is documented as needed for this function. It doesn't
1584        seem to matter on Linux, but it showed up on some releases of OS X.
1586     3. It seems that there are systems where bytes whose values are greater than
1587        127 match isprint() in the "C" locale. The "C" locale should be the
1588        default when a C program starts up. In most systems, only ASCII printing
1589        characters match isprint(). This difference caused the output from pcretest
1590        to vary, making some of the tests fail. I have changed pcretest so that:
1592        (a) When it is outputting text in the compiled version of a pattern, bytes
1593            other than 32-126 are always shown as hex escapes.
1595        (b) When it is outputting text that is a matched part of a subject string,
1596            it does the same, unless a different locale has been set for the match
1597            (using the /L modifier). In this case, it uses isprint() to decide.
1599     4. Fixed a major bug that caused incorrect computation of the amount of memory
1600        required for a compiled pattern when options that changed within the
1601        pattern affected the logic of the preliminary scan that determines the
1602        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1603        that the computed length was too small. The symptoms of this bug were
1604        either the PCRE error "internal error: code overflow" from pcre_compile(),
1605        or a glibc crash with a message such as "pcretest: free(): invalid next
1606        size (fast)". Examples of patterns that provoked this bug (shown in
1607        pcretest format) are:
1609          /(?-x: )/x
1610          /(?x)(?-x: \s*#\s*)/
1611          /((?i)[\x{c0}])/8
1612          /(?i:[\x{c0}])/8
1614        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1615        is now done differently.
1617     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1618        wrapper classes; (b) implement a new function in the C++ scanner that is
1619        more efficient than the old way of doing things because it avoids levels of
1620        recursion in the regex matching; (c) add a paragraph to the documentation
1621        for the FullMatch() function.
1623     6. The escape sequence \n was being treated as whatever was defined as
1624        "newline". Not only was this contrary to the documentation, which states
1625        that \n is character 10 (hex 0A), but it also went horribly wrong when
1626        "newline" was defined as CRLF. This has been fixed.
1628     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1629        was being set to -1 for the "end of line" case (supposedly a value that no
1630        character can have). Though this value is never used (the check for end of
1631        line is "zero bytes in current character"), it caused compiler complaints.
1632        I've changed it to 0xffffffff.
1634     8. In pcre_version.c, the version string was being built by a sequence of
1635        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1636        string (as it is for production releases) called a macro with an empty
1637        argument. The C standard says the result of this is undefined. The gcc
1638        compiler treats it as an empty string (which was what was wanted) but it is
1639        reported that Visual C gives an error. The source has been hacked around to
1640        avoid this problem.
1642     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1643        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1644        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1645        of them did).
1647    10. Originally, pcretest opened its input and output without "b"; then I was
1648        told that "b" was needed in some environments, so it was added for release
1649        5.0 to both the input and output. (It makes no difference on Unix-like
1650        systems.) Later I was told that it is wrong for the input on Windows. I've
1651        now abstracted the modes into two macros, to make it easier to fiddle with
1652        them, and removed "b" from the input mode under Windows.
1654    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1656    12. Added -help and --help to pcretest as an official way of being reminded
1657        of the options.
1659    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1660        and pcrecpp.cc because they annoy compilers at high warning levels.
1662    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1664    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1665        scripts are not C programs :-) and which was not noticed because it works
1666        on Linux.
1668    16. pcretest is supposed to handle any length of pattern and data line (as one
1669        line or as a continued sequence of lines) by extending its input buffer if
1670        necessary. This feature was broken for very long pattern lines, leading to
1671        a string of junk being passed to pcre_compile() if the pattern was longer
1672        than about 50K.
1674    17. I have done a major re-factoring of the way pcre_compile() computes the
1675        amount of memory needed for a compiled pattern. Previously, there was code
1676        that made a preliminary scan of the pattern in order to do this. That was
1677        OK when PCRE was new, but as the facilities have expanded, it has become
1678        harder and harder to keep it in step with the real compile phase, and there
1679        have been a number of bugs (see for example, 4 above). I have now found a
1680        cunning way of running the real compile function in a "fake" mode that
1681        enables it to compute how much memory it would need, while actually only
1682        ever using a few hundred bytes of working memory and without too many
1683        tests of the mode. This should make future maintenance and development
1684        easier. A side effect of this work is that the limit of 200 on the nesting
1685        depth of parentheses has been removed (though this was never a serious
1686        limitation, I suspect). However, there is a downside: pcre_compile() now
1687        runs more slowly than before (30% or more, depending on the pattern). I
1688        hope this isn't a big issue. There is no effect on runtime performance.
1690    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1691        newline (only possible for the last line of a file) and it was a
1692        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1694    19. Added additional timing features to pcretest. (1) The -tm option now times
1695        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1696        separate command line item, by a number that specifies the number of
1697        repeats to use when timing. The default is 50000; this gives better
1698        precision, but takes uncomfortably long for very large patterns.
1700    20. Extended pcre_study() to be more clever in cases where a branch of a
1701        subpattern has no definite first character. For example, (a*|b*)[cd] would
1702        previously give no result from pcre_study(). Now it recognizes that the
1703        first character must be a, b, c, or d.
1705    21. There was an incorrect error "recursive call could loop indefinitely" if
1706        a subpattern (or the entire pattern) that was being tested for matching an
1707        empty string contained only one non-empty item after a nested subpattern.
1708        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1709        incorrectly, because the \d was being skipped in the check.
1711    22. The pcretest program now has a new pattern option /B and a command line
1712        option -b, which is equivalent to adding /B to every pattern. This causes
1713        it to show the compiled bytecode, without the additional information that
1714        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1715        is the same as /B/I).
1717    23. A new optimization is now able automatically to treat some sequences such
1718        as a*b as a*+b. More specifically, if something simple (such as a character
1719        or a simple class like \d) has an unlimited quantifier, and is followed by
1720        something that cannot possibly match the quantified thing, the quantifier
1721        is automatically "possessified".
1723    24. A recursive reference to a subpattern whose number was greater than 39
1724        went wrong under certain circumstances in UTF-8 mode. This bug could also
1725        have affected the operation of pcre_study().
1727    25. Realized that a little bit of performance could be had by replacing
1728        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1730    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1732    27. Possessive quantifiers such as a++ were previously implemented by turning
1733        them into atomic groups such as ($>a+). Now they have their own opcodes,
1734        which improves performance. This includes the automatically created ones
1735        from 23 above.
1737    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1738        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1739        the first matched character to be a colon. This applied both to named and
1740        numbered groups.
1742    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1744    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1745        building PCRE on Windows easier, so I have included it in the distribution.
1747    31. There is now a check in pcretest against a ridiculously large number being
1748        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1749        loop, the loop is abandoned.
1751    32. Forward references to subpatterns in conditions such as (?(2)...) where
1752        subpattern 2 is defined later cause pcre_compile() to search forwards in
1753        the pattern for the relevant set of parentheses. This search went wrong
1754        when there were unescaped parentheses in a character class, parentheses
1755        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1757    33. "Subroutine" calls and backreferences were previously restricted to
1758        referencing subpatterns earlier in the regex. This restriction has now
1759        been removed.
1761    34. Added a number of extra features that are going to be in Perl 5.10. On the
1762        whole, these are just syntactic alternatives for features that PCRE had
1763        previously implemented using the Python syntax or my own invention. The
1764        other formats are all retained for compatibility.
1766        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1767            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1768            also .NET compatible.
1770        (b) A recursion or subroutine call to a named group can now be defined as
1771            (?&name) as well as (?P>name).
1773        (c) A backreference to a named group can now be defined as \k<name> or
1774            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1775            5.10, are also .NET compatible.
1777        (d) A conditional reference to a named group can now use the syntax
1778            (?(<name>) or (?('name') as well as (?(name).
1780        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1781            groups (named and numbered) that are never evaluated inline, but can be
1782            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1783            is always false. There may be only one alternative in such a group.
1785        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1786            as the simple (?(R). The condition is true only if the most recent
1787            recursion is that of the given number or name. It does not search out
1788            through the entire recursion stack.
1790        (g) The escape \gN or \g{N} has been added, where N is a positive or
1791            negative number, specifying an absolute or relative reference.
1793    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1794        some "unreachable code" warnings.
1796    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1797        things, this adds five new scripts.
1799    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1800        There were also incompatibilities regarding the handling of \Q..\E inside
1801        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1802        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1804    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1805        matches an empty string, and forcibly breaks the loop. There were bugs in
1806        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1807        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1808        separate and independent bugs (that affected different cases) have been
1809        fixed.
1811    39. Refactored the code to abolish the use of different opcodes for small
1812        capturing bracket numbers. This is a tidy that I avoided doing when I
1813        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1814        The new approach is not only tidier, it makes it possible to reduce the
1815        memory needed to fix the previous bug (38).
1817    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1818        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1819        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1820        mode.
1822    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1823        report.
1825    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1826        copy construction and assignment in the C++ wrapper.
1828    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1829        couple of bugs that could have given wrong results in the "--newline=crlf"
1830        case.
1832    44. Added a number of casts and did some reorganization of signed/unsigned int
1833        variables following suggestions from Dair Grant. Also renamed the variable
1834        "this" as "item" because it is a C++ keyword.
1836    45. Arranged for dftables to add
1838          #include "pcre_internal.h"
1840        to pcre_chartables.c because without it, gcc 4.x may remove the array
1841        definition from the final binary if PCRE is built into a static library and
1842        dead code stripping is activated.
1844    46. For an unanchored pattern, if a match attempt fails at the start of a
1845        newline sequence, and the newline setting is CRLF or ANY, and the next two
1846        characters are CRLF, advance by two characters instead of one.
1849    Version 6.7 04-Jul-06
1850    ---------------------
1852     1. In order to handle tests when input lines are enormously long, pcretest has
1853        been re-factored so that it automatically extends its buffers when
1854        necessary. The code is crude, but this _is_ just a test program. The
1855        default size has been increased from 32K to 50K.
1857     2. The code in pcre_study() was using the value of the re argument before
1858        testing it for NULL. (Of course, in any sensible call of the function, it
1859        won't be NULL.)
1861     3. The memmove() emulation function in pcre_internal.h, which is used on
1862        systems that lack both memmove() and bcopy() - that is, hardly ever -
1863        was missing a "static" storage class specifier.
1865     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1866        containing an extended class (one that cannot be represented by a bitmap
1867        because it contains high-valued characters or Unicode property items, e.g.
1868        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1869        pattern, but PCRE should not loop if you do not (it no longer does).
1870        [Detail: two cases were found: (a) a repeated subpattern containing an
1871        extended class; (b) a recursive reference to a subpattern that followed a
1872        previous extended class. It wasn't skipping over the extended class
1873        correctly when UTF-8 mode was not set.]
1875     5. A negated single-character class was not being recognized as fixed-length
1876        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1877        compile error "lookbehind assertion is not fixed length".
1879     6. The RunPerlTest auxiliary script was showing an unexpected difference
1880        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1881        write a Perl script that can interpret lines of an input file either as
1882        byte characters or as UTF-8, which is what "perltest" was being required to
1883        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1884        can't do is switch easily at run time between having the "use utf8;" pragma
1885        or not. In the end, I fudged it by using the RunPerlTest script to insert
1886        "use utf8;" explicitly for the UTF-8 tests.
1888     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1889        the end of the subject string, contrary to the documentation and to what
1890        Perl does. This was true of both matching functions. Now it matches only at
1891        the start of the subject and immediately after *internal* newlines.
1893     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1894        a pointer to an int instead of a pointer to an unsigned long int. This
1895        caused problems on 64-bit systems.
1897     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1898        instance of the 'standard' template library not being so standard".
1900    10. There was no check on the number of named subpatterns nor the maximum
1901        length of a subpattern name. The product of these values is used to compute
1902        the size of the memory block for a compiled pattern. By supplying a very
1903        long subpattern name and a large number of named subpatterns, the size
1904        computation could be caused to overflow. This is now prevented by limiting
1905        the length of names to 32 characters, and the number of named subpatterns
1906        to 10,000.
1908    11. Subpatterns that are repeated with specific counts have to be replicated in
1909        the compiled pattern. The size of memory for this was computed from the
1910        length of the subpattern and the repeat count. The latter is limited to
1911        65535, but there was no limit on the former, meaning that integer overflow
1912        could in principle occur. The compiled length of a repeated subpattern is
1913        now limited to 30,000 bytes in order to prevent this.
1915    12. Added the optional facility to have named substrings with the same name.
1917    13. Added the ability to use a named substring as a condition, using the
1918        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1919        are numbers (not recommended). Forward references are permitted.
1921    14. Added forward references in named backreferences (if you see what I mean).
1923    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1924        pattern could run off the end of the subject. For example, the pattern
1925        "(?s)(.{1,5})"8 did this with the subject "ab".
1927    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1928        PCRE_CASELESS was set when matching characters that were quantified with ?
1929        or *.
1931    17. A character class other than a single negated character that had a minimum
1932        but no maximum quantifier - for example [ab]{6,} - was not handled
1933        correctly by pce_dfa_exec(). It would match only one character.
1935    18. A valid (though odd) pattern that looked like a POSIX character
1936        class but used an invalid character after [ (for example [[,abc,]]) caused
1937        pcre_compile() to give the error "Failed: internal error: code overflow" or
1938        in some cases to crash with a glibc free() error. This could even happen if
1939        the pattern terminated after [[ but there just happened to be a sequence of
1940        letters, a binary zero, and a closing ] in the memory that followed.
1942    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1943        over the years. Originally (before any Unicode support), just the bottom 8
1944        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1945        output from "man perlunicode" includes this:
1947          The regular expression compiler produces polymorphic opcodes.  That
1948          is, the pattern adapts to the data and automatically switches to
1949          the Unicode character scheme when presented with Unicode data--or
1950          instead uses a traditional byte scheme when presented with byte
1951          data.
1953        Sadly, a wide octal escape does not cause a switch, and in a string with
1954        no other multibyte characters, these octal escapes are treated as before.
1955        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1956        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1957        Unicode string.
1959        I have not perpetrated such confusion in PCRE. Up till now, it took just
1960        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1961        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1962        translate to the appropriate multibyte character.
1964    29. Applied some refactoring to reduce the number of warnings from Microsoft
1965        and Borland compilers. This has included removing the fudge introduced
1966        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1967        a warning about an unused variable.
1969    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1970        characters since release 4.0, because Perl (from release 5.004) does not.
1971        [Or at least, is documented not to: some releases seem to be in conflict
1972        with the documentation.] However, when a pattern was studied with
1973        pcre_study() and all its branches started with \s, PCRE still included VT
1974        as a possible starting character. Of course, this did no harm; it just
1975        caused an unnecessary match attempt.
1977    22. Removed a now-redundant internal flag bit that recorded the fact that case
1978        dependency changed within the pattern. This was once needed for "required
1979        byte" processing, but is no longer used. This recovers a now-scarce options
1980        bit. Also moved the least significant internal flag bit to the most-
1981        significant bit of the word, which was not previously used (hangover from
1982        the days when it was an int rather than a uint) to free up another bit for
1983        the future.
1985    23. Added support for CRLF line endings as well as CR and LF. As well as the
1986        default being selectable at build time, it can now be changed at runtime
1987        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1988        specify that it is scanning data with non-default line endings.
1990    24. Changed the definition of CXXLINK to make it agree with the definition of
1991        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1993    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1994        recursions. This makes a big different to stack usage for some patterns.
1996    26. If a subpattern containing a named recursion or subroutine reference such
1997        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1998        the space required for the compiled pattern went wrong and gave too small a
1999        value. Depending on the environment, this could lead to "Failed: internal
2000        error: code overflow at offset 49" or "glibc detected double free or
2001        corruption" errors.
2003    27. Applied patches from Google (a) to support the new newline modes and (b) to
2004        advance over multibyte UTF-8 characters in GlobalReplace.
2006    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2007        difference for some implementation of PCRE in some Windows version.
2009    29. Added some extra testing facilities to pcretest:
2011        \q<number>   in a data line sets the "match limit" value
2012        \Q<number>   in a data line sets the "match recursion limt" value
2013        -S <number>  sets the stack size, where <number> is in megabytes
2015        The -S option isn't available for Windows.
2018    Version 6.6 06-Feb-06
2019    ---------------------
2021     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2022        in pcreposix.h. I have copied the definition from pcre.h.
2024     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2025        because pcre.h is no longer a built file.
2027     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2028        not normally included in the compiled code.
2031    Version 6.5 01-Feb-06
2032    ---------------------
2034     1. When using the partial match feature with pcre_dfa_exec(), it was not
2035        anchoring the second and subsequent partial matches at the new starting
2036        point. This could lead to incorrect results. For example, with the pattern
2037        /1234/, partially matching against "123" and then "a4" gave a match.
2039     2. Changes to pcregrep:
2041        (a) All non-match returns from pcre_exec() were being treated as failures
2042            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2043            error message is output. Some extra information is given for the
2044            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2045            probably the only errors that are likely to be caused by users (by
2046            specifying a regex that has nested indefinite repeats, for instance).
2047            If there are more than 20 of these errors, pcregrep is abandoned.
2049        (b) A binary zero was treated as data while matching, but terminated the
2050            output line if it was written out. This has been fixed: binary zeroes
2051            are now no different to any other data bytes.
2053        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2054            used to set a locale for matching. The --locale=xxxx long option has
2055            been added (no short equivalent) to specify a locale explicitly on the
2056            pcregrep command, overriding the environment variables.
2058        (d) When -B was used with -n, some line numbers in the output were one less
2059            than they should have been.
2061        (e) Added the -o (--only-matching) option.
2063        (f) If -A or -C was used with -c (count only), some lines of context were
2064            accidentally printed for the final match.
2066        (g) Added the -H (--with-filename) option.
2068        (h) The combination of options -rh failed to suppress file names for files
2069            that were found from directory arguments.
2071        (i) Added the -D (--devices) and -d (--directories) options.
2073        (j) Added the -F (--fixed-strings) option.
2075        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2077        (l) Added the --colo(u)r option.
2079        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2080            is not present by default.
2082     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2083        items such as (?R) or (?1), when the recursion could match a number of
2084        alternatives. If it matched one of the alternatives, but subsequently,
2085        outside the recursion, there was a failure, the code tried to back up into
2086        the recursion. However, because of the way PCRE is implemented, this is not
2087        possible, and the result was an incorrect result from the match.
2089        In order to prevent this happening, the specification of recursion has
2090        been changed so that all such subpatterns are automatically treated as
2091        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2093     4. I had overlooked the fact that, in some locales, there are characters for
2094        which isalpha() is true but neither isupper() nor islower() are true. In
2095        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2096        and ordfeminine) are like this. This affected the treatment of \w and \W
2097        when they appeared in character classes, but not when they appeared outside
2098        a character class. The bit map for "word" characters is now created
2099        separately from the results of isalnum() instead of just taking it from the
2100        upper, lower, and digit maps. (Plus the underscore character, of course.)
2102     5. The above bug also affected the handling of POSIX character classes such as
2103        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2104        permanent tables. Instead, the bit maps for such a class were previously
2105        created as the appropriate unions of the upper, lower, and digit bitmaps.
2106        Now they are created by subtraction from the [[:word:]] class, which has
2107        its own bitmap.
2109     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2110        It is created by subtracting the vertical space characters (\x09, \x0a,
2111        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2112        subtraction was done in the overall bitmap for a character class, meaning
2113        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2114        be recognized. This bug has been fixed.
2116     7. Patches from the folks at Google:
2118          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2119          real life, but is still worth protecting against".
2121          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2122          regular expressions".
2124          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2125          have it.
2127          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2128          "configure" and the latter not, in order to fix a problem somebody had
2129          with compiling the Arg class on HP-UX.
2131          (e) Improve the error-handling of the C++ wrapper a little bit.
2133          (f) New tests for checking recursion limiting.
2135     8. The pcre_memmove() function, which is used only if the environment does not
2136        have a standard memmove() function (and is therefore rarely compiled),
2137        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2138        returning a result (though PCRE never actually uses the result).
2140     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2141        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2142        returned instead of calling malloc() with an overflowing number that would
2143        most likely cause subsequent chaos.
2145    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2147    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2148        with this option is matched, the nmatch and pmatch options of regexec() are
2149        ignored.
2151    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2152        provided in case anyone wants to the the POSIX interface with UTF-8
2153        strings.
2155    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2156        C++ linking (needed for some HP-UX environments).
2158    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2159        (unused parameter) and in the pcre_printint() function (omitted "default"
2160        switch label when the default is to do nothing).
2162    15. Added some code to make it possible, when PCRE is compiled as a C++
2163        library, to replace subject pointers for pcre_exec() with a smart pointer
2164        class, thus making it possible to process discontinuous strings.
2166    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2167        much the same function. They were added by different people who were trying
2168        to make PCRE easy to compile on non-Unix systems. It has been suggested
2169        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2170        for compiling on Windows systems. I have therefore replaced it with
2171        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2172        defaults to "extern" for C or "extern C" for C++, which works fine on
2173        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2174        SCOPE with something explicit in config.h. In addition:
2176        (a) pcreposix.h still had just "extern" instead of either of these macros;
2177            I have replaced it with PCRE_DATA_SCOPE.
2179        (b) Functions such as _pcre_xclass(), which are internal to the library,
2180            but external in the C sense, all had PCRE_EXPORT in their definitions.
2181            This is apparently wrong for the Windows case, so I have removed it.
2182            (It makes no difference on Unix-like systems.)
2184    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2185        of recursive calls to match(). This is different to MATCH_LIMIT because
2186        that limits the total number of calls to match(), not all of which increase
2187        the depth of recursion. Limiting the recursion depth limits the amount of
2188        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2189        when PCRE is compiled, and changed at run time. A patch from Google adds
2190        this functionality to the C++ interface.
2192    18. Changes to the handling of Unicode character properties:
2194        (a) Updated the table to Unicode 4.1.0.
2196        (b) Recognize characters that are not in the table as "Cn" (undefined).
2198        (c) I revised the way the table is implemented to a much improved format
2199            which includes recognition of ranges. It now supports the ranges that
2200            are defined in UnicodeData.txt, and it also amalgamates other
2201            characters into ranges. This has reduced the number of entries in the
2202            table from around 16,000 to around 3,000, thus reducing its size
2203            considerably. I realized I did not need to use a tree structure after
2204            all - a binary chop search is just as efficient. Having reduced the
2205            number of entries, I extended their size from 6 bytes to 8 bytes to
2206            allow for more data.
2208        (d) Added support for Unicode script names via properties such as \p{Han}.
2210    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2211        matching that character.
2213    20. When matching a repeated Unicode property with a minimum greater than zero,
2214        (for example \pL{2,}), PCRE could look past the end of the subject if it
2215        reached it while seeking the minimum number of characters. This could
2216        happen only if some of the characters were more than one byte long, because
2217        there is a check for at least the minimum number of bytes.
2219    21. Refactored the implementation of \p and \P so as to be more general, to
2220        allow for more different types of property in future. This has changed the
2221        compiled form incompatibly. Anybody with saved compiled patterns that use
2222        \p or \P will have to recompile them.
2224    22. Added "Any" and "L&" to the supported property types.
2226    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2227        but give a compile time error if the value is greater than 0xff.
2229    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2230        accidentally not being installed or uninstalled.
2232    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2233        made were to insert the current release number. This seemed silly, because
2234        it made things harder for people building PCRE on systems that don't run
2235        "configure". I have turned pcre.h into a distributed file, no longer built
2236        by "configure", with the version identification directly included. There is
2237        no longer a pcre.h.in file.
2239        However, this change necessitated a change to the pcre-config script as
2240        well. It is built from pcre-config.in, and one of the substitutions was the
2241        release number. I have updated configure.ac so that ./configure now finds
2242        the release number by grepping pcre.h.
2244    26. Added the ability to run the tests under valgrind.
2247    Version 6.4 05-Sep-05
2248    ---------------------
2250     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2251        "--" to be printed when multiple files were scanned, even when none of the
2252        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2253        consider it to be a bug, and have restored the previous behaviour.
2255     2. A couple of code tidies to get rid of compiler warnings.
2257     3. The pcretest program used to cheat by referring to symbols in the library
2258        whose names begin with _pcre_. These are internal symbols that are not
2259        really supposed to be visible externally, and in some environments it is
2260        possible to suppress them. The cheating is now confined to including
2261        certain files from the library's source, which is a bit cleaner.
2263     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2264        file's purpose clearer.
2266     5. Reorganized pcre_ucp_findchar().
2269    Version 6.3 15-Aug-05
2270    ---------------------
2272     1. The file libpcre.pc.in did not have general read permission in the tarball.
2274     2. There were some problems when building without C++ support:
2276        (a) If C++ support was not built, "make install" and "make test" still
2277            tried to test it.
2279        (b) There were problems when the value of CXX was explicitly set. Some
2280            changes have been made to try to fix these, and ...
2282        (c) --disable-cpp can now be used to explicitly disable C++ support.
2284        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2285            backslash in a target when C++ was disabled. This confuses some
2286            versions of "make", apparently. Using an intermediate variable solves
2287            this. (Same for CPP_LOBJ.)
2289     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2290        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2291        necessary on certain architectures.
2293     4. Added a setting of -export-symbols-regex to the link command to remove
2294        those symbols that are exported in the C sense, but actually are local
2295        within the library, and not documented. Their names all begin with
2296        "_pcre_". This is not a perfect job, because (a) we have to except some
2297        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2298        available (and never for static libraries). I have made a note to try to
2299        find a way round (a) in the future.
2302    Version 6.2 01-Aug-05
2303    ---------------------
2305     1. There was no test for integer overflow of quantifier values. A construction
2306        such as {1111111111111111} would give undefined results. What is worse, if
2307        a minimum quantifier for a parenthesized subpattern overflowed and became
2308        negative, the calculation of the memory size went wrong. This could have
2309        led to memory overwriting.
2311     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2313     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2314        operating environments where this matters.
2316     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2317        PCRE options from within the C++ wrapper.
2319     5. Named capturing subpatterns were not being correctly counted when a pattern
2320        was compiled. This caused two problems: (a) If there were more than 100
2321        such subpatterns, the calculation of the memory needed for the whole
2322        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2323        back references of the form \12, where the number was greater than 9, were
2324        not recognized as back references, even though there were sufficient
2325        previous subpatterns.
2327     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2328        versions of gcc, e.g. 2.95.4.
2331    Version 6.1 21-Jun-05
2332    ---------------------
2334     1. There was one reference to the variable "posix" in pcretest.c that was not
2335        surrounded by "#if !defined NOPOSIX".
2337     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2338        the cross-check on the old pcre_info() function, for the benefit of the
2339        cut-down version of PCRE that is currently imported into Exim.
2341     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2342        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2343        patterns that start (?i) but I don't think that matters. The use of (?i) is
2344        just an example; this all applies to the other options as well.
2346     4. Since libtool seems to echo the compile commands it is issuing, the output
2347        from "make" can be reduced a bit by putting "@" in front of each libtool
2348        compile command.
2350     5. Patch from the folks at Google for configure.in to be a bit more thorough
2351        in checking for a suitable C++ installation before trying to compile the
2352        C++ stuff. This should fix a reported problem when a compiler was present,
2353        but no suitable headers.
2355     6. The man pages all had just "PCRE" as their title. I have changed them to
2356        be the relevant file name. I have also arranged that these names are
2357        retained in the file doc/pcre.txt, which is a concatenation in text format
2358        of all the man pages except the little individual ones for each function.
2360     7. The NON-UNIX-USE file had not been updated for the different set of source
2361        files that come with release 6. I also added a few comments about the C++
2362        wrapper.
2365    Version 6.0 07-Jun-05
2366    ---------------------
2368     1. Some minor internal re-organization to help with my DFA experiments.
2370     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2371        didn't matter for the library itself when fully configured, but did matter
2372        when compiling without UCP support, or within Exim, where the ucp files are
2373        not imported.
2375     3. Refactoring of the library code to split up the various functions into
2376        different source modules. The addition of the new DFA matching code (see
2377        below) to a single monolithic source would have made it really too
2378        unwieldy, quite apart from causing all the code to be include in a
2379        statically linked application, when only some functions are used. This is
2380        relevant even without the DFA addition now that patterns can be compiled in
2381        one application and matched in another.
2383        The downside of splitting up is that there have to be some external
2384        functions and data tables that are used internally in different modules of
2385        the library but which are not part of the API. These have all had their
2386        names changed to start with "_pcre_" so that they are unlikely to clash
2387        with other external names.
2389     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2390        a different (DFA) algorithm. Although it is slower than the original
2391        function, it does have some advantages for certain types of matching
2392        problem.
2394     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2395        including restarting after a partial match.
2397     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2398        defined when compiling for Windows was sent to me. I have put it into the
2399        code, though I have no means of testing or verifying it.
2401     7. Added the pcre_refcount() auxiliary function.
2403     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2404        match before or at the first newline in the subject string. In pcretest,
2405        the /f option on a pattern can be used to set this.
2407     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2408        would behave wrongly. This has been present in PCRE since release 4.0.
2410    10. A number of changes to the pcregrep command:
2412        (a) Refactored how -x works; insert ^(...)$ instead of setting
2413            PCRE_ANCHORED and checking the length, in preparation for adding
2414            something similar for -w.
2416        (b) Added the -w (match as a word) option.
2418        (c) Refactored the way lines are read and buffered so as to have more
2419            than one at a time available.
2421        (d) Implemented a pcregrep test script.
2423        (e) Added the -M (multiline match) option. This allows patterns to match
2424            over several lines of the subject. The buffering ensures that at least
2425            8K, or the rest of the document (whichever is the shorter) is available
2426            for matching (and similarly the previous 8K for lookbehind assertions).
2428        (f) Changed the --help output so that it now says
2430              -w, --word-regex(p)
2432            instead of two lines, one with "regex" and the other with "regexp"
2433            because that confused at least one person since the short forms are the
2434            same. (This required a bit of code, as the output is generated
2435            automatically from a table. It wasn't just a text change.)
2437        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2438            option but starts with a hyphen. Could be a pattern or a path name
2439            starting with a hyphen, for instance.
2441        (h) "-" can be given as a file name to represent stdin.
2443        (i) When file names are being printed, "(standard input)" is used for
2444            the standard input, for compatibility with GNU grep. Previously
2445            "<stdin>" was used.
2447        (j) The option --label=xxx can be used to supply a name to be used for
2448            stdin when file names are being printed. There is no short form.
2450        (k) Re-factored the options decoding logic because we are going to add
2451            two more options that take data. Such options can now be given in four
2452            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2454        (l) Added the -A, -B, and -C options for requesting that lines of context
2455            around matches be printed.
2457        (m) Added the -L option to print the names of files that do not contain
2458            any matching lines, that is, the complement of -l.
2460        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2461            continue to scan other files.
2463        (o) The -s option was incorrectly implemented. For compatibility with other
2464            greps, it now suppresses the error message for a non-existent or non-
2465            accessible file (but not the return code). There is a new option called
2466            -q that suppresses the output of matching lines, which was what -s was
2467            previously doing.
2469        (p) Added --include and --exclude options to specify files for inclusion
2470            and exclusion when recursing.
2472    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2473        Hopefully, it now does.
2475    12. Missing cast in pcre_study().
2477    13. Added an "uninstall" target to the makefile.
2479    14. Replaced "extern" in the function prototypes in Makefile.in with
2480        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2481        world, but is set differently for Windows.
2483    15. Added a second compiling function called pcre_compile2(). The only
2484        difference is that it has an extra argument, which is a pointer to an
2485        integer error code. When there is a compile-time failure, this is set
2486        non-zero, in addition to the error test pointer being set to point to an
2487        error message. The new argument may be NULL if no error number is required
2488        (but then you may as well call pcre_compile(), which is now just a
2489        wrapper). This facility is provided because some applications need a
2490        numeric error indication, but it has also enabled me to tidy up the way
2491        compile-time errors are handled in the POSIX wrapper.
2493    16. Added VPATH=.libs to the makefile; this should help when building with one
2494        prefix path and installing with another. (Or so I'm told by someone who
2495        knows more about this stuff than I do.)
2497    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2498        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2499        match everything, including newlines. This is not POSIX-compatible, but
2500        somebody wanted the feature. From pcretest it can be activated by using
2501        both the P and the s flags.
2503    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2505    19. libpcre.pc was being incorrectly installed as executable.
2507    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2508        it now also looks for '\r' so that it will work unmodified on Windows.
2510    21. Added Google's contributed C++ wrapper to the distribution.
2512    22. Added some untidy missing memory free() calls in pcretest, to keep
2513        Electric Fence happy when testing.
2517    Version 5.0 13-Sep-04
2518    ---------------------
2520     1. Internal change: literal characters are no longer packed up into items
2521        containing multiple characters in a single byte-string. Each character
2522        is now matched using a separate opcode. However, there may be more than one
2523        byte in the character in UTF-8 mode.
2525     2. The pcre_callout_block structure has two new fields: pattern_position and
2526        next_item_length. These contain the offset in the pattern to the next match
2527        item, and its length, respectively.
2529     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2530        insertion of callouts before each pattern item. Added the /C option to
2531        pcretest to make use of this.
2533     4. On the advice of a Windows user, the lines
2535          #if defined(_WIN32) || defined(WIN32)
2536          _setmode( _fileno( stdout ), 0x8000 );
2537          #endif  /* defined(_WIN32) || defined(WIN32) */
2539        have been added to the source of pcretest. This apparently does useful
2540        magic in relation to line terminators.
2542     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2543        for the benefit of those environments where the "b" makes a difference.
2545     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2546        to know about it. I have put a hack into configure.in that adds in code
2547        to set GCC=yes if CC=icc. This seems to end up at a point in the
2548        generated configure script that is early enough to affect the setting of
2549        compiler options, which is what is needed, but I have no means of testing
2550        whether it really works. (The user who reported this had patched the
2551        generated configure script, which of course I cannot do.)
2553        LATER: After change 22 below (new libtool files), the configure script
2554        seems to know about icc (and also ecc). Therefore, I have commented out
2555        this hack in configure.in.
2557     7. Added support for pkg-config (2 patches were sent in).
2559     8. Negated POSIX character classes that used a combination of internal tables
2560        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2561        [[:^ascii]]. Typically, they would match almost any characters. The other
2562        POSIX classes were not broken in this way.
2564     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2565        to find the match, as PCRE was deluded into thinking that the match had to
2566        start at the start point or following a newline. The same bug applied to
2567        patterns with negative forward assertions or any backward assertions
2568        preceding ".*" at the start, unless the pattern required a fixed first
2569        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2571    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2572        starting at the last subject character, bytes beyond the end of the subject
2573        string were read.
2575    11. Renamed the variable "class" as "classbits" to make life easier for C++
2576        users. (Previously there was a macro definition, but it apparently wasn't
2577        enough.)
2579    12. Added the new field "tables" to the extra data so that tables can be passed
2580        in at exec time, or the internal tables can be re-selected. This allows
2581        a compiled regex to be saved and re-used at a later time by a different
2582        program that might have everything at different addresses.
2584    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2585        -R library as well as a -L library.
2587    14. The debugging options of pcretest (-d on the command line or D on a
2588        pattern) showed incorrect output for anything following an extended class
2589        that contained multibyte characters and which was followed by a quantifier.
2591    15. Added optional support for general category Unicode character properties
2592        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2593        support. It adds about 90K to the size of the library. The meanings of the
2594        inbuilt class escapes such as \d and \s have NOT been changed.
2596    16. Updated pcredemo.c to include calls to free() to release the memory for the
2597        compiled pattern.
2599    17. The generated file chartables.c was being created in the source directory
2600        instead of in the building directory. This caused the build to fail if the
2601        source directory was different from the building directory, and was
2602        read-only.
2604    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2605        file. No doubt somebody will tell me if they don't make sense... Also added
2606        Dan Mooney's comments about building on OpenVMS.
2608    19. Added support for partial matching via the PCRE_PARTIAL option for
2609        pcre_exec() and the \P data escape in pcretest.
2611    20. Extended pcretest with 3 new pattern features:
2613        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2614              write the compiled pattern to the file whose name is "rest-of-line".
2615              This is a straight binary dump of the data, with the saved pointer to
2616              the character tables forced to be NULL. The study data, if any, is
2617              written too. After writing, pcretest reads a new pattern.
2619        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2620              compiled pattern from the given file. There must not be any
2621              occurrences of "<" in the file name (pretty unlikely); if there are,
2622              pcretest will instead treat the initial "<" as a pattern delimiter.
2623              After reading in the pattern, pcretest goes on to read data lines as
2624              usual.
2626        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2627              and 16-bit fields in a compiled pattern, to simulate a pattern that
2628              was compiled on a host of opposite endianness.
2630    21. The pcre-exec() function can now cope with patterns that were compiled on
2631        hosts of opposite endianness, with this restriction:
2633          As for any compiled expression that is saved and used later, the tables
2634          pointer field cannot be preserved; the extra_data field in the arguments
2635          to pcre_exec() should be used to pass in a tables address if a value
2636          other than the default internal tables were used at compile time.
2638    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2639        now diagnosed as an error. Previously, most of the time, a negative number
2640        would have been treated as zero, but if in addition "ovector" was passed as
2641        NULL, a crash could occur.
2643    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2644        new versions from the libtool 1.5 distribution (the last one is a copy of
2645        a file called libtool.m4). This seems to have fixed the need to patch
2646        "configure" to support Darwin 1.3 (which I used to do). However, I still
2647        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2648        workstation).
2650    24. Changed the PCRE licence to be the more standard "BSD" licence.
2653    Version 4.5 01-Dec-03
2654    ---------------------
2656     1. There has been some re-arrangement of the code for the match() function so
2657        that it can be compiled in a version that does not call itself recursively.
2658        Instead, it keeps those local variables that need separate instances for
2659        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2660        needs to "recurse". Keeping track of where control must go is done by means
2661        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2662        hide most of the details from the main code, and operates only if
2663        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2664        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2665        operating.
2667        To make it easier for callers to provide specially tailored get/free
2668        functions for this usage, two new functions, pcre_stack_malloc, and
2669        pcre_stack_free, are used. They are always called in strict stacking order,
2670        and the size of block requested is always the same.
2672        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2673        PCRE has been compiled to use the stack or the heap for recursion. The
2674        -C option of pcretest uses this to show which version is compiled.
2676        A new data escape \S, is added to pcretest; it causes the amounts of store
2677        obtained and freed by both kinds of malloc/free at match time to be added
2678        to the output.
2680     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2681        what's available on my current Linux desktop machine.
2683     3. When matching a UTF-8 string, the test for a valid string at the start has
2684        been extended. If start_offset is not zero, PCRE now checks that it points
2685        to a byte that is the start of a UTF-8 character. If not, it returns
2686        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2687        this is necessary because there may be backward assertions in the pattern.
2688        When matching the same subject several times, it may save resources to use
2689        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2691     4. The code for checking the validity of UTF-8 strings has been tightened so
2692        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2693        containing "overlong sequences".
2695     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2696        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2697        should have been "&", but it just so happened that all the cases this let
2698        through by mistake were picked up later in the function.
2700     6. I had used a variable called "isblank" - this is a C99 function, causing
2701        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2703     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2704        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2705        so the version is shown; (c) stop "make test" echoing "./RunTest".
2707     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2709     9. The prototype for memmove() for systems that don't have it was using
2710        size_t, but the inclusion of the header that defines size_t was later. I've
2711        moved the #includes for the C headers earlier to avoid this.
2713    10. Added some adjustments to the code to make it easier to compiler on certain
2714        special systems:
2716          (a) Some "const" qualifiers were missing.
2717          (b) Added the macro EXPORT before all exported functions; by default this
2718              is defined to be empty.
2719          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2720              that it reads its output file name as an argument instead of writing
2721              to the standard output and assuming this can be redirected.
2723    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2724        class containing characters with values greater than 255, PCRE compilation
2725        went into a loop.
2727    12. A recursive reference to a subpattern that was within another subpattern
2728        that had a minimum quantifier of zero caused PCRE to crash. For example,
2729        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2730        recursion. If the recursively-called subpattern itself had a zero repeat,
2731        that was OK.
2733    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2734        buffer into which it was copied (for escape processing) was still set at
2735        1024, so long lines caused crashes.
2737    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2738        "internal error: code overflow...". This applied to any character class
2739        that was followed by a possessive quantifier.
2741    15. Modified the Makefile to add libpcre.la as a prerequisite for
2742        libpcreposix.la because I was told this is needed for a parallel build to
2743        work.
2745    16. If a pattern that contained .* following optional items at the start was
2746        studied, the wrong optimizing data was generated, leading to matching
2747        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2748        matching string must start with a or b or c. The correct conclusion for
2749        this pattern is that a match can start with any character.
2752    Version 4.4 13-Aug-03
2753    ---------------------
2755     1. In UTF-8 mode, a character class containing characters with values between
2756        127 and 255 was not handled correctly if the compiled pattern was studied.
2757        In fixing this, I have also improved the studying algorithm for such
2758        classes (slightly).
2760     2. Three internal functions had redundant arguments passed to them. Removal
2761        might give a very teeny performance improvement.
2763     3. Documentation bug: the value of the capture_top field in a callout is *one
2764        more than* the number of the hightest numbered captured substring.
2766     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2767        in incorrectly linking with a previously installed version. They now link
2768        explicitly with libpcre.la.
2770     5. configure.in no longer needs to recognize Cygwin specially.
2772     6. A problem in pcre.in for Windows platforms is fixed.
2774     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2775        pcretest, it used to include the size of the study block as part of its
2776        output. Unfortunately, the structure contains a field that has a different
2777        size on different hardware architectures. This meant that the tests that
2778        showed this size failed. As the block is currently always of a fixed size,
2779        this information isn't actually particularly useful in pcretest output, so
2780        I have just removed it.
2782     8. Three pre-processor statements accidentally did not start in column 1.
2783        Sadly, there are *still* compilers around that complain, even though
2784        standard C has not required this for well over a decade. Sigh.
2786     9. In pcretest, the code for checking callouts passed small integers in the
2787        callout_data field, which is a void * field. However, some picky compilers
2788        complained about the casts involved for this on 64-bit systems. Now
2789        pcretest passes the address of the small integer instead, which should get
2790        rid of the warnings.
2792    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2793        both compile and run time, and gives an error if an invalid UTF-8 sequence
2794        is found. There is a option for disabling this check in cases where the
2795        string is known to be correct and/or the maximum performance is wanted.
2797    11. In response to a bug report, I changed one line in Makefile.in from
2799            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2800        to
2801            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2803        to look similar to other lines, but I have no way of telling whether this
2804        is the right thing to do, as I do not use Windows. No doubt I'll get told
2805        if it's wrong...
2808    Version 4.3 21-May-03
2809    ---------------------
2811    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2812       Makefile.
2814    2. Some refactoring to improve the quality of the code:
2816       (i)   The utf8_table... variables are now declared "const".
2818       (ii)  The code for \cx, which used the "case flipping" table to upper case
2819             lower case letters, now just substracts 32. This is ASCII-specific,
2820             but the whole concept of \cx is ASCII-specific, so it seems
2821             reasonable.
2823       (iii) PCRE was using its character types table to recognize decimal and
2824             hexadecimal digits in the pattern. This is silly, because it handles
2825             only 0-9, a-f, and A-F, but the character types table is locale-
2826             specific, which means strange things might happen. A private
2827             table is now used for this - though it costs 256 bytes, a table is
2828             much faster than multiple explicit tests. Of course, the standard
2829             character types table is still used for matching digits in subject
2830             strings against \d.
2832       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2833             ending in _t are). So I've renamed it as ESC_tee.
2835    3. The first argument for regexec() in the POSIX wrapper should have been
2836       defined as "const".
2838    4. Changed pcretest to use malloc() for its buffers so that they can be
2839       Electric Fenced for debugging.
2841    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2842       to read one or more bytes before the start of the subject string. Often this
2843       had no effect on PCRE's behaviour, but in some circumstances it could
2844       provoke a segmentation fault.
2846    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2847       to try to read one or more bytes before the start of the subject string.
2849    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2850       UTF-8 support could misbehave in various ways if the subject string
2851       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2852       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2853       back over UTF-8 characters.)
2856    Version 4.2 14-Apr-03
2857    ---------------------
2859    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2861    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2862         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2863         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2864         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2865         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2866           and BUILD_EXEEXT
2867         Note: automatic setting of the BUILD variables is not yet working
2868         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2869           compile-time but not at link-time
2870         [LINK]: use for linking executables only
2871         make different versions for Windows and non-Windows
2872         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2873           libraries
2874         [LINK_FOR_BUILD]: new variable
2875         [OBJEXT]: use throughout
2876         [EXEEXT]: use throughout
2877         <winshared>: new target
2878         <wininstall>: new target
2879         <dftables.o>: use native compiler
2880         <dftables>: use native linker
2881         <install>: handle Windows platform correctly
2882         <clean>: ditto
2883         <check>: ditto
2884         copy DLL to top builddir before testing
2886       As part of these changes, -no-undefined was removed again. This was reported
2887       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2888       in any case.
2890    3. Some tidies to get rid of compiler warnings:
2892       . In the match_data structure, match_limit was an unsigned long int, whereas
2893         match_call_count was an int. I've made them both unsigned long ints.
2895       . In pcretest the fact that a const uschar * doesn't automatically cast to
2896         a void * provoked a warning.
2898       . Turning on some more compiler warnings threw up some "shadow" variables
2899         and a few more missing casts.
2901    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2902       option, a class that contained a single character with a value between 128
2903       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2905    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2906       option, a class that contained several characters, but with at least one
2907       whose value was between 128 and 255 caused PCRE to crash.
2910    Version 4.1 12-Mar-03
2911    ---------------------
2913    1. Compiling with gcc -pedantic found a couple of places where casts were
2914    needed, and a string in dftables.c that was longer than standard compilers are
2915    required to support.
2917    2. Compiling with Sun's compiler found a few more places where the code could
2918    be tidied up in order to avoid warnings.
2920    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2921    first of these names is deprecated in the latest Autoconf in favour of the name
2922    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2923    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2924    analogy I have changed it to CFLAGS_FOR_BUILD.
2926    4. Added -no-undefined to the linking command in the Makefile, because this is
2927    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2928    linking step for the pcreposix library.
2930    5. PCRE was failing to diagnose the case of two named groups with the same
2931    name.
2933    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2934    literal character that is needed in the subject for a match, and scans along to
2935    ensure that it is present before embarking on the full matching process. This
2936    saves time in cases of nested unlimited repeats that are never going to match.
2937    Problem: the scan can take a lot of time if the subject is very long (e.g.
2938    megabytes), thus penalizing straightforward matches. It is now done only if the
2939    amount of subject to be scanned is less than 1000 bytes.
2941    7. A lesser problem with the same optimization is that it was recording the
2942    first character of an anchored pattern as "needed", thus provoking a search
2943    right along the subject, even when the first match of the pattern was going to
2944    fail. The "needed" character is now not set for anchored patterns, unless it
2945    follows something in the pattern that is of non-fixed length. Thus, it still
2946    fulfils its original purpose of finding quick non-matches in cases of nested
2947    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2950    Version 4.0 17-Feb-03
2951    ---------------------
2953    1. If a comment in an extended regex that started immediately after a meta-item
2954    extended to the end of string, PCRE compiled incorrect data. This could lead to
2955    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2957    2. Moved to autoconf 2.53 and libtool 1.4.2.
2959    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2960    the special perltest8 script is no longer needed - all the tests can be run
2961    from a single perltest script.
2963    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2964    by \s. It has now been removed in PCRE. This means it isn't recognized as
2965    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2966    class [:space:] *does* include VT, thereby creating a mess.
2968    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2969    space and tab.
2971    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2972    its new features into the main test script, reducing the number of scripts.
2974    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2975    were backward compatible, and made the (?i) apply to the whole pattern, as if
2976    /i were given. Now it behaves more logically, and applies the option setting
2977    only to what follows. PCRE has been changed to follow suit. However, if it
2978    finds options settings right at the start of the pattern, it extracts them into
2979    the global options, as before. Thus, they show up in the info data.
2981    8. Added support for the \Q...\E escape sequence. Characters in between are
2982    treated as literals. This is slightly different from Perl in that $ and @ are
2983    also handled as literals inside the quotes. In Perl, they will cause variable
2984    interpolation. Note the following examples:
2986        Pattern            PCRE matches      Perl matches
2988        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2989        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2990        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2992    For compatibility with Perl, \Q...\E sequences are recognized inside character
2993    classes as well as outside them.
2995    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2996    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2997    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2998    signed/unsigned warnings.
3000    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3001    option for pcretest, so I've replaced it by a simple function that does just
3002    that job.
3004    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3005    "pcregrep -".
3007    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3008    Java package. This provides some syntactic sugar for simple cases of what my
3009    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3010    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3011    item, you can use this simplified notation. Note that only makes sense with
3012    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3013    greediness, whatever the setting of the PCRE_UNGREEDY option.
3015    13. A change of greediness default within a pattern was not taking effect at
3016    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3017    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3018    was abstracted outside.
3020    14. PCRE now supports the \G assertion. It is true when the current matching
3021    position is at the start point of the match. This differs from \A when the
3022    starting offset is non-zero. Used with the /g option of pcretest (or similar
3023    code), it works in the same way as it does for Perl's /g option. If all
3024    alternatives of a regex begin with \G, the expression is anchored to the start
3025    match position, and the "anchored" flag is set in the compiled expression.
3027    15. Some bugs concerning the handling of certain option changes within patterns
3028    have been fixed. These applied to options other than (?ims). For example,
3029    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3030    been the other way round. Some of this was related to change 7 above.
3032    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3033    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3034    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3035    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3037    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3038    mode. Unlike ".", it always matches newline, whatever the setting of
3039    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3040    assertions. Perl allows it, but it doesn't (in general) work because it can't
3041    calculate the length of the lookbehind. At least, that's the case for Perl
3042    5.8.0 - I've been told they are going to document that it doesn't work in
3043    future.
3045    18. Added an error diagnosis for escapes that PCRE does not support: these are
3046    \L, \l, \N, \P, \p, \U, \u, and \X.
3048    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3049    reading past the end of the pattern in cases such as /[abcd/.
3051    20. PCRE was getting more memory than necessary for patterns with classes that
3052    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3054    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3055    compiling PCRE for use with Virtual Pascal.
3057    22. Small fix to the Makefile to make it work properly if the build is done
3058    outside the source tree.
3060    23. Added a new extension: a condition to go with recursion. If a conditional
3061    subpattern starts with (?(R) the "true" branch is used if recursion has
3062    happened, whereas the "false" branch is used only at the top level.
3064    24. When there was a very long string of literal characters (over 255 bytes
3065    without UTF support, over 250 bytes with UTF support), the computation of how
3066    much memory was required could be incorrect, leading to segfaults or other
3067    strange effects.
3069    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3070    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3071    there was a subsequent back reference to those brackets. This meant that, for
3072    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3073    possible to check for precisely this case. All we can do is abandon the
3074    optimization if .* occurs inside capturing brackets when there are any back
3075    references whatsoever. (See below for a better fix that came later.)
3077    26. The handling of the optimization for finding the first character of a
3078    non-anchored pattern, and for finding a character that is required later in the
3079    match were failing in some cases. This didn't break the matching; it just
3080    failed to optimize when it could. The way this is done has been re-implemented.
3082    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3084    28. Added a new feature that provides some of the functionality that Perl
3085    provides with (?{...}). The facility is termed a "callout". The way it is done
3086    in PCRE is for the caller to provide an optional function, by setting
3087    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3088    global variable. By default it is unset, which disables all calling out. To get
3089    the function called, the regex must include (?C) at appropriate points. This
3090    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3091    This provides a means of identifying different callout points. When PCRE
3092    reaches such a point in the regex, if pcre_callout has been set, the external
3093    function is called. It is provided with data in a structure called
3094    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3095    matching continues; if it returns a non-zero value, the match at the current
3096    point fails. However, backtracking will occur if possible. [This was changed
3097    later and other features added - see item 49 below.]
3099    29. pcretest is upgraded to test the callout functionality. It provides a
3100    callout function that displays information. By default, it shows the start of
3101    the match and the current position in the text. There are some new data escapes
3102    to vary what happens:
3104        \C+         in addition, show current contents of captured substrings
3105        \C-         do not supply a callout function
3106        \C!n        return 1 when callout number n is reached
3107        \C!n!m      return 1 when callout number n is reached for the mth time
3109    30. If pcregrep was called with the -l option and just a single file name, it
3110    output "<stdin>" if a match was found, instead of the file name.
3112    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3113    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3114    pcre_exec(). This saves a malloc/free per call. The default value of
3115    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3116    when configuring.
3118    32. The default maximum size of a compiled pattern is 64K. There have been a
3119    few cases of people hitting this limit. The code now uses macros to handle the
3120    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3121    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3122    configuring. Tests 2 and 5 work only with 2-byte links because they output
3123    debugging information about compiled patterns.
3125    33. Internal code re-arrangements:
3127    (a) Moved the debugging function for printing out a compiled regex into
3128        its own source file (printint.c) and used #include to pull it into
3129        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3130        separate copies.
3132    (b) Defined the list of op-code names for debugging as a macro in
3133        internal.h so that it is next to the definition of the opcodes.
3135    (c) Defined a table of op-code lengths for simpler skipping along compiled
3136        code. This is again a macro in internal.h so that it is next to the
3137        definition of the opcodes.
3139    34. Added support for recursive calls to individual subpatterns, along the
3140    lines of Robin Houston's patch (but implemented somewhat differently).
3142    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3143    allow it to read and process whole directories in Win32. This code was
3144    contributed by Lionel Fourquaux; it has not been tested by me.
3146    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3147    used to name a group. Names consist of alphanumerics and underscores, and must
3148    be unique. Back references use the syntax (?P=name) and recursive calls use
3149    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3150    numbers. The function pcre_fullinfo() can be used after compilation to extract
3151    a name/number map. There are three relevant calls:
3153      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3154      PCRE_INFO_NAMECOUNT            yields the number of entries
3155      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3157    The map is a vector of fixed-size entries. The size of each entry depends on
3158    the length of the longest name used. The first two bytes of each entry are the
3159    group number, most significant byte first. There follows the corresponding
3160    name, zero terminated. The names are in alphabetical order.
3162    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3163    case instead of 255. Making it the same both with and without UTF-8 support
3164    means that the same test output works with both.
3166    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3167    calling malloc() with a zero argument.
3169    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3170    optimization. I've improved things by keeping a bitmap of backreferences with
3171    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3172    fact referenced, the optimization can be applied. It is unlikely that a
3173    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3174    the match to follow \n) will appear inside brackets with a number greater than
3175    31, but if it does, any back reference > 31 suppresses the optimization.
3177    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3178    of disabling numbered capturing parentheses. Any opening parenthesis that is
3179    not followed by ? behaves as if it were followed by ?: but named parentheses
3180    can still be used for capturing (and they will acquire numbers in the usual
3181    way).
3183    41. Redesigned the return codes from the match() function into yes/no/error so
3184    that errors can be passed back from deep inside the nested calls. A malloc
3185    failure while inside a recursive subpattern call now causes the
3186    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3188    42. It is now possible to set a limit on the number of times the match()
3189    function is called in a call to pcre_exec(). This facility makes it possible to
3190    limit the amount of recursion and backtracking, though not in a directly
3191    obvious way, because the match() function is used in a number of different
3192    circumstances. The count starts from zero for each position in the subject
3193    string (for non-anchored patterns). The default limit is, for compatibility, a
3194    large number, namely 10 000 000. You can change this in two ways:
3196    (a) When configuring PCRE before making, you can use --with-match-limit=n
3197        to set a default value for the compiled library.
3199    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3200        a different value is set. See 45 below.
3202    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3204    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3205    of things that can be changed at compile time. The first argument specifies
3206    what is wanted and the second points to where the information is to be placed.
3207    The current list of available information is:
3209      PCRE_CONFIG_UTF8
3211    The output is an integer that is set to one if UTF-8 support is available;
3212    otherwise it is set to zero.
3216    The output is an integer that it set to the value of the code that is used for
3217    newline. It is either LF (10) or CR (13).
3221    The output is an integer that contains the number of bytes used for internal
3222    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3226    The output is an integer that contains the threshold above which the POSIX
3227    interface uses malloc() for output vectors. See item 31 above.
3231    The output is an unsigned integer that contains the default limit of the number
3232    of match() calls in a pcre_exec() execution. See 42 above.
3234    44. pcretest has been upgraded by the addition of the -C option. This causes it
3235    to extract all the available output from the new pcre_config() function, and to
3236    output it. The program then exits immediately.
3238    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3239    order to support additional features. One way would have been to define
3240    pcre_exec2() (for example) with extra arguments, but this would not have been
3241    extensible, and would also have required all calls to the original function to
3242    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3243    is used for passing in "extra" data from pcre_study().
3245    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3246    contains the following fields:
3248      flags         a bitmap indicating which of the following fields are set
3249      study_data    opaque data from pcre_study()
3250      match_limit   a way of specifying a limit on match() calls for a specific
3251                      call to pcre_exec()
3252      callout_data  data for callouts (see 49 below)
3254    The flag bits are also defined in pcre.h, and are
3260    The pcre_study() function now returns one of these new pcre_extra blocks, with
3261    the actual study data pointed to by the study_data field, and the
3262    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3263    before. That is, this change is entirely upwards-compatible and requires no
3264    change to existing code.
3266    If you want to pass in additional data to pcre_exec(), you can either place it
3267    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3268    block.
3270    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3271    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3272    times with different match limits, until it finds the minimum value needed for
3273    pcre_exec() to complete. The value is then output. This can be instructive; for
3274    most simple matches the number is quite small, but for pathological cases it
3275    gets very large very quickly.
3277    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3278    returns the size of the data block pointed to by the study_data field in a
3279    pcre_extra block, that is, the value that was passed as the argument to
3280    pcre_malloc() when PCRE was getting memory in which to place the information
3281    created by pcre_study(). The fourth argument should point to a size_t variable.
3282    pcretest has been extended so that this information is shown after a successful
3283    pcre_study() call when information about the compiled regex is being displayed.
3285    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3286    because what follows is always an absolute path. (Later: it turns out that this
3287    is more than cosmetic for MinGW, because it doesn't like empty path
3288    components.)
3290    49. Some changes have been made to the callout feature (see 28 above):
3292    (i)  A callout function now has three choices for what it returns:
3294           0  =>  success, carry on matching
3295         > 0  =>  failure at this point, but backtrack if possible
3296         < 0  =>  serious error, return this value from pcre_exec()
3298         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3299         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3300         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3301         use by callout functions. It will never be used by PCRE itself.
3303    (ii) The pcre_extra structure (see 45 above) has a void * field called
3304         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3305         pcre_callout_block structure has a field of the same name. The contents of
3306         the field passed in the pcre_extra structure are passed to the callout
3307         function in the corresponding field in the callout block. This makes it
3308         easier to use the same callout-containing regex from multiple threads. For
3309         testing, the pcretest program has a new data escape
3311           \C*n        pass the number n (may be negative) as callout_data
3313         If the callout function in pcretest receives a non-zero value as
3314         callout_data, it returns that value.
3316    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3317    there were some redundant $(CFLAGS) in commands that are now specified as
3318    $(LINK), which already includes $(CFLAGS).
3320    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3321    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3322    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3323    one-byte characters throughout. Note that case-insensitive matching applies
3324    only to characters whose values are less than 256. PCRE doesn't support the
3325    notion of cases for higher-valued characters.
3327    (i)   A character class whose characters are all within 0-255 is handled as
3328          a bit map, and the map is inverted for negative classes. Previously, a
3329          character > 255 always failed to match such a class; however it should
3330          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3332    (ii)  A negated character class with a single character < 255 is coded as
3333          "not this character" (OP_NOT). This wasn't working properly when the test
3334          character was multibyte, either singly or repeated.
3336    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3337          mode, for example: \x{100}{2,3}.
3339    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3340          singly or repeated) now correctly test multibyte characters. However,
3341          PCRE doesn't recognize any characters with values greater than 255 as
3342          digits, spaces, or word characters. Such characters always match \D, \S,
3343          and \W, and never match \d, \s, or \w.
3345    (v)   Classes may now contain characters and character ranges with values
3346          greater than 255. For example: [ab\x{100}-\x{400}].
3348    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3349          PCRE in UTF-8 mode.
3351    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3352    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3353    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3354    value.)
3356    53. The single man page has become too large. I have therefore split it up into
3357    a number of separate man pages. These also give rise to individual HTML pages;
3358    these are now put in a separate directory, and there is an index.html page that
3359    lists them all. Some hyperlinking between the pages has been installed.
3361    54. Added convenience functions for handling named capturing parentheses.
3363    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3364    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3365    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3366    are faulted.
3368    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3369    calling configure. These values are used when compiling the dftables.c program
3370    which is run to generate the source of the default character tables. They
3371    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3372    you will need to set these values.
3374    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3377    Version 3.9 02-Jan-02
3378    ---------------------
3380    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3382    2. If --disable-static was given, the building process failed when trying to
3383    build pcretest and pcregrep. (For some reason it was using libtool to compile
3384    them, which is not right, as they aren't part of the library.)
3387    Version 3.8 18-Dec-01
3388    ---------------------
3390    1. The experimental UTF-8 code was completely screwed up. It was packing the
3391    bytes in the wrong order. How dumb can you get?
3394    Version 3.7 29-Oct-01
3395    ---------------------
3397    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3398    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3399    this didn't happen under Solaris 8, where I normally test things.
3401    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3402    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3403    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3404    this fix an hour or so after the initial 3.7 release.)
3407    Version 3.6 23-Oct-01
3408    ---------------------
3410    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3411    offsets passed as NULL with zero offset count.
3413    2. The config.guess and config.sub files had not been updated when I moved to
3414    the latest autoconf.
3417    Version 3.5 15-Aug-01
3418    ---------------------
3420    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3421    had been forgotten.
3423    2. By using declared but undefined structures, we can avoid using "void"
3424    definitions in pcre.h while keeping the internal definitions of the structures
3425    private.
3427    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3428    user point of view, this means that both static and shared libraries are built
3429    by default, but this can be individually controlled. More of the work of
3430    handling this static/shared cases is now inside libtool instead of PCRE's make
3431    file.
3433    4. The pcretest utility is now installed along with pcregrep because it is
3434    useful for users (to test regexs) and by doing this, it automatically gets
3435    relinked by libtool. The documentation has been turned into a man page, so
3436    there are now .1, .txt, and .html versions in /doc.
3438    5. Upgrades to pcregrep:
3439       (i)   Added long-form option names like gnu grep.
3440       (ii)  Added --help to list all options with an explanatory phrase.
3441       (iii) Added -r, --recursive to recurse into sub-directories.
3442       (iv)  Added -f, --file to read patterns from a file.
3444    6. pcre_exec() was referring to its "code" argument before testing that
3445    argument for NULL (and giving an error if it was NULL).
3447    7. Upgraded Makefile.in to allow for compiling in a different directory from
3448    the source directory.
3450    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3451    options bits, the pointer it was passed was to an int instead of to an unsigned
3452    long int. This mattered only on 64-bit systems.
3454    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3455    generated) instead of pcre.in, which it its source. Also made the same change
3456    in several of the .c files.
3458    10. A new release of gcc defines printf() as a macro, which broke pcretest
3459    because it had an ifdef in the middle of a string argument for printf(). Fixed
3460    by using separate calls to printf().
3462    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3463    script, to force use of CR or LF instead of \n in the source. On non-Unix
3464    systems, the value can be set in config.h.
3466    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3467    absolute limit. Changed the text of the error message to make this clear, and
3468    likewise updated the man page.
3470    13. The limit of 99 on the number of capturing subpatterns has been removed.
3471    The new limit is 65535, which I hope will not be a "real" limit.
3474    Version 3.4 22-Aug-00
3475    ---------------------
3477    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3479    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3482    Version 3.3 01-Aug-00
3483    ---------------------
3485    1. If an octal character was given, but the value was greater than \377, it
3486    was not getting masked to the least significant bits, as documented. This could
3487    lead to crashes in some systems.
3489    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3490    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3492    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3493    These just pass their arguments on to (pcre_free)(), but they are provided
3494    because some uses of PCRE bind it to non-C systems that can call its functions,
3495    but cannot call free() or pcre_free() directly.
3497    4. Add "make test" as a synonym for "make check". Corrected some comments in
3498    the Makefile.
3500    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3501    Makefile.
3503    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3504    command called pgrep for grepping around the active processes.
3506    7. Added the beginnings of support for UTF-8 character strings.
3508    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3509    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3510    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3511    out for the ar command.)
3514    Version 3.2 12-May-00
3515    ---------------------
3517    This is purely a bug fixing release.
3519    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3520    of ZA. This was just one example of several cases that could provoke this bug,
3521    which was introduced by change 9 of version 2.00. The code for breaking
3522    infinite loops after an iteration that matches an empty string was't working
3523    correctly.
3525    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3526    when matched against abbab (for example). After matching an empty string, it
3527    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3528    caused it to match further down the string than it should.
3530    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3531    was there because it doesn't seem to be needed, and it causes trouble on some
3532    systems, as it is not a Standard C header. It has been removed.
3534    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3535    were reported on the Macintosh. The changes were from
3537      while ((c = *(++ptr)) != 0 && c != '\n');
3538    to
3539      while ((c = *(++ptr)) != 0 && c != '\n') ;
3541    Totally extraordinary, but if that's what it takes...
3543    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3544    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3545    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3546    assumes the way PCRE uses memmove() (always moving upwards).
3548    6. PCRE is being used in one environment where strchr() is not available. There
3549    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3550    faster code anyway.
3553    Version 3.1 09-Feb-00
3554    ---------------------
3556    The only change in this release is the fixing of some bugs in Makefile.in for
3557    the "install" target:
3559    (1) It was failing to install pcreposix.h.
3561    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3564    Version 3.0 01-Feb-00
3565    ---------------------
3567    1. Add support for the /+ modifier to perltest (to output $` like it does in
3568    pcretest).
3570    2. Add support for the /g modifier to perltest.
3572    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3573    matches null strings.
3575    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3576    pattern. Perl treats empty patterns specially - it reuses the most recent
3577    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3578    effect.
3580    5. The POSIX interface was broken in that it was just handing over the POSIX
3581    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3582    required a bigger vector, with some working space on the end. This means that
3583    the POSIX wrapper now has to get and free some memory, and copy the results.
3585    6. Added some simple autoconf support, placing the test data and the
3586    documentation in separate directories, re-organizing some of the
3587    information files, and making it build pcre-config (a GNU standard). Also added
3588    libtool support for building PCRE as a shared library, which is now the
3589    default.