ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 11 by nigel, Sat Feb 24 21:38:17 2007 UTC revision 598 by ph10, Sat May 7 15:37:31 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.13 30-Apr-2011
5    ------------------------
7    1.  The Unicode data tables have been updated to Unicode 6.0.0.
9    2.  Two minor typos in pcre_internal.h have been fixed.
11    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
12        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
13        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
15    4.  There were a number of related bugs in the code for matching backrefences
16        caselessly in UTF-8 mode when codes for the characters concerned were
17        different numbers of bytes. For example, U+023A and U+2C65 are an upper
18        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
19        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
20        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
21        2-byte code at the end of the subject (it thought there wasn't enough data
22        left).
24    5.  Comprehensive information about what went wrong is now returned by
25        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
26        as the output vector has at least 2 elements. The offset of the start of
27        the failing character and a reason code are placed in the vector.
29    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
30        now returned is for the first byte of the failing character, instead of the
31        last byte inspected. This is an incompatible change, but I hope it is small
32        enough not to be a problem. It makes the returned offset consistent with
33        pcre_exec() and pcre_dfa_exec().
35    7.  pcretest now gives a text phrase as well as the error number when
36        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
37        failure, the offset and reason code are output.
41    Version 8.12 15-Jan-2011
42    ------------------------
44    1.  Fixed some typos in the markup of the man pages, and wrote a script that
45        checks for such things as part of the documentation building process.
47    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
48        --match-limit and --recursion-limit options (added for 8.11). In
49        particular, this made one of the standard tests fail. (The integer value
50        went into the wrong half of a long int.)
52    3.  If the --colour option was given to pcregrep with -v (invert match), it
53        did strange things, either producing crazy output, or crashing. It should,
54        of course, ignore a request for colour when reporting lines that do not
55        match.
57    4.  Another pcregrep bug caused similar problems if --colour was specified with
58        -M (multiline) and the pattern match finished with a line ending.
60    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
61        matched in multiline mode, the following line was shown as part of the
62        match. This seems wrong, so I have changed it.
64    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
65        the check for further matches in the same line (so they could be coloured)
66        to overrun the end of the current line. If another match was found, it was
67        incorrectly shown (and then shown again when found in the next line).
69    7.  If pcregrep was compiled under Windows, there was a reference to the
70        function pcregrep_exit() before it was defined. I am assuming this was
71        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
72        reported by a user. I've moved the definition above the reference.
75    Version 8.11 10-Dec-2010
76    ------------------------
78    1.  (*THEN) was not working properly if there were untried alternatives prior
79        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
80        backtracked to try for "b" instead of moving to the next alternative branch
81        at the same level (in this case, to look for "c"). The Perl documentation
82        is clear that when (*THEN) is backtracked onto, it goes to the "next
83        alternative in the innermost enclosing group".
85    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
86        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
87        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
88        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
89        (*THEN).
91    3.  If \s appeared in a character class, it removed the VT character from
92        the class, even if it had been included by some previous item, for example
93        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
94        of \s, but is part of the POSIX "space" class.)
96    4.  A partial match never returns an empty string (because you can always
97        match an empty string at the end of the subject); however the checking for
98        an empty string was starting at the "start of match" point. This has been
99        changed to the "earliest inspected character" point, because the returned
100        data for a partial match starts at this character. This means that, for
101        example, /(?<=abc)def/ gives a partial match for the subject "abc"
102        (previously it gave "no match").
104    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
105        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
106        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
107        has an implication that the given string is incomplete (because a partial
108        match is preferred over a full match). For this reason, these items now
109        give a partial match in this situation. [Aside: previously, the one case
110        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
111        match rather than a full match, which was wrong by the old rules, but is
112        now correct.]
114    6.  There was a bug in the handling of #-introduced comments, recognized when
115        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
116        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
117        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
118        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
119        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
120        places in pcre_compile().
122    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
123        comments when looking ahead for named forward references to subpatterns,
124        the only newline sequence it recognized was NL. It now handles newlines
125        according to the set newline convention.
127    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
128        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
129        cater for a lack of strerror(). These oversights have been fixed.
131    9.  Added --match-limit and --recursion-limit to pcregrep.
133    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
135    11. When the -o option was used, pcregrep was setting a return code of 1, even
136        when matches were found, and --line-buffered was not being honoured.
138    12. Added an optional parentheses number to the -o and --only-matching options
139        of pcregrep.
141    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
142        can match an empty string. The code to do it in pcretest and pcredemo
143        needed fixing:
145        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
146            only one byte after an empty string match just before CRLF (this case
147            just got forgotten; "any" and "anycrlf" were OK).
149        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
150            mode when an empty string match preceded an ASCII character followed by
151            a non-ASCII character. (The code for advancing by one character rather
152            than one byte was nonsense.)
154        (c) The pcredemo.c sample program did not have any code at all to handle
155            the cases when CRLF is a valid newline sequence.
157    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
158        as a starting offset was within the subject string. There is now a new
159        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
160        negative or greater than the length of the string. In order to test this,
161        pcretest is extended to allow the setting of negative starting offsets.
163    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
164        starting offset points to the beginning of a UTF-8 character was
165        unnecessarily clumsy. I tidied it up.
167    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
168        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
170    17. Nobody had reported that the --include_dir option, which was added in
171        release 7.7 should have been called --include-dir (hyphen, not underscore)
172        for compatibility with GNU grep. I have changed it to --include-dir, but
173        left --include_dir as an undocumented synonym, and the same for
174        --exclude-dir, though that is not available in GNU grep, at least as of
175        release 2.5.4.
177    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
178        characters from a string of bytes) have been redefined so as not to use
179        loops, in order to improve performance in some environments. At the same
180        time, I abstracted some of the common code into auxiliary macros to save
181        repetition (this should not affect the compiled code).
183    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
184        compile-time error is now given if \c is not followed by an ASCII
185        character, that is, a byte less than 128. (In EBCDIC mode, the code is
186        different, and any byte value is allowed.)
188    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
189        START_OPTIMIZE option, which is now allowed at compile time - but just
190        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
191        to pcregrep and other applications that have no direct access to PCRE
192        options. The new /Y option in pcretest sets this option when calling
193        pcre_compile().
195    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
196        back references. Groups containing recursive back references were forced to
197        be atomic by that change, but in the case of named groups, the amount of
198        memory required was incorrectly computed, leading to "Failed: internal
199        error: code overflow". This has been fixed.
201    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
202        pcretest.c, to avoid build problems in some Borland environments.
205    Version 8.10 25-Jun-2010
206    ------------------------
208    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
209        THEN.
211    2.  (*ACCEPT) was not working when inside an atomic group.
213    3.  Inside a character class, \B is treated as a literal by default, but
214        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
215        causes the error). The code is unchanged, but I tidied the documentation.
217    4.  Inside a character class, PCRE always treated \R and \X as literals,
218        whereas Perl faults them if its -w option is set. I have changed PCRE so
219        that it faults them when PCRE_EXTRA is set.
221    5.  Added support for \N, which always matches any character other than
222        newline. (It is the same as "." when PCRE_DOTALL is not set.)
224    6.  When compiling pcregrep with newer versions of gcc which may have
225        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
226        declared with attribute warn_unused_result" were given. Just casting the
227        result to (void) does not stop the warnings; a more elaborate fudge is
228        needed. I've used a macro to implement this.
230    7.  Minor change to pcretest.c to avoid a compiler warning.
232    8.  Added four artifical Unicode properties to help with an option to make
233        \s etc use properties (see next item). The new properties are: Xan
234        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
236    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
237        use Unicode properties. (*UCP) at the start of a pattern can be used to set
238        this option. Modified pcretest to add /W to test this facility. Added
239        REG_UCP to make it available via the POSIX interface.
241    10. Added --line-buffered to pcregrep.
243    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
244        studied, and the match started with a letter with a code point greater than
245        127 whose first byte was different to the first byte of the other case of
246        the letter, the other case of this starting letter was not recognized
247        (#976).
249    12. If a pattern that was studied started with a repeated Unicode property
250        test, for example, \p{Nd}+, there was the theoretical possibility of
251        setting up an incorrect bitmap of starting bytes, but fortunately it could
252        not have actually happened in practice until change 8 above was made (it
253        added property types that matched character-matching opcodes).
255    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
256        possible starting bytes for non-anchored patterns.
258    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
259        \R, and also a number of cases that involve Unicode properties, both
260        explicit and implicit when PCRE_UCP is set.
262    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
263        input, it could crash or give wrong results if characters with values
264        greater than 0xc0 were present in the subject string. (Detail: it assumed
265        UTF-8 input when processing these items.)
267    16. Added a lot of (int) casts to avoid compiler warnings in systems where
268        size_t is 64-bit (#991).
270    17. Added a check for running out of memory when PCRE is compiled with
271        --disable-stack-for-recursion (#990).
273    18. If the last data line in a file for pcretest does not have a newline on
274        the end, a newline was missing in the output.
276    19. The default pcre_chartables.c file recognizes only ASCII characters (values
277        less than 128) in its various bitmaps. However, there is a facility for
278        generating tables according to the current locale when PCRE is compiled. It
279        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
280        characters, are recognized by isspace() and therefore were getting set in
281        these tables, and indeed these tables seem to approximate to ISO 8859. This
282        caused a problem in UTF-8 mode when pcre_study() was used to create a list
283        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
284        which of course cannot start UTF-8 characters. I have changed the code so
285        that only real ASCII characters (less than 128) and the correct starting
286        bytes for UTF-8 encodings are set for characters greater than 127 when in
287        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
288        altogether.)
290    20. Added the /T option to pcretest so as to be able to run tests with non-
291        standard character tables, thus making it possible to include the tests
292        used for 19 above in the standard set of tests.
294    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
295        reference to a subpattern the other side of a comment that contains an
296        opening parenthesis caused either an internal compiling error, or a
297        reference to the wrong subpattern.
300    Version 8.02 19-Mar-2010
301    ------------------------
303    1.  The Unicode data tables have been updated to Unicode 5.2.0.
305    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
306        configured.
308    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
309        original author of that file, following a query about its status.
311    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
312        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
314    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
315        quantifier applied to a forward-referencing subroutine call, could compile
316        incorrect code or give the error "internal error: previously-checked
317        referenced subpattern not found".
319    6.  Both MS Visual Studio and Symbian OS have problems with initializing
320        variables to point to external functions. For these systems, therefore,
321        pcre_malloc etc. are now initialized to local functions that call the
322        relevant global functions.
324    7.  There were two entries missing in the vectors called coptable and poptable
325        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
326        I've fixed the data, and added a kludgy way of testing at compile time that
327        the lengths are correct (equal to the number of opcodes).
329    8.  Following on from 7, I added a similar kludge to check the length of the
330        eint vector in pcreposix.c.
332    9.  Error texts for pcre_compile() are held as one long string to avoid too
333        much relocation at load time. To find a text, the string is searched,
334        counting zeros. There was no check for running off the end of the string,
335        which could happen if a new error number was added without updating the
336        string.
338    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
340    11. \K was not working if it appeared in an atomic group or in a group that
341        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
342        \K is "not well defined" if used in an assertion. PCRE now accepts it if
343        the assertion is positive, but not if it is negative.
345    12. Change 11 fortuitously reduced the size of the stack frame used in the
346        "match()" function of pcre_exec.c by one pointer. Forthcoming
347        implementation of support for (*MARK) will need an extra pointer on the
348        stack; I have reserved it now, so that the stack frame size does not
349        decrease.
351    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
352        item in branch that calls a recursion is a subroutine call - as in the
353        second branch in the above example - was incorrectly given the compile-
354        time error "recursive call could loop indefinitely" because pcre_compile()
355        was not correctly checking the subroutine for matching a non-empty string.
357    14. The checks for overrunning compiling workspace could trigger after an
358        overrun had occurred. This is a "should never occur" error, but it can be
359        triggered by pathological patterns such as hundreds of nested parentheses.
360        The checks now trigger 100 bytes before the end of the workspace.
362    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
365    Version 8.01 19-Jan-2010
366    ------------------------
368    1.  If a pattern contained a conditional subpattern with only one branch (in
369        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
370        computed the wrong minimum data length (which is of course zero for such
371        subpatterns). This could cause incorrect "no match" results.
373    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
374        the pattern is reset in the first branch, pcre_compile() failed with
375        "internal error: code overflow at offset...". This happened only when
376        the reset was to the original external option setting. (An optimization
377        abstracts leading options settings into an external setting, which was the
378        cause of this.)
380    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
381        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
382        assertion pattern did not match (meaning that the assertion was true), it
383        was incorrectly treated as false if the SKIP had been reached during the
384        matching. This also applied to assertions used as conditions.
386    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
387        assertion subpattern, including such a pattern used as a condition,
388        unpredictable results occurred, instead of the error return
391    5.  The C++ GlobalReplace function was not working like Perl for the special
392        situation when an empty string is matched. It now does the fancy magic
393        stuff that is necessary.
395    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
396        removed. (These were left over from very, very early versions of PCRE.)
398    7.  Some cosmetic changes to the code to make life easier when compiling it
399        as part of something else:
401        (a) Change DEBUG to PCRE_DEBUG.
403        (b) In pcre_compile(), rename the member of the "branch_chain" structure
404            called "current" as "current_branch", to prevent a collision with the
405            Linux macro when compiled as a kernel module.
407        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
408            prevent a collision with the Linux macro when compiled as a kernel
409            module.
411    8.  In pcre_compile() there are some checks for integer overflows that used to
412        cast potentially large values to (double). This has been changed to that
413        when building, a check for int64_t is made, and if it is found, it is used
414        instead, thus avoiding the use of floating point arithmetic. (There is no
415        other use of FP in PCRE.) If int64_t is not found, the fallback is to
416        double.
418    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
419        2005 (difference between two addresses compared to an unsigned value).
421    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
422        custom one, because of the following reported problem in Windows:
424          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
425              under Win32.
426          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
427              therefore missing the function definition.
428          - The compiler thus generates a "C" signature for the test function.
429          - The linker fails to find the "C" function.
430          - PCRE fails to configure if asked to do so against libbz2.
432    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
433        messages were output:
435          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
436          rerunning libtoolize, to keep the correct libtool macros in-tree.
437          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
439        I have done both of these things.
441    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
442        most of the time, it *can* run out if it is given a pattern that contains a
443        runaway infinite recursion. I updated the discussion in the pcrestack man
444        page.
446    13. Now that we have gone to the x.xx style of version numbers, the minor
447        version may start with zero. Using 08 or 09 is a bad idea because users
448        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
449        interpreted as invalid octal numbers. I've updated the previous comment in
450        configure.ac, and also added a check that gives an error if 08 or 09 are
451        used.
453    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
454        causing partial matching to fail when the end of the subject matched \W
455        in a UTF-8 pattern where \W was quantified with a minimum of 3.
457    15. There were some discrepancies between the declarations in pcre_internal.h
458        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
459        their definitions. The declarations used "const uschar *" and the
460        definitions used USPTR. Even though USPTR is normally defined as "const
461        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
462        reported that: "This difference in casting confuses some C++ compilers, for
463        example, SunCC recognizes above declarations as different functions and
464        generates broken code for hbpcre." I have changed the declarations to use
465        USPTR.
467    16. GNU libtool is named differently on some systems. The autogen.sh script now
468        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
469        (FreeBSD).
471    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
472        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
473        comment: "Figure out how to create a longlong from a string: strtoll and
474        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
475        instance, but it only takes 2 args instead of 3!"
477    18. A subtle bug concerned with back references has been fixed by a change of
478        specification, with a corresponding code fix. A pattern such as
479        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
480        refers, was giving matches when it shouldn't. For example, xa=xaaa would
481        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
482        same bug. Such groups have to be quantified to be useful, or contained
483        inside another quantified group. (If there's no repetition, the reference
484        can never match.) The problem arises because, having left the group and
485        moved on to the rest of the pattern, a later failure that backtracks into
486        the group uses the captured value from the final iteration of the group
487        rather than the correct earlier one. I have fixed this in PCRE by forcing
488        any group that contains a reference to itself to be an atomic group; that
489        is, there cannot be any backtracking into it once it has completed. This is
490        similar to recursive and subroutine calls.
493    Version 8.00 19-Oct-09
494    ----------------------
496    1.  The table for translating pcre_compile() error codes into POSIX error codes
497        was out-of-date, and there was no check on the pcre_compile() error code
498        being within the table. This could lead to an OK return being given in
499        error.
501    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
502        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
503        in a Windows environment.
505    3.  The pcregrep --count option prints the count for each file even when it is
506        zero, as does GNU grep. However, pcregrep was also printing all files when
507        --files-with-matches was added. Now, when both options are given, it prints
508        counts only for those files that have at least one match. (GNU grep just
509        prints the file name in this circumstance, but including the count seems
510        more useful - otherwise, why use --count?) Also ensured that the
511        combination -clh just lists non-zero counts, with no names.
513    4.  The long form of the pcregrep -F option was incorrectly implemented as
514        --fixed_strings instead of --fixed-strings. This is an incompatible change,
515        but it seems right to fix it, and I didn't think it was worth preserving
516        the old behaviour.
518    5.  The command line items --regex=pattern and --regexp=pattern were not
519        recognized by pcregrep, which required --regex pattern or --regexp pattern
520        (with a space rather than an '='). The man page documented the '=' forms,
521        which are compatible with GNU grep; these now work.
523    6.  No libpcreposix.pc file was created for pkg-config; there was just
524        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
526    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
527        when UCP support is not needed, by modifying the Python script that
528        generates it from Unicode data files. This should not matter if the module
529        is correctly used as a library, but I received one complaint about 50K of
530        unwanted data. My guess is that the person linked everything into his
531        program rather than using a library. Anyway, it does no harm.
533    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
534        was a minimum greater than 1 for a wide character in a possessive
535        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
536        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
537        character. Chaos in the form of incorrect output or a compiling loop could
538        result.
540    9.  The restrictions on what a pattern can contain when partial matching is
541        requested for pcre_exec() have been removed. All patterns can now be
542        partially matched by this function. In addition, if there are at least two
543        slots in the offset vector, the offset of the earliest inspected character
544        for the match and the offset of the end of the subject are set in them when
545        PCRE_ERROR_PARTIAL is returned.
547    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
548        synonymous with PCRE_PARTIAL, for backwards compatibility, and
549        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
550        and may be more useful for multi-segment matching.
552    11. Partial matching with pcre_exec() is now more intuitive. A partial match
553        used to be given if ever the end of the subject was reached; now it is
554        given only if matching could not proceed because another character was
555        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
556        string "Z", which now yields "no match" instead of "partial match". In the
557        case of pcre_dfa_exec(), "no match" is given if every matching path for the
558        final character ended with (*FAIL).
560    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
561        if the pattern had a "must contain" character that was already found in the
562        earlier partial match, unless partial matching was again requested. For
563        example, with the pattern /dog.(body)?/, the "must contain" character is
564        "g". If the first part-match was for the string "dog", restarting with
565        "sbody" failed. This bug has been fixed.
567    13. The string returned by pcre_dfa_exec() after a partial match has been
568        changed so that it starts at the first inspected character rather than the
569        first character of the match. This makes a difference only if the pattern
570        starts with a lookbehind assertion or \b or \B (\K is not supported by
571        pcre_dfa_exec()). It's an incompatible change, but it makes the two
572        matching functions compatible, and I think it's the right thing to do.
574    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
575        so that the demonstration program is easily available in environments where
576        PCRE has not been installed from source.
578    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
579        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
580        library.
582    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
583        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
584        is not the first non-POSIX option to be added. Clearly some people find
585        these options useful.
587    17. If a caller to the POSIX matching function regexec() passes a non-zero
588        value for nmatch with a NULL value for pmatch, the value of
589        nmatch is forced to zero.
591    18. RunGrepTest did not have a test for the availability of the -u option of
592        the diff command, as RunTest does. It now checks in the same way as
593        RunTest, and also checks for the -b option.
595    19. If an odd number of negated classes containing just a single character
596        interposed, within parentheses, between a forward reference to a named
597        subpattern and the definition of the subpattern, compilation crashed with
598        an internal error, complaining that it could not find the referenced
599        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
600        [The bug was that it was starting one character too far in when skipping
601        over the character class, thus treating the ] as data rather than
602        terminating the class. This meant it could skip too much.]
604    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
605        /g option in pcretest when the pattern contains \K, which makes it possible
606        to have an empty string match not at the start, even when the pattern is
607        anchored. Updated pcretest and pcredemo to use this option.
609    21. If the maximum number of capturing subpatterns in a recursion was greater
610        than the maximum at the outer level, the higher number was returned, but
611        with unset values at the outer level. The correct (outer level) value is
612        now given.
614    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
615        PCRE did not set those parentheses (unlike Perl). I have now found a way to
616        make it do so. The string so far is captured, making this feature
617        compatible with Perl.
619    23. The tests have been re-organized, adding tests 11 and 12, to make it
620        possible to check the Perl 5.10 features against Perl 5.10.
622    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
623        pattern matches a fixed length string. PCRE did not allow this; now it
624        does. Neither allows recursion.
626    25. I finally figured out how to implement a request to provide the minimum
627        length of subject string that was needed in order to match a given pattern.
628        (It was back references and recursion that I had previously got hung up
629        on.) This code has now been added to pcre_study(); it finds a lower bound
630        to the length of subject needed. It is not necessarily the greatest lower
631        bound, but using it to avoid searching strings that are too short does give
632        some useful speed-ups. The value is available to calling programs via
633        pcre_fullinfo().
635    26. While implementing 25, I discovered to my embarrassment that pcretest had
636        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
637        study optimizations had never been tested with that matching function.
638        Oops. What is worse, even when it was passed study data, there was a bug in
639        pcre_dfa_exec() that meant it never actually used it. Double oops. There
640        were also very few tests of studied patterns with pcre_dfa_exec().
642    27. If (?| is used to create subpatterns with duplicate numbers, they are now
643        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
644        on the other side of the coin, they are no longer allowed to have different
645        names, because these cannot be distinguished in PCRE, and this has caused
646        confusion. (This is a difference from Perl.)
648    28. When duplicate subpattern names are present (necessarily with different
649        numbers, as required by 27 above), and a test is made by name in a
650        conditional pattern, either for a subpattern having been matched, or for
651        recursion in such a pattern, all the associated numbered subpatterns are
652        tested, and the overall condition is true if the condition is true for any
653        one of them. This is the way Perl works, and is also more like the way
654        testing by number works.
657    Version 7.9 11-Apr-09
658    ---------------------
660    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
661        (pcretest), all targets were linked against these libraries. This included
662        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
663        libraries. This caused unwanted dependencies to be created. This problem
664        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
665        pcretest is linked with readline.
667    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
668        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
669        moved outside it again, because FALSE and TRUE are already defined in AIX,
670        but BOOL is not.
672    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
673        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
675    4.  The pcregrep documentation said spaces were inserted as well as colons (or
676        hyphens) following file names and line numbers when outputting matching
677        lines. This is not true; no spaces are inserted. I have also clarified the
678        wording for the --colour (or --color) option.
680    5.  In pcregrep, when --colour was used with -o, the list of matching strings
681        was not coloured; this is different to GNU grep, so I have changed it to be
682        the same.
684    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
685        each matching line was coloured. Now it goes on to look for further matches
686        of any of the test patterns, which is the same behaviour as GNU grep.
688    7.  A pattern that could match an empty string could cause pcregrep to loop; it
689        doesn't make sense to accept an empty string match in pcregrep, so I have
690        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
691        seems to be how GNU grep behaves.
693    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
694        start or after a newline", because the conditional assertion was not being
695        correctly handled. The rule now is that both the assertion and what follows
696        in the first alternative must satisfy the test.
698    9.  If auto-callout was enabled in a pattern with a conditional group whose
699        condition was an assertion, PCRE could crash during matching, both with
700        pcre_exec() and pcre_dfa_exec().
702    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
703        used for matching.
705    11. Unicode property support in character classes was not working for
706        characters (bytes) greater than 127 when not in UTF-8 mode.
708    12. Added the -M command line option to pcretest.
710    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
712    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
714    16. Added comments and documentation about mis-use of no_arg in the C++
715        wrapper.
717    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
718        from Martin Jerabek that uses macro names for all relevant character and
719        string constants.
721    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
722        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
723        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
724        these, but not everybody uses configure.
726    19. A conditional group that had only one branch was not being correctly
727        recognized as an item that could match an empty string. This meant that an
728        enclosing group might also not be so recognized, causing infinite looping
729        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
730        with the subject "ab", where knowledge that the repeated group can match
731        nothing is needed in order to break the loop.
733    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
734        exec(), but without supplying a callout function, matching went wrong.
736    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
737        leak if the size of the offset vector was greater than 30. When the vector
738        is smaller, the saved offsets during recursion go onto a local stack
739        vector, but for larger vectors malloc() is used. It was failing to free
740        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
741        error, in fact).
743    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
744        heapframe that is used only when UTF-8 support is enabled. This caused no
745        problem, but was untidy.
747    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
748        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
749        included within another project.
751    24. Steven Van Ingelgem's patches to add more options to the CMake support,
752        slightly modified by me:
754          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
755              not building pcregrep.
757          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
758              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
760    25. Forward references, both numeric and by name, in patterns that made use of
761        duplicate group numbers, could behave incorrectly or give incorrect errors,
762        because when scanning forward to find the reference group, PCRE was not
763        taking into account the duplicate group numbers. A pattern such as
764        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
766    26. Changed a few more instances of "const unsigned char *" to USPTR, making
767        the feature of a custom pointer more persuasive (as requested by a user).
769    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
770        pcretest.c, inside #ifndefs, because it seems they are sometimes already
771        pre-defined.
773    28. Added support for (*UTF8) at the start of a pattern.
775    29. Arrange for flags added by the "release type" setting in CMake to be shown
776        in the configuration summary.
779    Version 7.8 05-Sep-08
780    ---------------------
782    1.  Replaced UCP searching code with optimized version as implemented for Ad
783        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
784        stage table and inline lookup instead of a function, giving speed ups of 2
785        to 5 times on some simple patterns that I tested. Permission was given to
786        distribute the MultiStage2.py script that generates the tables (it's not in
787        the tarball, but is in the Subversion repository).
789    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
790        scripts.
792    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
793        a group with a zero qualifier. The result of the study could be incorrect,
794        or the function might crash, depending on the pattern.
796    4.  Caseless matching was not working for non-ASCII characters in back
797        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
798        It now works when Unicode Property Support is available.
800    5.  In pcretest, an escape such as \x{de} in the data was always generating
801        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
802        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
803        truncation.
805    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
807    7.  Added two (int) casts to pcregrep when printing the difference of two
808        pointers, in case they are 64-bit values.
810    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
811        test 2 if it fails.
813    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
814        and a #define of that name to empty if it is not externally set. This is to
815        allow users of MSVC to set it if necessary.
817    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
818        the convenience functions in the pcre_get.c source file.
820    11. An option change at the start of a pattern that had top-level alternatives
821        could cause overwriting and/or a crash. This command provoked a crash in
822        some environments:
824          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
826        This potential security problem was recorded as CVE-2008-2371.
828    12. For a pattern where the match had to start at the beginning or immediately
829        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
830        pcre_dfa_exec() could read past the end of the passed subject if there was
831        no match. To help with detecting such bugs (e.g. with valgrind), I modified
832        pcretest so that it places the subject at the end of its malloc-ed buffer.
834    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
835        exec() might read past the end of the data buffer in UTF-8 mode.
837    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
838        the data contained the byte 0x85 as part of a UTF-8 character within its
839        first line. This applied both to normal and DFA matching.
841    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
842        /^[^d]*?$/8 failed to match "abc".
844    16. Added a missing copyright notice to pcrecpp_internal.h.
846    17. Make it more clear in the documentation that values returned from
847        pcre_exec() in ovector are byte offsets, not character counts.
849    18. Tidied a few places to stop certain compilers from issuing warnings.
851    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
852        supplied by Stefan Weber. I made a further small update for 7.8 because
853        there is a change of source arrangements: the pcre_searchfuncs.c module is
854        replaced by pcre_ucd.c.
857    Version 7.7 07-May-08
858    ---------------------
860    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
861        a string to a long long, pretend we don't even have a long long." This is
862        done by checking for the strtoq, strtoll, and _strtoi64 functions.
864    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
865        pre-7.6 versions, which defined a global no_arg variable instead of putting
866        it in the RE class. (See also #8 below.)
868    3.  Remove a line of dead code, identified by coverity and reported by Nuno
869        Lopes.
871    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
873        (1) The include/exclude patterns were being applied to the whole pathnames
874            of files, instead of just to the final components.
876        (2) If there was more than one level of directory, the subdirectories were
877            skipped unless they satisfied the include/exclude conditions. This is
878            inconsistent with GNU grep (and could even be seen as contrary to the
879            pcregrep specification - which I improved to make it absolutely clear).
880            The action now is always to scan all levels of directory, and just
881            apply the include/exclude patterns to regular files.
883    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
884        --exclude_dir in the tests to avoid scanning .svn directories.
886    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
887        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
888        doesn't support NULs in patterns.
890    7.  Added some missing "const"s to declarations of static tables in
891        pcre_compile.c and pcre_dfa_exec.c.
893    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
894        caused by fix #2  above. (Subsequently also a second patch to fix the
895        first patch. And a third patch - this was a messy problem.)
897    9.  Applied Craig's patch to remove the use of push_back().
899    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
900        matching function regexec().
902    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
903        which, however, unlike Perl's \g{...}, are subroutine calls, not back
904        references. PCRE supports relative numbers with this syntax (I don't think
905        Oniguruma does).
907    12. Previously, a group with a zero repeat such as (...){0} was completely
908        omitted from the compiled regex. However, this means that if the group
909        was called as a subroutine from elsewhere in the pattern, things went wrong
910        (an internal error was given). Such groups are now left in the compiled
911        pattern, with a new opcode that causes them to be skipped at execution
912        time.
914    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
915        to the way PCRE behaves:
917        (a) A lone ] character is dis-allowed (Perl treats it as data).
919        (b) A back reference to an unmatched subpattern matches an empty string
920            (Perl fails the current match path).
922        (c) A data ] in a character class must be notated as \] because if the
923            first data character in a class is ], it defines an empty class. (In
924            Perl it is not possible to have an empty class.) The empty class []
925            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
926            The negative empty class [^] matches any one character, independently
927            of the DOTALL setting.
929    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
930        non-existent subpattern following a character class starting with ']' and
931        containing () gave an internal compiling error instead of "reference to
932        non-existent subpattern". Fortunately, when the pattern did exist, the
933        compiled code was correct. (When scanning forwards to check for the
934        existencd of the subpattern, it was treating the data ']' as terminating
935        the class, so got the count wrong. When actually compiling, the reference
936        was subsequently set up correctly.)
938    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
939        it was being rejected as not supported by pcre_dfa_exec(), even though
940        other assertions are supported. I have made pcre_dfa_exec() support
941        (*FAIL).
943    16. The implementation of 13c above involved the invention of a new opcode,
944        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
945        cannot be changed at match time, I realized I could make a small
946        improvement to matching performance by compiling OP_ALLANY instead of
947        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
948        on the OP_ANY path.
950    17. Compiling pcretest on Windows with readline support failed without the
951        following two fixes: (1) Make the unistd.h include conditional on
952        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
954    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
955        ncurses library to be included for pcretest when ReadLine support is
956        requested, but also to allow for it to be overridden. This patch came from
957        Daniel Bergström.
959    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
960        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
961        any errors with the current Unicode tables. Thanks to Peter Kankowski for
962        spotting this.
965    Version 7.6 28-Jan-08
966    ---------------------
968    1.  A character class containing a very large number of characters with
969        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
970        overflow.
972    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
973        HAVE_LONG_LONG is not defined.
975    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
976        bring them up to date and include new features. This patch includes:
978        - Fixed PH's badly added libz and libbz2 support.
979        - Fixed a problem with static linking.
980        - Added pcredemo. [But later removed - see 7 below.]
981        - Fixed dftables problem and added an option.
982        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
983            HAVE_LONG_LONG.
984        - Added readline support for pcretest.
985        - Added an listing of the option settings after cmake has run.
987    4.  A user submitted a patch to Makefile that makes it easy to create
988        "pcre.dll" under mingw when using Configure/Make. I added stuff to
989        Makefile.am that cause it to include this special target, without
990        affecting anything else. Note that the same mingw target plus all
991        the other distribution libraries and programs are now supported
992        when configuring with CMake (see 6 below) instead of with
993        Configure/Make.
995    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
996        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
997        exported in the Windows port". It has not yet been confirmed that the patch
998        solves the problem, but it does no harm.
1000    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1001        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1002        with CMake, and also correct the comment about stack recursion.
1004    7.  Remove the automatic building of pcredemo from the ./configure system and
1005        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1006        of a program that users should build themselves after PCRE is installed, so
1007        building it automatically is not really right. What is more, it gave
1008        trouble in some build environments.
1010    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1013    Version 7.5 10-Jan-08
1014    ---------------------
1016    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1017        values in parens when parsing an RE using the C++ wrapper."
1019    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1020        Characters greater than 255 were excluded from the class instead of being
1021        included.
1023    3.  The same bug as (2) above applied to negated POSIX classes such as
1024        [:^space:].
1026    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1027        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1028        I have changed it.
1030    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1031        first named subpattern) and a construct such as (?&a) would reference the
1032        first named subpattern whose name started with "a" (in other words, the
1033        length check was missing). Both these problems are fixed. "Subpattern name
1034        expected" is now given for (?&) (a zero-length name), and this patch also
1035        makes it give the same error for \k'' (previously it complained that that
1036        was a reference to a non-existent subpattern).
1038    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1039        this is right because (?- can be followed by option settings as well as by
1040        digits. I have, however, made the messages clearer.
1042    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1043        than the number used in the conditional) now cause a compile-time error.
1044        This is actually not compatible with Perl, which accepts such patterns, but
1045        treats the conditional as always being FALSE (as PCRE used to), but it
1046        seems to me that giving a diagnostic is better.
1048    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1049        and messages.
1051    9.  Fix two occurrences of "backslash" in comments that should have been
1052        "backspace".
1054    10. Remove two redundant lines of code that can never be obeyed (their function
1055        was moved elsewhere).
1057    11. The program that makes PCRE's Unicode character property table had a bug
1058        which caused it to generate incorrect table entries for sequences of
1059        characters that have the same character type, but are in different scripts.
1060        It amalgamated them into a single range, with the script of the first of
1061        them. In other words, some characters were in the wrong script. There were
1062        thirteen such cases, affecting characters in the following ranges:
1064          U+002b0 - U+002c1
1065          U+0060c - U+0060d
1066          U+0061e - U+00612
1067          U+0064b - U+0065e
1068          U+0074d - U+0076d
1069          U+01800 - U+01805
1070          U+01d00 - U+01d77
1071          U+01d9b - U+01dbf
1072          U+0200b - U+0200f
1073          U+030fc - U+030fe
1074          U+03260 - U+0327f
1075          U+0fb46 - U+0fbb1
1076          U+10450 - U+1049d
1078    12. The -o option (show only the matching part of a line) for pcregrep was not
1079        compatible with GNU grep in that, if there was more than one match in a
1080        line, it showed only the first of them. It now behaves in the same way as
1081        GNU grep.
1083    13. If the -o and -v options were combined for pcregrep, it printed a blank
1084        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1085        does the same. The return code can be used to tell if there were any
1086        non-matching lines.
1088    14. Added --file-offsets and --line-offsets to pcregrep.
1090    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1091        infinitely looping recursion. The bug was that positive lookaheads were not
1092        being skipped when checking for a possible empty match (negative lookaheads
1093        and both kinds of lookbehind were skipped).
1095    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1096        inclusion of <windows.h> to before rather than after the definition of
1097        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1099    17. Specifying a possessive quantifier with a specific limit for a Unicode
1100        character property caused pcre_compile() to compile bad code, which led at
1101        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1102        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1103        caused the error; without that there was no problem.
1105    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1107    19. Added --enable-pcretest-libreadline.
1109    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1110        RE::GlobalReplace(). As a result, the number of replacements returned was
1111        double what it should be. I removed one of the increments, but Craig sent a
1112        later patch that removed the other one (the right fix) and added unit tests
1113        that check the return values (which was not done before).
1115    21. Several CMake things:
1117        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1118            the names libpcre and libpcreposix, not just pcre and pcreposix.
1120        (2) The above change means that pcretest and pcregrep are now correctly
1121            linked with the newly-built libraries, not previously installed ones.
1125    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1126        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1127        UTF-8 newline character). The key issue is that the pattern starts .*;
1128        this means that the match must be either at the beginning, or after a
1129        newline. The bug was in the code for advancing after a failed match and
1130        checking that the new position followed a newline. It was not taking
1131        account of UTF-8 characters correctly.
1133    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1134        character classes. PCRE was not treating the sequence [:...:] as a
1135        character class unless the ... were all letters. Perl, however, seems to
1136        allow any characters between [: and :], though of course it rejects as
1137        unknown any "names" that contain non-letters, because all the known class
1138        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1139        for example, whereas PCRE did not - it did not recognize a POSIX character
1140        class. This seemed a bit dangerous, so the code has been changed to be
1141        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1142        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1143        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1144        Perl does, and where it didn't before.
1146    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1147        Windows environments %n is disabled by default.
1150    Version 7.4 21-Sep-07
1151    ---------------------
1153    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1154        means that a class such as [\s] counted as "explicit reference to CR or
1155        LF". That isn't really right - the whole point of the change was to try to
1156        help when there was an actual mention of one of the two characters. So now
1157        the change happens only if \r or \n (or a literal CR or LF) character is
1158        encountered.
1160    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1161        of both had grown to the point where there were only 3 bits left.
1162        Fortunately, there was spare space in the data structure, and so I have
1163        moved the internal flags into a new 16-bit field to free up more option
1164        bits.
1166    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1167        but did not set the internal JCHANGED flag - either of these is enough to
1168        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1169        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1170        start sets both bits.
1172    4.  Added options (at build time, compile time, exec time) to change \R from
1173        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1175    5.  doc/pcresyntax.html was missing from the distribution.
1177    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1178        compatibility, even though it is no longer used.
1180    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1181        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1182        windows.h file is present (where different names are used). [This was
1183        reversed later after testing - see 16 below.]
1185    8.  Changed all #include <config.h> to #include "config.h". There were also
1186        some further <pcre.h> cases that I changed to "pcre.h".
1188    9.  When pcregrep was used with the --colour option, it missed the line ending
1189        sequence off the lines that it output.
1191    10. It was pointed out to me that arrays of string pointers cause lots of
1192        relocations when a shared library is dynamically loaded. A technique of
1193        using a single long string with a table of offsets can drastically reduce
1194        these. I have refactored PCRE in four places to do this. The result is
1195        dramatic:
1197          Originally:                          290
1198          After changing UCP table:            187
1199          After changing error message table:   43
1200          After changing table of "verbs"       36
1201          After changing table of Posix names   22
1203        Thanks to the folks working on Gregex for glib for this insight.
1205    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1206        unicode-properties was also set.
1208    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1210    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1211        checked only for CRLF.
1213    14. Added casts to pcretest.c to avoid compiler warnings.
1215    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1217    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1218        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1219        entirely. This removes changes made in 7 above.
1221    17. The CMake files have been updated, and there is now more information about
1222        building with CMake in the NON-UNIX-USE document.
1225    Version 7.3 28-Aug-07
1226    ---------------------
1228     1. In the rejigging of the build system that eventually resulted in 7.1, the
1229        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1230        brackets there is not right, since it causes compilers to look for an
1231        installed pcre.h, not the version that is in the source that is being
1232        compiled (which of course may be different). I have changed it back to:
1234          #include "pcre.h"
1236        I have a vague recollection that the change was concerned with compiling in
1237        different directories, but in the new build system, that is taken care of
1238        by the VPATH setting the Makefile.
1240     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1241        when the subject happened to end in the byte 0x85 (e.g. if the last
1242        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1243        characters but of course it shouldn't be taken as a newline when it is part
1244        of another character. The bug was that, for an unlimited repeat of . in
1245        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1246        characters when looking for a newline.
1248     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1250     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1251        in debug output.
1253     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1254        long printing in the pcrecpp unittest when running under MinGW.
1256     6. ESC_K was left out of the EBCDIC table.
1258     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1259        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1260        limit also applies to "virtual nesting" when a pattern is recursive, and in
1261        this case 1000 isn't so big. I have been able to remove this limit at the
1262        expense of backing off one optimization in certain circumstances. Normally,
1263        when pcre_exec() would call its internal match() function recursively and
1264        immediately return the result unconditionally, it uses a "tail recursion"
1265        feature to save stack. However, when a subpattern that can match an empty
1266        string has an unlimited repetition quantifier, it no longer makes this
1267        optimization. That gives it a stack frame in which to save the data for
1268        checking that an empty string has been matched. Previously this was taken
1269        from the 1000-entry workspace that had been reserved. So now there is no
1270        explicit limit, but more stack is used.
1272     8. Applied Daniel's patches to solve problems with the import/export magic
1273        syntax that is required for Windows, and which was going wrong for the
1274        pcreposix and pcrecpp parts of the library. These were overlooked when this
1275        problem was solved for the main library.
1277     9. There were some crude static tests to avoid integer overflow when computing
1278        the size of patterns that contain repeated groups with explicit upper
1279        limits. As the maximum quantifier is 65535, the maximum group length was
1280        set at 30,000 so that the product of these two numbers did not overflow a
1281        32-bit integer. However, it turns out that people want to use groups that
1282        are longer than 30,000 bytes (though not repeat them that many times).
1283        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1284        made it possible to implement the integer overflow checks in a much more
1285        dynamic way, which I have now done. The artificial limitation on group
1286        length has been removed - we now have only the limit on the total length of
1287        the compiled pattern, which depends on the LINK_SIZE setting.
1289    10. Fixed a bug in the documentation for get/copy named substring when
1290        duplicate names are permitted. If none of the named substrings are set, the
1291        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1292        empty string.
1294    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1295        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1296        because the ] is interpreted as the first data character and the
1297        terminating ] is not found. PCRE has been made compatible with Perl in this
1298        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1299        cause memory overwriting.
1301    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1302        string has been matched (to stop an infinite loop). It was not recognizing
1303        a conditional subpattern that could match an empty string if that
1304        subpattern was within another subpattern. For example, it looped when
1305        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1306        condition was not nested. This bug has been fixed.
1308    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1309        past the start of the subject in the presence of bytes with the top bit
1310        set, for example "\x8aBCD".
1312    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1313        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1315    14. Optimized (?!) to (*FAIL).
1317    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1318        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1319        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1320        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1321        does: it's just the validity check that is more restrictive.
1323    16. Inserted checks for integer overflows during escape sequence (backslash)
1324        processing, and also fixed erroneous offset values for syntax errors during
1325        backslash processing.
1327    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1328        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1330    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1331        caused an overrun.
1333    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1334        something other than just ASCII characters) inside a group that had an
1335        unlimited repeat caused a loop at compile time (while checking to see
1336        whether the group could match an empty string).
1338    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1339        [\P{Any}] did so. (Error in the code for printing property names.)
1341    21. An orphan \E inside a character class could cause a crash.
1343    22. A repeated capturing bracket such as (A)? could cause a wild memory
1344        reference during compilation.
1346    23. There are several functions in pcre_compile() that scan along a compiled
1347        expression for various reasons (e.g. to see if it's fixed length for look
1348        behind). There were bugs in these functions when a repeated \p or \P was
1349        present in the pattern. These operators have additional parameters compared
1350        with \d, etc, and these were not being taken into account when moving along
1351        the compiled data. Specifically:
1353        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1354            length.
1356        (b) An item such as \pL+ within a repeated group could cause crashes or
1357            loops.
1359        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1360            "reference to non-existent subpattern" error.
1362        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1364    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1365        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1367    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1369    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1370        character were causing crashes (broken optimization).
1372    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1373        \p or \P) caused a compile-time loop.
1375    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1376        break. For example, the unstudied pattern [\r\n]A does not match the string
1377        "\r\nA" because change 7.0/46 below moves the current point on by two
1378        characters after failing to match at the start. However, the pattern \nA
1379        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1380        the same is true. There doesn't seem any very clean way out of this, but
1381        what I have chosen to do makes the common cases work: PCRE now takes note
1382        of whether there can be an explicit match for \r or \n anywhere in the
1383        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1384        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1385        pattern has explicit CR or LF references.
1387    29. Added (*CR) etc for changing newline setting at start of pattern.
1390    Version 7.2 19-Jun-07
1391    ---------------------
1393     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1394        which is apparently normally available under Windows.
1396     2. Re-jig the pcregrep tests with different newline settings in an attempt
1397        to make them independent of the local environment's newline setting.
1399     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1401     4. Some of the "internals" tests were previously cut out when the link size
1402        was not 2, because the output contained actual offsets. The recent new
1403        "Z" feature of pcretest means that these can be cut out, making the tests
1404        usable with all link sizes.
1406     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1407        stack recursion. This gives a massive performance boost under BSD, but just
1408        a small improvement under Linux. However, it saves one field in the frame
1409        in all cases.
1411     6. Added more features from the forthcoming Perl 5.10:
1413        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1414            recursion call. It refers to the nth most recently opened parentheses.
1416        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1417            to be opened parentheses.
1419        (c) Conditions that refer to capturing parentheses can be specified
1420            relatively, for example, (?(-2)... or (?(+3)...
1422        (d) \K resets the start of the current match so that everything before
1423            is not part of it.
1425        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1427        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1428            reference syntax.
1430        (g) (?| introduces a group in which the numbering of parentheses in each
1431            alternative starts with the same number.
1433        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1435     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1438     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1439        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1440        for detecting groups that can match an empty string.
1442     9. A pattern with a very large number of alternatives (more than several
1443        hundred) was running out of internal workspace during the pre-compile
1444        phase, where pcre_compile() figures out how much memory will be needed. A
1445        bit of new cunning has reduced the workspace needed for groups with
1446        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1447        workspace instead of running out of the 4096 that are available.
1449    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1451    11. Applied patch from Google to remove an optimization that didn't quite work.
1452        The report of the bug said:
1454          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1455          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1456          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1458    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1459        it matched the wrong number of bytes.
1462    Version 7.1 24-Apr-07
1463    ---------------------
1465     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1466        that is more "standard", making use of automake and other Autotools. There
1467        is some re-arrangement of the files and adjustment of comments consequent
1468        on this.
1470     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1471        for recursive directory scanning broke on some systems because the files
1472        are not scanned in any specific order and on different systems the order
1473        was different. A call to "sort" has been inserted into RunGrepTest for the
1474        approprate test as a short-term fix. In the longer term there may be an
1475        alternative.
1477     3. I had an email from Eric Raymond about problems translating some of PCRE's
1478        man pages to HTML (despite the fact that I distribute HTML pages, some
1479        people do their own conversions for various reasons). The problems
1480        concerned the use of low-level troff macros .br and .in. I have therefore
1481        removed all such uses from the man pages (some were redundant, some could
1482        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1483        HTML has been updated to handle .nf/.fi and to complain if it encounters
1484        .br or .in.
1486     4. Updated comments in configure.ac that get placed in config.h.in and also
1487        arranged for config.h to be included in the distribution, with the name
1488        config.h.generic, for the benefit of those who have to compile without
1489        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1491     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1492        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1493        makevp.bat for the current PCRE, using the additional files
1494        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1496     6. A Windows user reported a minor discrepancy with test 2, which turned out
1497        to be caused by a trailing space on an input line that had got lost in his
1498        copy. The trailing space was an accident, so I've just removed it.
1500     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1501        that is needed.
1503     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1504        as "const" (a) because they are and (b) because it helps the PHP
1505        maintainers who have recently made a script to detect big data structures
1506        in the php code that should be moved to the .rodata section. I remembered
1507        to update Builducptable as well, so it won't revert if ucptable.h is ever
1508        re-created.
1510     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1511        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1512        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1513        support is not required. This saves 1.5-2K of code, which is important in
1514        some applications.
1516        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1517        so as not to refer to the tables, even though these functions will never be
1518        called when UTF-8 support is disabled. Otherwise there are problems with a
1519        shared library.
1521    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1523        (a) It was defining its arguments as char * instead of void *.
1525        (b) It was assuming that all moves were upwards in memory; this was true
1526            a long time ago when I wrote it, but is no longer the case.
1528        The emulated memove() is provided for those environments that have neither
1529        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1530        is clearly not the case, as these two bugs were recently reported.
1532    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1533        and Detrail to create the HTML documentation, the .txt form of the man
1534        pages, and it removes trailing spaces from listed files. It also creates
1535        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1536        case, it wraps all the #defines with #ifndefs. This script should be run
1537        before "make dist".
1539    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1540        with Unicode property support.
1542        (a) For a maximizing quantifier, if the two different cases of the
1543            character were of different lengths in their UTF-8 codings (there are
1544            some cases like this - I found 11), and the matching function had to
1545            back up over a mixture of the two cases, it incorrectly assumed they
1546            were both the same length.
1548        (b) When PCRE was configured to use the heap rather than the stack for
1549            recursion during matching, it was not correctly preserving the data for
1550            the other case of a UTF-8 character when checking ahead for a match
1551            while processing a minimizing repeat. If the check also involved
1552            matching a wide character, but failed, corruption could cause an
1553            erroneous result when trying to check for a repeat of the original
1554            character.
1556    13. Some tidying changes to the testing mechanism:
1558        (a) The RunTest script now detects the internal link size and whether there
1559            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1560            values substituted by "configure". (The RunGrepTest script already did
1561            this for UTF-8.) The configure.ac script no longer substitutes the
1562            relevant variables.
1564        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1565            with length and offset values. This means that the output is different
1566            for different internal link sizes. Test 2 is skipped for link sizes
1567            other than 2 because of this, bypassing the problem. Unfortunately,
1568            there was also a test in test 3 (the locale tests) that used /B and
1569            failed for link sizes other than 2. Rather than cut the whole test out,
1570            I have added a new /Z option to pcretest that replaces the length and
1571            offset values with spaces. This is now used to make test 3 independent
1572            of link size. (Test 2 will be tidied up later.)
1574    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1575        segmentation fault instead of returning the appropriate error message.
1577    15. In multiline mode when the newline sequence was set to "any", the pattern
1578        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1579        This doesn't seem right; it now treats the CRLF combination as the line
1580        ending, and so does not match in that case. It's only a pattern such as ^$
1581        that would hit this one: something like ^ABC$ would have failed after \r
1582        and then tried again after \r\n.
1584    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1585        in an attempt to make files that differ only in their line terminators
1586        compare equal. This works on Linux.
1588    17. Under certain error circumstances pcregrep might try to free random memory
1589        as it exited. This is now fixed, thanks to valgrind.
1591    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1592        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1593        was because its rules for how to advance for /g after matching an empty
1594        string at the end of a line did not allow for this case. They now check for
1595        it specially.
1597    20. pcretest is supposed to handle patterns and data of any length, by
1598        extending its buffers when necessary. It was getting this wrong when the
1599        buffer for a data line had to be extended.
1601    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1602        CRLF as a newline sequence.
1604    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1605        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1606        I have nevertheless tidied it up.
1608    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1610    24. Added a man page for pcre-config.
1613    Version 7.0 19-Dec-06
1614    ---------------------
1616     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1617        moving to gcc 4.1.1.
1619     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1620        sys/time.h, which is documented as needed for this function. It doesn't
1621        seem to matter on Linux, but it showed up on some releases of OS X.
1623     3. It seems that there are systems where bytes whose values are greater than
1624        127 match isprint() in the "C" locale. The "C" locale should be the
1625        default when a C program starts up. In most systems, only ASCII printing
1626        characters match isprint(). This difference caused the output from pcretest
1627        to vary, making some of the tests fail. I have changed pcretest so that:
1629        (a) When it is outputting text in the compiled version of a pattern, bytes
1630            other than 32-126 are always shown as hex escapes.
1632        (b) When it is outputting text that is a matched part of a subject string,
1633            it does the same, unless a different locale has been set for the match
1634            (using the /L modifier). In this case, it uses isprint() to decide.
1636     4. Fixed a major bug that caused incorrect computation of the amount of memory
1637        required for a compiled pattern when options that changed within the
1638        pattern affected the logic of the preliminary scan that determines the
1639        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1640        that the computed length was too small. The symptoms of this bug were
1641        either the PCRE error "internal error: code overflow" from pcre_compile(),
1642        or a glibc crash with a message such as "pcretest: free(): invalid next
1643        size (fast)". Examples of patterns that provoked this bug (shown in
1644        pcretest format) are:
1646          /(?-x: )/x
1647          /(?x)(?-x: \s*#\s*)/
1648          /((?i)[\x{c0}])/8
1649          /(?i:[\x{c0}])/8
1651        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1652        is now done differently.
1654     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1655        wrapper classes; (b) implement a new function in the C++ scanner that is
1656        more efficient than the old way of doing things because it avoids levels of
1657        recursion in the regex matching; (c) add a paragraph to the documentation
1658        for the FullMatch() function.
1660     6. The escape sequence \n was being treated as whatever was defined as
1661        "newline". Not only was this contrary to the documentation, which states
1662        that \n is character 10 (hex 0A), but it also went horribly wrong when
1663        "newline" was defined as CRLF. This has been fixed.
1665     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1666        was being set to -1 for the "end of line" case (supposedly a value that no
1667        character can have). Though this value is never used (the check for end of
1668        line is "zero bytes in current character"), it caused compiler complaints.
1669        I've changed it to 0xffffffff.
1671     8. In pcre_version.c, the version string was being built by a sequence of
1672        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1673        string (as it is for production releases) called a macro with an empty
1674        argument. The C standard says the result of this is undefined. The gcc
1675        compiler treats it as an empty string (which was what was wanted) but it is
1676        reported that Visual C gives an error. The source has been hacked around to
1677        avoid this problem.
1679     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1680        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1681        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1682        of them did).
1684    10. Originally, pcretest opened its input and output without "b"; then I was
1685        told that "b" was needed in some environments, so it was added for release
1686        5.0 to both the input and output. (It makes no difference on Unix-like
1687        systems.) Later I was told that it is wrong for the input on Windows. I've
1688        now abstracted the modes into two macros, to make it easier to fiddle with
1689        them, and removed "b" from the input mode under Windows.
1691    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1693    12. Added -help and --help to pcretest as an official way of being reminded
1694        of the options.
1696    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1697        and pcrecpp.cc because they annoy compilers at high warning levels.
1699    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1701    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1702        scripts are not C programs :-) and which was not noticed because it works
1703        on Linux.
1705    16. pcretest is supposed to handle any length of pattern and data line (as one
1706        line or as a continued sequence of lines) by extending its input buffer if
1707        necessary. This feature was broken for very long pattern lines, leading to
1708        a string of junk being passed to pcre_compile() if the pattern was longer
1709        than about 50K.
1711    17. I have done a major re-factoring of the way pcre_compile() computes the
1712        amount of memory needed for a compiled pattern. Previously, there was code
1713        that made a preliminary scan of the pattern in order to do this. That was
1714        OK when PCRE was new, but as the facilities have expanded, it has become
1715        harder and harder to keep it in step with the real compile phase, and there
1716        have been a number of bugs (see for example, 4 above). I have now found a
1717        cunning way of running the real compile function in a "fake" mode that
1718        enables it to compute how much memory it would need, while actually only
1719        ever using a few hundred bytes of working memory and without too many
1720        tests of the mode. This should make future maintenance and development
1721        easier. A side effect of this work is that the limit of 200 on the nesting
1722        depth of parentheses has been removed (though this was never a serious
1723        limitation, I suspect). However, there is a downside: pcre_compile() now
1724        runs more slowly than before (30% or more, depending on the pattern). I
1725        hope this isn't a big issue. There is no effect on runtime performance.
1727    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1728        newline (only possible for the last line of a file) and it was a
1729        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1731    19. Added additional timing features to pcretest. (1) The -tm option now times
1732        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1733        separate command line item, by a number that specifies the number of
1734        repeats to use when timing. The default is 50000; this gives better
1735        precision, but takes uncomfortably long for very large patterns.
1737    20. Extended pcre_study() to be more clever in cases where a branch of a
1738        subpattern has no definite first character. For example, (a*|b*)[cd] would
1739        previously give no result from pcre_study(). Now it recognizes that the
1740        first character must be a, b, c, or d.
1742    21. There was an incorrect error "recursive call could loop indefinitely" if
1743        a subpattern (or the entire pattern) that was being tested for matching an
1744        empty string contained only one non-empty item after a nested subpattern.
1745        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1746        incorrectly, because the \d was being skipped in the check.
1748    22. The pcretest program now has a new pattern option /B and a command line
1749        option -b, which is equivalent to adding /B to every pattern. This causes
1750        it to show the compiled bytecode, without the additional information that
1751        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1752        is the same as /B/I).
1754    23. A new optimization is now able automatically to treat some sequences such
1755        as a*b as a*+b. More specifically, if something simple (such as a character
1756        or a simple class like \d) has an unlimited quantifier, and is followed by
1757        something that cannot possibly match the quantified thing, the quantifier
1758        is automatically "possessified".
1760    24. A recursive reference to a subpattern whose number was greater than 39
1761        went wrong under certain circumstances in UTF-8 mode. This bug could also
1762        have affected the operation of pcre_study().
1764    25. Realized that a little bit of performance could be had by replacing
1765        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1767    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1769    27. Possessive quantifiers such as a++ were previously implemented by turning
1770        them into atomic groups such as ($>a+). Now they have their own opcodes,
1771        which improves performance. This includes the automatically created ones
1772        from 23 above.
1774    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1775        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1776        the first matched character to be a colon. This applied both to named and
1777        numbered groups.
1779    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1781    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1782        building PCRE on Windows easier, so I have included it in the distribution.
1784    31. There is now a check in pcretest against a ridiculously large number being
1785        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1786        loop, the loop is abandoned.
1788    32. Forward references to subpatterns in conditions such as (?(2)...) where
1789        subpattern 2 is defined later cause pcre_compile() to search forwards in
1790        the pattern for the relevant set of parentheses. This search went wrong
1791        when there were unescaped parentheses in a character class, parentheses
1792        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1794    33. "Subroutine" calls and backreferences were previously restricted to
1795        referencing subpatterns earlier in the regex. This restriction has now
1796        been removed.
1798    34. Added a number of extra features that are going to be in Perl 5.10. On the
1799        whole, these are just syntactic alternatives for features that PCRE had
1800        previously implemented using the Python syntax or my own invention. The
1801        other formats are all retained for compatibility.
1803        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1804            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1805            also .NET compatible.
1807        (b) A recursion or subroutine call to a named group can now be defined as
1808            (?&name) as well as (?P>name).
1810        (c) A backreference to a named group can now be defined as \k<name> or
1811            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1812            5.10, are also .NET compatible.
1814        (d) A conditional reference to a named group can now use the syntax
1815            (?(<name>) or (?('name') as well as (?(name).
1817        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1818            groups (named and numbered) that are never evaluated inline, but can be
1819            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1820            is always false. There may be only one alternative in such a group.
1822        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1823            as the simple (?(R). The condition is true only if the most recent
1824            recursion is that of the given number or name. It does not search out
1825            through the entire recursion stack.
1827        (g) The escape \gN or \g{N} has been added, where N is a positive or
1828            negative number, specifying an absolute or relative reference.
1830    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1831        some "unreachable code" warnings.
1833    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1834        things, this adds five new scripts.
1836    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1837        There were also incompatibilities regarding the handling of \Q..\E inside
1838        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1839        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1841    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1842        matches an empty string, and forcibly breaks the loop. There were bugs in
1843        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1844        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1845        separate and independent bugs (that affected different cases) have been
1846        fixed.
1848    39. Refactored the code to abolish the use of different opcodes for small
1849        capturing bracket numbers. This is a tidy that I avoided doing when I
1850        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1851        The new approach is not only tidier, it makes it possible to reduce the
1852        memory needed to fix the previous bug (38).
1854    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1855        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1856        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1857        mode.
1859    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1860        report.
1862    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1863        copy construction and assignment in the C++ wrapper.
1865    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1866        couple of bugs that could have given wrong results in the "--newline=crlf"
1867        case.
1869    44. Added a number of casts and did some reorganization of signed/unsigned int
1870        variables following suggestions from Dair Grant. Also renamed the variable
1871        "this" as "item" because it is a C++ keyword.
1873    45. Arranged for dftables to add
1875          #include "pcre_internal.h"
1877        to pcre_chartables.c because without it, gcc 4.x may remove the array
1878        definition from the final binary if PCRE is built into a static library and
1879        dead code stripping is activated.
1881    46. For an unanchored pattern, if a match attempt fails at the start of a
1882        newline sequence, and the newline setting is CRLF or ANY, and the next two
1883        characters are CRLF, advance by two characters instead of one.
1886    Version 6.7 04-Jul-06
1887    ---------------------
1889     1. In order to handle tests when input lines are enormously long, pcretest has
1890        been re-factored so that it automatically extends its buffers when
1891        necessary. The code is crude, but this _is_ just a test program. The
1892        default size has been increased from 32K to 50K.
1894     2. The code in pcre_study() was using the value of the re argument before
1895        testing it for NULL. (Of course, in any sensible call of the function, it
1896        won't be NULL.)
1898     3. The memmove() emulation function in pcre_internal.h, which is used on
1899        systems that lack both memmove() and bcopy() - that is, hardly ever -
1900        was missing a "static" storage class specifier.
1902     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1903        containing an extended class (one that cannot be represented by a bitmap
1904        because it contains high-valued characters or Unicode property items, e.g.
1905        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1906        pattern, but PCRE should not loop if you do not (it no longer does).
1907        [Detail: two cases were found: (a) a repeated subpattern containing an
1908        extended class; (b) a recursive reference to a subpattern that followed a
1909        previous extended class. It wasn't skipping over the extended class
1910        correctly when UTF-8 mode was not set.]
1912     5. A negated single-character class was not being recognized as fixed-length
1913        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1914        compile error "lookbehind assertion is not fixed length".
1916     6. The RunPerlTest auxiliary script was showing an unexpected difference
1917        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1918        write a Perl script that can interpret lines of an input file either as
1919        byte characters or as UTF-8, which is what "perltest" was being required to
1920        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1921        can't do is switch easily at run time between having the "use utf8;" pragma
1922        or not. In the end, I fudged it by using the RunPerlTest script to insert
1923        "use utf8;" explicitly for the UTF-8 tests.
1925     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1926        the end of the subject string, contrary to the documentation and to what
1927        Perl does. This was true of both matching functions. Now it matches only at
1928        the start of the subject and immediately after *internal* newlines.
1930     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1931        a pointer to an int instead of a pointer to an unsigned long int. This
1932        caused problems on 64-bit systems.
1934     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1935        instance of the 'standard' template library not being so standard".
1937    10. There was no check on the number of named subpatterns nor the maximum
1938        length of a subpattern name. The product of these values is used to compute
1939        the size of the memory block for a compiled pattern. By supplying a very
1940        long subpattern name and a large number of named subpatterns, the size
1941        computation could be caused to overflow. This is now prevented by limiting
1942        the length of names to 32 characters, and the number of named subpatterns
1943        to 10,000.
1945    11. Subpatterns that are repeated with specific counts have to be replicated in
1946        the compiled pattern. The size of memory for this was computed from the
1947        length of the subpattern and the repeat count. The latter is limited to
1948        65535, but there was no limit on the former, meaning that integer overflow
1949        could in principle occur. The compiled length of a repeated subpattern is
1950        now limited to 30,000 bytes in order to prevent this.
1952    12. Added the optional facility to have named substrings with the same name.
1954    13. Added the ability to use a named substring as a condition, using the
1955        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1956        are numbers (not recommended). Forward references are permitted.
1958    14. Added forward references in named backreferences (if you see what I mean).
1960    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1961        pattern could run off the end of the subject. For example, the pattern
1962        "(?s)(.{1,5})"8 did this with the subject "ab".
1964    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1965        PCRE_CASELESS was set when matching characters that were quantified with ?
1966        or *.
1968    17. A character class other than a single negated character that had a minimum
1969        but no maximum quantifier - for example [ab]{6,} - was not handled
1970        correctly by pce_dfa_exec(). It would match only one character.
1972    18. A valid (though odd) pattern that looked like a POSIX character
1973        class but used an invalid character after [ (for example [[,abc,]]) caused
1974        pcre_compile() to give the error "Failed: internal error: code overflow" or
1975        in some cases to crash with a glibc free() error. This could even happen if
1976        the pattern terminated after [[ but there just happened to be a sequence of
1977        letters, a binary zero, and a closing ] in the memory that followed.
1979    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1980        over the years. Originally (before any Unicode support), just the bottom 8
1981        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1982        output from "man perlunicode" includes this:
1984          The regular expression compiler produces polymorphic opcodes.  That
1985          is, the pattern adapts to the data and automatically switches to
1986          the Unicode character scheme when presented with Unicode data--or
1987          instead uses a traditional byte scheme when presented with byte
1988          data.
1990        Sadly, a wide octal escape does not cause a switch, and in a string with
1991        no other multibyte characters, these octal escapes are treated as before.
1992        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1993        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1994        Unicode string.
1996        I have not perpetrated such confusion in PCRE. Up till now, it took just
1997        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1998        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1999        translate to the appropriate multibyte character.
2001    29. Applied some refactoring to reduce the number of warnings from Microsoft
2002        and Borland compilers. This has included removing the fudge introduced
2003        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2004        a warning about an unused variable.
2006    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2007        characters since release 4.0, because Perl (from release 5.004) does not.
2008        [Or at least, is documented not to: some releases seem to be in conflict
2009        with the documentation.] However, when a pattern was studied with
2010        pcre_study() and all its branches started with \s, PCRE still included VT
2011        as a possible starting character. Of course, this did no harm; it just
2012        caused an unnecessary match attempt.
2014    22. Removed a now-redundant internal flag bit that recorded the fact that case
2015        dependency changed within the pattern. This was once needed for "required
2016        byte" processing, but is no longer used. This recovers a now-scarce options
2017        bit. Also moved the least significant internal flag bit to the most-
2018        significant bit of the word, which was not previously used (hangover from
2019        the days when it was an int rather than a uint) to free up another bit for
2020        the future.
2022    23. Added support for CRLF line endings as well as CR and LF. As well as the
2023        default being selectable at build time, it can now be changed at runtime
2024        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2025        specify that it is scanning data with non-default line endings.
2027    24. Changed the definition of CXXLINK to make it agree with the definition of
2028        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2030    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2031        recursions. This makes a big different to stack usage for some patterns.
2033    26. If a subpattern containing a named recursion or subroutine reference such
2034        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2035        the space required for the compiled pattern went wrong and gave too small a
2036        value. Depending on the environment, this could lead to "Failed: internal
2037        error: code overflow at offset 49" or "glibc detected double free or
2038        corruption" errors.
2040    27. Applied patches from Google (a) to support the new newline modes and (b) to
2041        advance over multibyte UTF-8 characters in GlobalReplace.
2043    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2044        difference for some implementation of PCRE in some Windows version.
2046    29. Added some extra testing facilities to pcretest:
2048        \q<number>   in a data line sets the "match limit" value
2049        \Q<number>   in a data line sets the "match recursion limt" value
2050        -S <number>  sets the stack size, where <number> is in megabytes
2052        The -S option isn't available for Windows.
2055    Version 6.6 06-Feb-06
2056    ---------------------
2058     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2059        in pcreposix.h. I have copied the definition from pcre.h.
2061     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2062        because pcre.h is no longer a built file.
2064     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2065        not normally included in the compiled code.
2068    Version 6.5 01-Feb-06
2069    ---------------------
2071     1. When using the partial match feature with pcre_dfa_exec(), it was not
2072        anchoring the second and subsequent partial matches at the new starting
2073        point. This could lead to incorrect results. For example, with the pattern
2074        /1234/, partially matching against "123" and then "a4" gave a match.
2076     2. Changes to pcregrep:
2078        (a) All non-match returns from pcre_exec() were being treated as failures
2079            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2080            error message is output. Some extra information is given for the
2081            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2082            probably the only errors that are likely to be caused by users (by
2083            specifying a regex that has nested indefinite repeats, for instance).
2084            If there are more than 20 of these errors, pcregrep is abandoned.
2086        (b) A binary zero was treated as data while matching, but terminated the
2087            output line if it was written out. This has been fixed: binary zeroes
2088            are now no different to any other data bytes.
2090        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2091            used to set a locale for matching. The --locale=xxxx long option has
2092            been added (no short equivalent) to specify a locale explicitly on the
2093            pcregrep command, overriding the environment variables.
2095        (d) When -B was used with -n, some line numbers in the output were one less
2096            than they should have been.
2098        (e) Added the -o (--only-matching) option.
2100        (f) If -A or -C was used with -c (count only), some lines of context were
2101            accidentally printed for the final match.
2103        (g) Added the -H (--with-filename) option.
2105        (h) The combination of options -rh failed to suppress file names for files
2106            that were found from directory arguments.
2108        (i) Added the -D (--devices) and -d (--directories) options.
2110        (j) Added the -F (--fixed-strings) option.
2112        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2114        (l) Added the --colo(u)r option.
2116        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2117            is not present by default.
2119     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2120        items such as (?R) or (?1), when the recursion could match a number of
2121        alternatives. If it matched one of the alternatives, but subsequently,
2122        outside the recursion, there was a failure, the code tried to back up into
2123        the recursion. However, because of the way PCRE is implemented, this is not
2124        possible, and the result was an incorrect result from the match.
2126        In order to prevent this happening, the specification of recursion has
2127        been changed so that all such subpatterns are automatically treated as
2128        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2130     4. I had overlooked the fact that, in some locales, there are characters for
2131        which isalpha() is true but neither isupper() nor islower() are true. In
2132        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2133        and ordfeminine) are like this. This affected the treatment of \w and \W
2134        when they appeared in character classes, but not when they appeared outside
2135        a character class. The bit map for "word" characters is now created
2136        separately from the results of isalnum() instead of just taking it from the
2137        upper, lower, and digit maps. (Plus the underscore character, of course.)
2139     5. The above bug also affected the handling of POSIX character classes such as
2140        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2141        permanent tables. Instead, the bit maps for such a class were previously
2142        created as the appropriate unions of the upper, lower, and digit bitmaps.
2143        Now they are created by subtraction from the [[:word:]] class, which has
2144        its own bitmap.
2146     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2147        It is created by subtracting the vertical space characters (\x09, \x0a,
2148        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2149        subtraction was done in the overall bitmap for a character class, meaning
2150        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2151        be recognized. This bug has been fixed.
2153     7. Patches from the folks at Google:
2155          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2156          real life, but is still worth protecting against".
2158          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2159          regular expressions".
2161          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2162          have it.
2164          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2165          "configure" and the latter not, in order to fix a problem somebody had
2166          with compiling the Arg class on HP-UX.
2168          (e) Improve the error-handling of the C++ wrapper a little bit.
2170          (f) New tests for checking recursion limiting.
2172     8. The pcre_memmove() function, which is used only if the environment does not
2173        have a standard memmove() function (and is therefore rarely compiled),
2174        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2175        returning a result (though PCRE never actually uses the result).
2177     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2178        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2179        returned instead of calling malloc() with an overflowing number that would
2180        most likely cause subsequent chaos.
2182    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2184    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2185        with this option is matched, the nmatch and pmatch options of regexec() are
2186        ignored.
2188    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2189        provided in case anyone wants to the the POSIX interface with UTF-8
2190        strings.
2192    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2193        C++ linking (needed for some HP-UX environments).
2195    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2196        (unused parameter) and in the pcre_printint() function (omitted "default"
2197        switch label when the default is to do nothing).
2199    15. Added some code to make it possible, when PCRE is compiled as a C++
2200        library, to replace subject pointers for pcre_exec() with a smart pointer
2201        class, thus making it possible to process discontinuous strings.
2203    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2204        much the same function. They were added by different people who were trying
2205        to make PCRE easy to compile on non-Unix systems. It has been suggested
2206        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2207        for compiling on Windows systems. I have therefore replaced it with
2208        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2209        defaults to "extern" for C or "extern C" for C++, which works fine on
2210        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2211        SCOPE with something explicit in config.h. In addition:
2213        (a) pcreposix.h still had just "extern" instead of either of these macros;
2214            I have replaced it with PCRE_DATA_SCOPE.
2216        (b) Functions such as _pcre_xclass(), which are internal to the library,
2217            but external in the C sense, all had PCRE_EXPORT in their definitions.
2218            This is apparently wrong for the Windows case, so I have removed it.
2219            (It makes no difference on Unix-like systems.)
2221    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2222        of recursive calls to match(). This is different to MATCH_LIMIT because
2223        that limits the total number of calls to match(), not all of which increase
2224        the depth of recursion. Limiting the recursion depth limits the amount of
2225        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2226        when PCRE is compiled, and changed at run time. A patch from Google adds
2227        this functionality to the C++ interface.
2229    18. Changes to the handling of Unicode character properties:
2231        (a) Updated the table to Unicode 4.1.0.
2233        (b) Recognize characters that are not in the table as "Cn" (undefined).
2235        (c) I revised the way the table is implemented to a much improved format
2236            which includes recognition of ranges. It now supports the ranges that
2237            are defined in UnicodeData.txt, and it also amalgamates other
2238            characters into ranges. This has reduced the number of entries in the
2239            table from around 16,000 to around 3,000, thus reducing its size
2240            considerably. I realized I did not need to use a tree structure after
2241            all - a binary chop search is just as efficient. Having reduced the
2242            number of entries, I extended their size from 6 bytes to 8 bytes to
2243            allow for more data.
2245        (d) Added support for Unicode script names via properties such as \p{Han}.
2247    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2248        matching that character.
2250    20. When matching a repeated Unicode property with a minimum greater than zero,
2251        (for example \pL{2,}), PCRE could look past the end of the subject if it
2252        reached it while seeking the minimum number of characters. This could
2253        happen only if some of the characters were more than one byte long, because
2254        there is a check for at least the minimum number of bytes.
2256    21. Refactored the implementation of \p and \P so as to be more general, to
2257        allow for more different types of property in future. This has changed the
2258        compiled form incompatibly. Anybody with saved compiled patterns that use
2259        \p or \P will have to recompile them.
2261    22. Added "Any" and "L&" to the supported property types.
2263    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2264        but give a compile time error if the value is greater than 0xff.
2266    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2267        accidentally not being installed or uninstalled.
2269    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2270        made were to insert the current release number. This seemed silly, because
2271        it made things harder for people building PCRE on systems that don't run
2272        "configure". I have turned pcre.h into a distributed file, no longer built
2273        by "configure", with the version identification directly included. There is
2274        no longer a pcre.h.in file.
2276        However, this change necessitated a change to the pcre-config script as
2277        well. It is built from pcre-config.in, and one of the substitutions was the
2278        release number. I have updated configure.ac so that ./configure now finds
2279        the release number by grepping pcre.h.
2281    26. Added the ability to run the tests under valgrind.
2284    Version 6.4 05-Sep-05
2285    ---------------------
2287     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2288        "--" to be printed when multiple files were scanned, even when none of the
2289        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2290        consider it to be a bug, and have restored the previous behaviour.
2292     2. A couple of code tidies to get rid of compiler warnings.
2294     3. The pcretest program used to cheat by referring to symbols in the library
2295        whose names begin with _pcre_. These are internal symbols that are not
2296        really supposed to be visible externally, and in some environments it is
2297        possible to suppress them. The cheating is now confined to including
2298        certain files from the library's source, which is a bit cleaner.
2300     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2301        file's purpose clearer.
2303     5. Reorganized pcre_ucp_findchar().
2306    Version 6.3 15-Aug-05
2307    ---------------------
2309     1. The file libpcre.pc.in did not have general read permission in the tarball.
2311     2. There were some problems when building without C++ support:
2313        (a) If C++ support was not built, "make install" and "make test" still
2314            tried to test it.
2316        (b) There were problems when the value of CXX was explicitly set. Some
2317            changes have been made to try to fix these, and ...
2319        (c) --disable-cpp can now be used to explicitly disable C++ support.
2321        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2322            backslash in a target when C++ was disabled. This confuses some
2323            versions of "make", apparently. Using an intermediate variable solves
2324            this. (Same for CPP_LOBJ.)
2326     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2327        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2328        necessary on certain architectures.
2330     4. Added a setting of -export-symbols-regex to the link command to remove
2331        those symbols that are exported in the C sense, but actually are local
2332        within the library, and not documented. Their names all begin with
2333        "_pcre_". This is not a perfect job, because (a) we have to except some
2334        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2335        available (and never for static libraries). I have made a note to try to
2336        find a way round (a) in the future.
2339    Version 6.2 01-Aug-05
2340    ---------------------
2342     1. There was no test for integer overflow of quantifier values. A construction
2343        such as {1111111111111111} would give undefined results. What is worse, if
2344        a minimum quantifier for a parenthesized subpattern overflowed and became
2345        negative, the calculation of the memory size went wrong. This could have
2346        led to memory overwriting.
2348     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2350     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2351        operating environments where this matters.
2353     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2354        PCRE options from within the C++ wrapper.
2356     5. Named capturing subpatterns were not being correctly counted when a pattern
2357        was compiled. This caused two problems: (a) If there were more than 100
2358        such subpatterns, the calculation of the memory needed for the whole
2359        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2360        back references of the form \12, where the number was greater than 9, were
2361        not recognized as back references, even though there were sufficient
2362        previous subpatterns.
2364     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2365        versions of gcc, e.g. 2.95.4.
2368    Version 6.1 21-Jun-05
2369    ---------------------
2371     1. There was one reference to the variable "posix" in pcretest.c that was not
2372        surrounded by "#if !defined NOPOSIX".
2374     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2375        the cross-check on the old pcre_info() function, for the benefit of the
2376        cut-down version of PCRE that is currently imported into Exim.
2378     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2379        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2380        patterns that start (?i) but I don't think that matters. The use of (?i) is
2381        just an example; this all applies to the other options as well.
2383     4. Since libtool seems to echo the compile commands it is issuing, the output
2384        from "make" can be reduced a bit by putting "@" in front of each libtool
2385        compile command.
2387     5. Patch from the folks at Google for configure.in to be a bit more thorough
2388        in checking for a suitable C++ installation before trying to compile the
2389        C++ stuff. This should fix a reported problem when a compiler was present,
2390        but no suitable headers.
2392     6. The man pages all had just "PCRE" as their title. I have changed them to
2393        be the relevant file name. I have also arranged that these names are
2394        retained in the file doc/pcre.txt, which is a concatenation in text format
2395        of all the man pages except the little individual ones for each function.
2397     7. The NON-UNIX-USE file had not been updated for the different set of source
2398        files that come with release 6. I also added a few comments about the C++
2399        wrapper.
2402    Version 6.0 07-Jun-05
2403    ---------------------
2405     1. Some minor internal re-organization to help with my DFA experiments.
2407     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2408        didn't matter for the library itself when fully configured, but did matter
2409        when compiling without UCP support, or within Exim, where the ucp files are
2410        not imported.
2412     3. Refactoring of the library code to split up the various functions into
2413        different source modules. The addition of the new DFA matching code (see
2414        below) to a single monolithic source would have made it really too
2415        unwieldy, quite apart from causing all the code to be include in a
2416        statically linked application, when only some functions are used. This is
2417        relevant even without the DFA addition now that patterns can be compiled in
2418        one application and matched in another.
2420        The downside of splitting up is that there have to be some external
2421        functions and data tables that are used internally in different modules of
2422        the library but which are not part of the API. These have all had their
2423        names changed to start with "_pcre_" so that they are unlikely to clash
2424        with other external names.
2426     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2427        a different (DFA) algorithm. Although it is slower than the original
2428        function, it does have some advantages for certain types of matching
2429        problem.
2431     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2432        including restarting after a partial match.
2434     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2435        defined when compiling for Windows was sent to me. I have put it into the
2436        code, though I have no means of testing or verifying it.
2438     7. Added the pcre_refcount() auxiliary function.
2440     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2441        match before or at the first newline in the subject string. In pcretest,
2442        the /f option on a pattern can be used to set this.
2444     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2445        would behave wrongly. This has been present in PCRE since release 4.0.
2447    10. A number of changes to the pcregrep command:
2449        (a) Refactored how -x works; insert ^(...)$ instead of setting
2450            PCRE_ANCHORED and checking the length, in preparation for adding
2451            something similar for -w.
2453        (b) Added the -w (match as a word) option.
2455        (c) Refactored the way lines are read and buffered so as to have more
2456            than one at a time available.
2458        (d) Implemented a pcregrep test script.
2460        (e) Added the -M (multiline match) option. This allows patterns to match
2461            over several lines of the subject. The buffering ensures that at least
2462            8K, or the rest of the document (whichever is the shorter) is available
2463            for matching (and similarly the previous 8K for lookbehind assertions).
2465        (f) Changed the --help output so that it now says
2467              -w, --word-regex(p)
2469            instead of two lines, one with "regex" and the other with "regexp"
2470            because that confused at least one person since the short forms are the
2471            same. (This required a bit of code, as the output is generated
2472            automatically from a table. It wasn't just a text change.)
2474        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2475            option but starts with a hyphen. Could be a pattern or a path name
2476            starting with a hyphen, for instance.
2478        (h) "-" can be given as a file name to represent stdin.
2480        (i) When file names are being printed, "(standard input)" is used for
2481            the standard input, for compatibility with GNU grep. Previously
2482            "<stdin>" was used.
2484        (j) The option --label=xxx can be used to supply a name to be used for
2485            stdin when file names are being printed. There is no short form.
2487        (k) Re-factored the options decoding logic because we are going to add
2488            two more options that take data. Such options can now be given in four
2489            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2491        (l) Added the -A, -B, and -C options for requesting that lines of context
2492            around matches be printed.
2494        (m) Added the -L option to print the names of files that do not contain
2495            any matching lines, that is, the complement of -l.
2497        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2498            continue to scan other files.
2500        (o) The -s option was incorrectly implemented. For compatibility with other
2501            greps, it now suppresses the error message for a non-existent or non-
2502            accessible file (but not the return code). There is a new option called
2503            -q that suppresses the output of matching lines, which was what -s was
2504            previously doing.
2506        (p) Added --include and --exclude options to specify files for inclusion
2507            and exclusion when recursing.
2509    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2510        Hopefully, it now does.
2512    12. Missing cast in pcre_study().
2514    13. Added an "uninstall" target to the makefile.
2516    14. Replaced "extern" in the function prototypes in Makefile.in with
2517        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2518        world, but is set differently for Windows.
2520    15. Added a second compiling function called pcre_compile2(). The only
2521        difference is that it has an extra argument, which is a pointer to an
2522        integer error code. When there is a compile-time failure, this is set
2523        non-zero, in addition to the error test pointer being set to point to an
2524        error message. The new argument may be NULL if no error number is required
2525        (but then you may as well call pcre_compile(), which is now just a
2526        wrapper). This facility is provided because some applications need a
2527        numeric error indication, but it has also enabled me to tidy up the way
2528        compile-time errors are handled in the POSIX wrapper.
2530    16. Added VPATH=.libs to the makefile; this should help when building with one
2531        prefix path and installing with another. (Or so I'm told by someone who
2532        knows more about this stuff than I do.)
2534    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2535        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2536        match everything, including newlines. This is not POSIX-compatible, but
2537        somebody wanted the feature. From pcretest it can be activated by using
2538        both the P and the s flags.
2540    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2542    19. libpcre.pc was being incorrectly installed as executable.
2544    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2545        it now also looks for '\r' so that it will work unmodified on Windows.
2547    21. Added Google's contributed C++ wrapper to the distribution.
2549    22. Added some untidy missing memory free() calls in pcretest, to keep
2550        Electric Fence happy when testing.
2554    Version 5.0 13-Sep-04
2555    ---------------------
2557     1. Internal change: literal characters are no longer packed up into items
2558        containing multiple characters in a single byte-string. Each character
2559        is now matched using a separate opcode. However, there may be more than one
2560        byte in the character in UTF-8 mode.
2562     2. The pcre_callout_block structure has two new fields: pattern_position and
2563        next_item_length. These contain the offset in the pattern to the next match
2564        item, and its length, respectively.
2566     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2567        insertion of callouts before each pattern item. Added the /C option to
2568        pcretest to make use of this.
2570     4. On the advice of a Windows user, the lines
2572          #if defined(_WIN32) || defined(WIN32)
2573          _setmode( _fileno( stdout ), 0x8000 );
2574          #endif  /* defined(_WIN32) || defined(WIN32) */
2576        have been added to the source of pcretest. This apparently does useful
2577        magic in relation to line terminators.
2579     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2580        for the benefit of those environments where the "b" makes a difference.
2582     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2583        to know about it. I have put a hack into configure.in that adds in code
2584        to set GCC=yes if CC=icc. This seems to end up at a point in the
2585        generated configure script that is early enough to affect the setting of
2586        compiler options, which is what is needed, but I have no means of testing
2587        whether it really works. (The user who reported this had patched the
2588        generated configure script, which of course I cannot do.)
2590        LATER: After change 22 below (new libtool files), the configure script
2591        seems to know about icc (and also ecc). Therefore, I have commented out
2592        this hack in configure.in.
2594     7. Added support for pkg-config (2 patches were sent in).
2596     8. Negated POSIX character classes that used a combination of internal tables
2597        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2598        [[:^ascii]]. Typically, they would match almost any characters. The other
2599        POSIX classes were not broken in this way.
2601     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2602        to find the match, as PCRE was deluded into thinking that the match had to
2603        start at the start point or following a newline. The same bug applied to
2604        patterns with negative forward assertions or any backward assertions
2605        preceding ".*" at the start, unless the pattern required a fixed first
2606        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2608    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2609        starting at the last subject character, bytes beyond the end of the subject
2610        string were read.
2612    11. Renamed the variable "class" as "classbits" to make life easier for C++
2613        users. (Previously there was a macro definition, but it apparently wasn't
2614        enough.)
2616    12. Added the new field "tables" to the extra data so that tables can be passed
2617        in at exec time, or the internal tables can be re-selected. This allows
2618        a compiled regex to be saved and re-used at a later time by a different
2619        program that might have everything at different addresses.
2621    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2622        -R library as well as a -L library.
2624    14. The debugging options of pcretest (-d on the command line or D on a
2625        pattern) showed incorrect output for anything following an extended class
2626        that contained multibyte characters and which was followed by a quantifier.
2628    15. Added optional support for general category Unicode character properties
2629        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2630        support. It adds about 90K to the size of the library. The meanings of the
2631        inbuilt class escapes such as \d and \s have NOT been changed.
2633    16. Updated pcredemo.c to include calls to free() to release the memory for the
2634        compiled pattern.
2636    17. The generated file chartables.c was being created in the source directory
2637        instead of in the building directory. This caused the build to fail if the
2638        source directory was different from the building directory, and was
2639        read-only.
2641    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2642        file. No doubt somebody will tell me if they don't make sense... Also added
2643        Dan Mooney's comments about building on OpenVMS.
2645    19. Added support for partial matching via the PCRE_PARTIAL option for
2646        pcre_exec() and the \P data escape in pcretest.
2648    20. Extended pcretest with 3 new pattern features:
2650        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2651              write the compiled pattern to the file whose name is "rest-of-line".
2652              This is a straight binary dump of the data, with the saved pointer to
2653              the character tables forced to be NULL. The study data, if any, is
2654              written too. After writing, pcretest reads a new pattern.
2656        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2657              compiled pattern from the given file. There must not be any
2658              occurrences of "<" in the file name (pretty unlikely); if there are,
2659              pcretest will instead treat the initial "<" as a pattern delimiter.
2660              After reading in the pattern, pcretest goes on to read data lines as
2661              usual.
2663        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2664              and 16-bit fields in a compiled pattern, to simulate a pattern that
2665              was compiled on a host of opposite endianness.
2667    21. The pcre-exec() function can now cope with patterns that were compiled on
2668        hosts of opposite endianness, with this restriction:
2670          As for any compiled expression that is saved and used later, the tables
2671          pointer field cannot be preserved; the extra_data field in the arguments
2672          to pcre_exec() should be used to pass in a tables address if a value
2673          other than the default internal tables were used at compile time.
2675    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2676        now diagnosed as an error. Previously, most of the time, a negative number
2677        would have been treated as zero, but if in addition "ovector" was passed as
2678        NULL, a crash could occur.
2680    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2681        new versions from the libtool 1.5 distribution (the last one is a copy of
2682        a file called libtool.m4). This seems to have fixed the need to patch
2683        "configure" to support Darwin 1.3 (which I used to do). However, I still
2684        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2685        workstation).
2687    24. Changed the PCRE licence to be the more standard "BSD" licence.
2690    Version 4.5 01-Dec-03
2691    ---------------------
2693     1. There has been some re-arrangement of the code for the match() function so
2694        that it can be compiled in a version that does not call itself recursively.
2695        Instead, it keeps those local variables that need separate instances for
2696        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2697        needs to "recurse". Keeping track of where control must go is done by means
2698        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2699        hide most of the details from the main code, and operates only if
2700        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2701        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2702        operating.
2704        To make it easier for callers to provide specially tailored get/free
2705        functions for this usage, two new functions, pcre_stack_malloc, and
2706        pcre_stack_free, are used. They are always called in strict stacking order,
2707        and the size of block requested is always the same.
2709        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2710        PCRE has been compiled to use the stack or the heap for recursion. The
2711        -C option of pcretest uses this to show which version is compiled.
2713        A new data escape \S, is added to pcretest; it causes the amounts of store
2714        obtained and freed by both kinds of malloc/free at match time to be added
2715        to the output.
2717     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2718        what's available on my current Linux desktop machine.
2720     3. When matching a UTF-8 string, the test for a valid string at the start has
2721        been extended. If start_offset is not zero, PCRE now checks that it points
2722        to a byte that is the start of a UTF-8 character. If not, it returns
2723        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2724        this is necessary because there may be backward assertions in the pattern.
2725        When matching the same subject several times, it may save resources to use
2726        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2728     4. The code for checking the validity of UTF-8 strings has been tightened so
2729        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2730        containing "overlong sequences".
2732     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2733        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2734        should have been "&", but it just so happened that all the cases this let
2735        through by mistake were picked up later in the function.
2737     6. I had used a variable called "isblank" - this is a C99 function, causing
2738        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2740     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2741        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2742        so the version is shown; (c) stop "make test" echoing "./RunTest".
2744     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2746     9. The prototype for memmove() for systems that don't have it was using
2747        size_t, but the inclusion of the header that defines size_t was later. I've
2748        moved the #includes for the C headers earlier to avoid this.
2750    10. Added some adjustments to the code to make it easier to compiler on certain
2751        special systems:
2753          (a) Some "const" qualifiers were missing.
2754          (b) Added the macro EXPORT before all exported functions; by default this
2755              is defined to be empty.
2756          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2757              that it reads its output file name as an argument instead of writing
2758              to the standard output and assuming this can be redirected.
2760    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2761        class containing characters with values greater than 255, PCRE compilation
2762        went into a loop.
2764    12. A recursive reference to a subpattern that was within another subpattern
2765        that had a minimum quantifier of zero caused PCRE to crash. For example,
2766        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2767        recursion. If the recursively-called subpattern itself had a zero repeat,
2768        that was OK.
2770    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2771        buffer into which it was copied (for escape processing) was still set at
2772        1024, so long lines caused crashes.
2774    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2775        "internal error: code overflow...". This applied to any character class
2776        that was followed by a possessive quantifier.
2778    15. Modified the Makefile to add libpcre.la as a prerequisite for
2779        libpcreposix.la because I was told this is needed for a parallel build to
2780        work.
2782    16. If a pattern that contained .* following optional items at the start was
2783        studied, the wrong optimizing data was generated, leading to matching
2784        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2785        matching string must start with a or b or c. The correct conclusion for
2786        this pattern is that a match can start with any character.
2789    Version 4.4 13-Aug-03
2790    ---------------------
2792     1. In UTF-8 mode, a character class containing characters with values between
2793        127 and 255 was not handled correctly if the compiled pattern was studied.
2794        In fixing this, I have also improved the studying algorithm for such
2795        classes (slightly).
2797     2. Three internal functions had redundant arguments passed to them. Removal
2798        might give a very teeny performance improvement.
2800     3. Documentation bug: the value of the capture_top field in a callout is *one
2801        more than* the number of the hightest numbered captured substring.
2803     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2804        in incorrectly linking with a previously installed version. They now link
2805        explicitly with libpcre.la.
2807     5. configure.in no longer needs to recognize Cygwin specially.
2809     6. A problem in pcre.in for Windows platforms is fixed.
2811     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2812        pcretest, it used to include the size of the study block as part of its
2813        output. Unfortunately, the structure contains a field that has a different
2814        size on different hardware architectures. This meant that the tests that
2815        showed this size failed. As the block is currently always of a fixed size,
2816        this information isn't actually particularly useful in pcretest output, so
2817        I have just removed it.
2819     8. Three pre-processor statements accidentally did not start in column 1.
2820        Sadly, there are *still* compilers around that complain, even though
2821        standard C has not required this for well over a decade. Sigh.
2823     9. In pcretest, the code for checking callouts passed small integers in the
2824        callout_data field, which is a void * field. However, some picky compilers
2825        complained about the casts involved for this on 64-bit systems. Now
2826        pcretest passes the address of the small integer instead, which should get
2827        rid of the warnings.
2829    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2830        both compile and run time, and gives an error if an invalid UTF-8 sequence
2831        is found. There is a option for disabling this check in cases where the
2832        string is known to be correct and/or the maximum performance is wanted.
2834    11. In response to a bug report, I changed one line in Makefile.in from
2836            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2837        to
2838            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2840        to look similar to other lines, but I have no way of telling whether this
2841        is the right thing to do, as I do not use Windows. No doubt I'll get told
2842        if it's wrong...
2845    Version 4.3 21-May-03
2846    ---------------------
2848    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2849       Makefile.
2851    2. Some refactoring to improve the quality of the code:
2853       (i)   The utf8_table... variables are now declared "const".
2855       (ii)  The code for \cx, which used the "case flipping" table to upper case
2856             lower case letters, now just substracts 32. This is ASCII-specific,
2857             but the whole concept of \cx is ASCII-specific, so it seems
2858             reasonable.
2860       (iii) PCRE was using its character types table to recognize decimal and
2861             hexadecimal digits in the pattern. This is silly, because it handles
2862             only 0-9, a-f, and A-F, but the character types table is locale-
2863             specific, which means strange things might happen. A private
2864             table is now used for this - though it costs 256 bytes, a table is
2865             much faster than multiple explicit tests. Of course, the standard
2866             character types table is still used for matching digits in subject
2867             strings against \d.
2869       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2870             ending in _t are). So I've renamed it as ESC_tee.
2872    3. The first argument for regexec() in the POSIX wrapper should have been
2873       defined as "const".
2875    4. Changed pcretest to use malloc() for its buffers so that they can be
2876       Electric Fenced for debugging.
2878    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2879       to read one or more bytes before the start of the subject string. Often this
2880       had no effect on PCRE's behaviour, but in some circumstances it could
2881       provoke a segmentation fault.
2883    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2884       to try to read one or more bytes before the start of the subject string.
2886    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2887       UTF-8 support could misbehave in various ways if the subject string
2888       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2889       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2890       back over UTF-8 characters.)
2893    Version 4.2 14-Apr-03
2894    ---------------------
2896    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2898    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2899         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2900         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2901         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2902         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2903           and BUILD_EXEEXT
2904         Note: automatic setting of the BUILD variables is not yet working
2905         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2906           compile-time but not at link-time
2907         [LINK]: use for linking executables only
2908         make different versions for Windows and non-Windows
2909         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2910           libraries
2911         [LINK_FOR_BUILD]: new variable
2912         [OBJEXT]: use throughout
2913         [EXEEXT]: use throughout
2914         <winshared>: new target
2915         <wininstall>: new target
2916         <dftables.o>: use native compiler
2917         <dftables>: use native linker
2918         <install>: handle Windows platform correctly
2919         <clean>: ditto
2920         <check>: ditto
2921         copy DLL to top builddir before testing
2923       As part of these changes, -no-undefined was removed again. This was reported
2924       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2925       in any case.
2927    3. Some tidies to get rid of compiler warnings:
2929       . In the match_data structure, match_limit was an unsigned long int, whereas
2930         match_call_count was an int. I've made them both unsigned long ints.
2932       . In pcretest the fact that a const uschar * doesn't automatically cast to
2933         a void * provoked a warning.
2935       . Turning on some more compiler warnings threw up some "shadow" variables
2936         and a few more missing casts.
2938    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2939       option, a class that contained a single character with a value between 128
2940       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2942    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2943       option, a class that contained several characters, but with at least one
2944       whose value was between 128 and 255 caused PCRE to crash.
2947    Version 4.1 12-Mar-03
2948    ---------------------
2950    1. Compiling with gcc -pedantic found a couple of places where casts were
2951    needed, and a string in dftables.c that was longer than standard compilers are
2952    required to support.
2954    2. Compiling with Sun's compiler found a few more places where the code could
2955    be tidied up in order to avoid warnings.
2957    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2958    first of these names is deprecated in the latest Autoconf in favour of the name
2959    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2960    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2961    analogy I have changed it to CFLAGS_FOR_BUILD.
2963    4. Added -no-undefined to the linking command in the Makefile, because this is
2964    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2965    linking step for the pcreposix library.
2967    5. PCRE was failing to diagnose the case of two named groups with the same
2968    name.
2970    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2971    literal character that is needed in the subject for a match, and scans along to
2972    ensure that it is present before embarking on the full matching process. This
2973    saves time in cases of nested unlimited repeats that are never going to match.
2974    Problem: the scan can take a lot of time if the subject is very long (e.g.
2975    megabytes), thus penalizing straightforward matches. It is now done only if the
2976    amount of subject to be scanned is less than 1000 bytes.
2978    7. A lesser problem with the same optimization is that it was recording the
2979    first character of an anchored pattern as "needed", thus provoking a search
2980    right along the subject, even when the first match of the pattern was going to
2981    fail. The "needed" character is now not set for anchored patterns, unless it
2982    follows something in the pattern that is of non-fixed length. Thus, it still
2983    fulfils its original purpose of finding quick non-matches in cases of nested
2984    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2987    Version 4.0 17-Feb-03
2988    ---------------------
2990    1. If a comment in an extended regex that started immediately after a meta-item
2991    extended to the end of string, PCRE compiled incorrect data. This could lead to
2992    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2994    2. Moved to autoconf 2.53 and libtool 1.4.2.
2996    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2997    the special perltest8 script is no longer needed - all the tests can be run
2998    from a single perltest script.
3000    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3001    by \s. It has now been removed in PCRE. This means it isn't recognized as
3002    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3003    class [:space:] *does* include VT, thereby creating a mess.
3005    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3006    space and tab.
3008    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3009    its new features into the main test script, reducing the number of scripts.
3011    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3012    were backward compatible, and made the (?i) apply to the whole pattern, as if
3013    /i were given. Now it behaves more logically, and applies the option setting
3014    only to what follows. PCRE has been changed to follow suit. However, if it
3015    finds options settings right at the start of the pattern, it extracts them into
3016    the global options, as before. Thus, they show up in the info data.
3018    8. Added support for the \Q...\E escape sequence. Characters in between are
3019    treated as literals. This is slightly different from Perl in that $ and @ are
3020    also handled as literals inside the quotes. In Perl, they will cause variable
3021    interpolation. Note the following examples:
3023        Pattern            PCRE matches      Perl matches
3025        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3026        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3027        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3029    For compatibility with Perl, \Q...\E sequences are recognized inside character
3030    classes as well as outside them.
3032    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3033    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3034    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3035    signed/unsigned warnings.
3037    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3038    option for pcretest, so I've replaced it by a simple function that does just
3039    that job.
3041    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3042    "pcregrep -".
3044    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3045    Java package. This provides some syntactic sugar for simple cases of what my
3046    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3047    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3048    item, you can use this simplified notation. Note that only makes sense with
3049    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3050    greediness, whatever the setting of the PCRE_UNGREEDY option.
3052    13. A change of greediness default within a pattern was not taking effect at
3053    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3054    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3055    was abstracted outside.
3057    14. PCRE now supports the \G assertion. It is true when the current matching
3058    position is at the start point of the match. This differs from \A when the
3059    starting offset is non-zero. Used with the /g option of pcretest (or similar
3060    code), it works in the same way as it does for Perl's /g option. If all
3061    alternatives of a regex begin with \G, the expression is anchored to the start
3062    match position, and the "anchored" flag is set in the compiled expression.
3064    15. Some bugs concerning the handling of certain option changes within patterns
3065    have been fixed. These applied to options other than (?ims). For example,
3066    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3067    been the other way round. Some of this was related to change 7 above.
3069    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3070    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3071    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3072    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3074    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3075    mode. Unlike ".", it always matches newline, whatever the setting of
3076    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3077    assertions. Perl allows it, but it doesn't (in general) work because it can't
3078    calculate the length of the lookbehind. At least, that's the case for Perl
3079    5.8.0 - I've been told they are going to document that it doesn't work in
3080    future.
3082    18. Added an error diagnosis for escapes that PCRE does not support: these are
3083    \L, \l, \N, \P, \p, \U, \u, and \X.
3085    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3086    reading past the end of the pattern in cases such as /[abcd/.
3088    20. PCRE was getting more memory than necessary for patterns with classes that
3089    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3091    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3092    compiling PCRE for use with Virtual Pascal.
3094    22. Small fix to the Makefile to make it work properly if the build is done
3095    outside the source tree.
3097    23. Added a new extension: a condition to go with recursion. If a conditional
3098    subpattern starts with (?(R) the "true" branch is used if recursion has
3099    happened, whereas the "false" branch is used only at the top level.
3101    24. When there was a very long string of literal characters (over 255 bytes
3102    without UTF support, over 250 bytes with UTF support), the computation of how
3103    much memory was required could be incorrect, leading to segfaults or other
3104    strange effects.
3106    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3107    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3108    there was a subsequent back reference to those brackets. This meant that, for
3109    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3110    possible to check for precisely this case. All we can do is abandon the
3111    optimization if .* occurs inside capturing brackets when there are any back
3112    references whatsoever. (See below for a better fix that came later.)
3114    26. The handling of the optimization for finding the first character of a
3115    non-anchored pattern, and for finding a character that is required later in the
3116    match were failing in some cases. This didn't break the matching; it just
3117    failed to optimize when it could. The way this is done has been re-implemented.
3119    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3121    28. Added a new feature that provides some of the functionality that Perl
3122    provides with (?{...}). The facility is termed a "callout". The way it is done
3123    in PCRE is for the caller to provide an optional function, by setting
3124    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3125    global variable. By default it is unset, which disables all calling out. To get
3126    the function called, the regex must include (?C) at appropriate points. This
3127    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3128    This provides a means of identifying different callout points. When PCRE
3129    reaches such a point in the regex, if pcre_callout has been set, the external
3130    function is called. It is provided with data in a structure called
3131    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3132    matching continues; if it returns a non-zero value, the match at the current
3133    point fails. However, backtracking will occur if possible. [This was changed
3134    later and other features added - see item 49 below.]
3136    29. pcretest is upgraded to test the callout functionality. It provides a
3137    callout function that displays information. By default, it shows the start of
3138    the match and the current position in the text. There are some new data escapes
3139    to vary what happens:
3141        \C+         in addition, show current contents of captured substrings
3142        \C-         do not supply a callout function
3143        \C!n        return 1 when callout number n is reached
3144        \C!n!m      return 1 when callout number n is reached for the mth time
3146    30. If pcregrep was called with the -l option and just a single file name, it
3147    output "<stdin>" if a match was found, instead of the file name.
3149    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3150    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3151    pcre_exec(). This saves a malloc/free per call. The default value of
3152    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3153    when configuring.
3155    32. The default maximum size of a compiled pattern is 64K. There have been a
3156    few cases of people hitting this limit. The code now uses macros to handle the
3157    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3158    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3159    configuring. Tests 2 and 5 work only with 2-byte links because they output
3160    debugging information about compiled patterns.
3162    33. Internal code re-arrangements:
3164    (a) Moved the debugging function for printing out a compiled regex into
3165        its own source file (printint.c) and used #include to pull it into
3166        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3167        separate copies.
3169    (b) Defined the list of op-code names for debugging as a macro in
3170        internal.h so that it is next to the definition of the opcodes.
3172    (c) Defined a table of op-code lengths for simpler skipping along compiled
3173        code. This is again a macro in internal.h so that it is next to the
3174        definition of the opcodes.
3176    34. Added support for recursive calls to individual subpatterns, along the
3177    lines of Robin Houston's patch (but implemented somewhat differently).
3179    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3180    allow it to read and process whole directories in Win32. This code was
3181    contributed by Lionel Fourquaux; it has not been tested by me.
3183    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3184    used to name a group. Names consist of alphanumerics and underscores, and must
3185    be unique. Back references use the syntax (?P=name) and recursive calls use
3186    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3187    numbers. The function pcre_fullinfo() can be used after compilation to extract
3188    a name/number map. There are three relevant calls:
3190      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3191      PCRE_INFO_NAMECOUNT            yields the number of entries
3192      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3194    The map is a vector of fixed-size entries. The size of each entry depends on
3195    the length of the longest name used. The first two bytes of each entry are the
3196    group number, most significant byte first. There follows the corresponding
3197    name, zero terminated. The names are in alphabetical order.
3199    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3200    case instead of 255. Making it the same both with and without UTF-8 support
3201    means that the same test output works with both.
3203    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3204    calling malloc() with a zero argument.
3206    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3207    optimization. I've improved things by keeping a bitmap of backreferences with
3208    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3209    fact referenced, the optimization can be applied. It is unlikely that a
3210    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3211    the match to follow \n) will appear inside brackets with a number greater than
3212    31, but if it does, any back reference > 31 suppresses the optimization.
3214    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3215    of disabling numbered capturing parentheses. Any opening parenthesis that is
3216    not followed by ? behaves as if it were followed by ?: but named parentheses
3217    can still be used for capturing (and they will acquire numbers in the usual
3218    way).
3220    41. Redesigned the return codes from the match() function into yes/no/error so
3221    that errors can be passed back from deep inside the nested calls. A malloc
3222    failure while inside a recursive subpattern call now causes the
3223    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3225    42. It is now possible to set a limit on the number of times the match()
3226    function is called in a call to pcre_exec(). This facility makes it possible to
3227    limit the amount of recursion and backtracking, though not in a directly
3228    obvious way, because the match() function is used in a number of different
3229    circumstances. The count starts from zero for each position in the subject
3230    string (for non-anchored patterns). The default limit is, for compatibility, a
3231    large number, namely 10 000 000. You can change this in two ways:
3233    (a) When configuring PCRE before making, you can use --with-match-limit=n
3234        to set a default value for the compiled library.
3236    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3237        a different value is set. See 45 below.
3239    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3241    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3242    of things that can be changed at compile time. The first argument specifies
3243    what is wanted and the second points to where the information is to be placed.
3244    The current list of available information is:
3246      PCRE_CONFIG_UTF8
3248    The output is an integer that is set to one if UTF-8 support is available;
3249    otherwise it is set to zero.
3253    The output is an integer that it set to the value of the code that is used for
3254    newline. It is either LF (10) or CR (13).
3258    The output is an integer that contains the number of bytes used for internal
3259    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3263    The output is an integer that contains the threshold above which the POSIX
3264    interface uses malloc() for output vectors. See item 31 above.
3268    The output is an unsigned integer that contains the default limit of the number
3269    of match() calls in a pcre_exec() execution. See 42 above.
3271    44. pcretest has been upgraded by the addition of the -C option. This causes it
3272    to extract all the available output from the new pcre_config() function, and to
3273    output it. The program then exits immediately.
3275    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3276    order to support additional features. One way would have been to define
3277    pcre_exec2() (for example) with extra arguments, but this would not have been
3278    extensible, and would also have required all calls to the original function to
3279    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3280    is used for passing in "extra" data from pcre_study().
3282    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3283    contains the following fields:
3285      flags         a bitmap indicating which of the following fields are set
3286      study_data    opaque data from pcre_study()
3287      match_limit   a way of specifying a limit on match() calls for a specific
3288                      call to pcre_exec()
3289      callout_data  data for callouts (see 49 below)
3291    The flag bits are also defined in pcre.h, and are
3297    The pcre_study() function now returns one of these new pcre_extra blocks, with
3298    the actual study data pointed to by the study_data field, and the
3299    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3300    before. That is, this change is entirely upwards-compatible and requires no
3301    change to existing code.
3303    If you want to pass in additional data to pcre_exec(), you can either place it
3304    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3305    block.
3307    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3308    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3309    times with different match limits, until it finds the minimum value needed for
3310    pcre_exec() to complete. The value is then output. This can be instructive; for
3311    most simple matches the number is quite small, but for pathological cases it
3312    gets very large very quickly.
3314    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3315    returns the size of the data block pointed to by the study_data field in a
3316    pcre_extra block, that is, the value that was passed as the argument to
3317    pcre_malloc() when PCRE was getting memory in which to place the information
3318    created by pcre_study(). The fourth argument should point to a size_t variable.
3319    pcretest has been extended so that this information is shown after a successful
3320    pcre_study() call when information about the compiled regex is being displayed.
3322    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3323    because what follows is always an absolute path. (Later: it turns out that this
3324    is more than cosmetic for MinGW, because it doesn't like empty path
3325    components.)
3327    49. Some changes have been made to the callout feature (see 28 above):
3329    (i)  A callout function now has three choices for what it returns:
3331           0  =>  success, carry on matching
3332         > 0  =>  failure at this point, but backtrack if possible
3333         < 0  =>  serious error, return this value from pcre_exec()
3335         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3336         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3337         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3338         use by callout functions. It will never be used by PCRE itself.
3340    (ii) The pcre_extra structure (see 45 above) has a void * field called
3341         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3342         pcre_callout_block structure has a field of the same name. The contents of
3343         the field passed in the pcre_extra structure are passed to the callout
3344         function in the corresponding field in the callout block. This makes it
3345         easier to use the same callout-containing regex from multiple threads. For
3346         testing, the pcretest program has a new data escape
3348           \C*n        pass the number n (may be negative) as callout_data
3350         If the callout function in pcretest receives a non-zero value as
3351         callout_data, it returns that value.
3353    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3354    there were some redundant $(CFLAGS) in commands that are now specified as
3355    $(LINK), which already includes $(CFLAGS).
3357    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3358    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3359    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3360    one-byte characters throughout. Note that case-insensitive matching applies
3361    only to characters whose values are less than 256. PCRE doesn't support the
3362    notion of cases for higher-valued characters.
3364    (i)   A character class whose characters are all within 0-255 is handled as
3365          a bit map, and the map is inverted for negative classes. Previously, a
3366          character > 255 always failed to match such a class; however it should
3367          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3369    (ii)  A negated character class with a single character < 255 is coded as
3370          "not this character" (OP_NOT). This wasn't working properly when the test
3371          character was multibyte, either singly or repeated.
3373    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3374          mode, for example: \x{100}{2,3}.
3376    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3377          singly or repeated) now correctly test multibyte characters. However,
3378          PCRE doesn't recognize any characters with values greater than 255 as
3379          digits, spaces, or word characters. Such characters always match \D, \S,
3380          and \W, and never match \d, \s, or \w.
3382    (v)   Classes may now contain characters and character ranges with values
3383          greater than 255. For example: [ab\x{100}-\x{400}].
3385    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3386          PCRE in UTF-8 mode.
3388    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3389    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3390    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3391    value.)
3393    53. The single man page has become too large. I have therefore split it up into
3394    a number of separate man pages. These also give rise to individual HTML pages;
3395    these are now put in a separate directory, and there is an index.html page that
3396    lists them all. Some hyperlinking between the pages has been installed.
3398    54. Added convenience functions for handling named capturing parentheses.
3400    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3401    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3402    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3403    are faulted.
3405    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3406    calling configure. These values are used when compiling the dftables.c program
3407    which is run to generate the source of the default character tables. They
3408    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3409    you will need to set these values.
3411    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3414    Version 3.9 02-Jan-02
3415    ---------------------
3417    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3419    2. If --disable-static was given, the building process failed when trying to
3420    build pcretest and pcregrep. (For some reason it was using libtool to compile
3421    them, which is not right, as they aren't part of the library.)
3424    Version 3.8 18-Dec-01
3425    ---------------------
3427    1. The experimental UTF-8 code was completely screwed up. It was packing the
3428    bytes in the wrong order. How dumb can you get?
3431    Version 3.7 29-Oct-01
3432    ---------------------
3434    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3435    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3436    this didn't happen under Solaris 8, where I normally test things.
3438    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3439    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3440    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3441    this fix an hour or so after the initial 3.7 release.)
3444    Version 3.6 23-Oct-01
3445    ---------------------
3447    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3448    offsets passed as NULL with zero offset count.
3450    2. The config.guess and config.sub files had not been updated when I moved to
3451    the latest autoconf.
3454    Version 3.5 15-Aug-01
3455    ---------------------
3457    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3458    had been forgotten.
3460    2. By using declared but undefined structures, we can avoid using "void"
3461    definitions in pcre.h while keeping the internal definitions of the structures
3462    private.
3464    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3465    user point of view, this means that both static and shared libraries are built
3466    by default, but this can be individually controlled. More of the work of
3467    handling this static/shared cases is now inside libtool instead of PCRE's make
3468    file.
3470    4. The pcretest utility is now installed along with pcregrep because it is
3471    useful for users (to test regexs) and by doing this, it automatically gets
3472    relinked by libtool. The documentation has been turned into a man page, so
3473    there are now .1, .txt, and .html versions in /doc.
3475    5. Upgrades to pcregrep:
3476       (i)   Added long-form option names like gnu grep.
3477       (ii)  Added --help to list all options with an explanatory phrase.
3478       (iii) Added -r, --recursive to recurse into sub-directories.
3479       (iv)  Added -f, --file to read patterns from a file.
3481    6. pcre_exec() was referring to its "code" argument before testing that
3482    argument for NULL (and giving an error if it was NULL).
3484    7. Upgraded Makefile.in to allow for compiling in a different directory from
3485    the source directory.
3487    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3488    options bits, the pointer it was passed was to an int instead of to an unsigned
3489    long int. This mattered only on 64-bit systems.
3491    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3492    generated) instead of pcre.in, which it its source. Also made the same change
3493    in several of the .c files.
3495    10. A new release of gcc defines printf() as a macro, which broke pcretest
3496    because it had an ifdef in the middle of a string argument for printf(). Fixed
3497    by using separate calls to printf().
3499    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3500    script, to force use of CR or LF instead of \n in the source. On non-Unix
3501    systems, the value can be set in config.h.
3503    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3504    absolute limit. Changed the text of the error message to make this clear, and
3505    likewise updated the man page.
3507    13. The limit of 99 on the number of capturing subpatterns has been removed.
3508    The new limit is 65535, which I hope will not be a "real" limit.
3511    Version 3.4 22-Aug-00
3512    ---------------------
3514    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3516    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3519    Version 3.3 01-Aug-00
3520    ---------------------
3522    1. If an octal character was given, but the value was greater than \377, it
3523    was not getting masked to the least significant bits, as documented. This could
3524    lead to crashes in some systems.
3526    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3527    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3529    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3530    These just pass their arguments on to (pcre_free)(), but they are provided
3531    because some uses of PCRE bind it to non-C systems that can call its functions,
3532    but cannot call free() or pcre_free() directly.
3534    4. Add "make test" as a synonym for "make check". Corrected some comments in
3535    the Makefile.
3537    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3538    Makefile.
3540    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3541    command called pgrep for grepping around the active processes.
3543    7. Added the beginnings of support for UTF-8 character strings.
3545    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3546    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3547    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3548    out for the ar command.)
3551    Version 3.2 12-May-00
3552    ---------------------
3554    This is purely a bug fixing release.
3556    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3557    of ZA. This was just one example of several cases that could provoke this bug,
3558    which was introduced by change 9 of version 2.00. The code for breaking
3559    infinite loops after an iteration that matches an empty string was't working
3560    correctly.
3562    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3563    when matched against abbab (for example). After matching an empty string, it
3564    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3565    caused it to match further down the string than it should.
3567    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3568    was there because it doesn't seem to be needed, and it causes trouble on some
3569    systems, as it is not a Standard C header. It has been removed.
3571    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3572    were reported on the Macintosh. The changes were from
3574      while ((c = *(++ptr)) != 0 && c != '\n');
3575    to
3576      while ((c = *(++ptr)) != 0 && c != '\n') ;
3578    Totally extraordinary, but if that's what it takes...
3580    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3581    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3582    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3583    assumes the way PCRE uses memmove() (always moving upwards).
3585    6. PCRE is being used in one environment where strchr() is not available. There
3586    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3587    faster code anyway.
3590    Version 3.1 09-Feb-00
3591    ---------------------
3593    The only change in this release is the fixing of some bugs in Makefile.in for
3594    the "install" target:
3596    (1) It was failing to install pcreposix.h.
3598    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3601    Version 3.0 01-Feb-00
3602    ---------------------