/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 27 by nigel, Sat Feb 24 21:38:49 2007 UTC revision 529 by ph10, Mon May 31 17:28:08 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.10 03 May-2010
5    ------------------------
6    
7    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8        THEN.
9    
10    2.  (*ACCEPT) was not working when inside an atomic group.
11    
12    3.  Inside a character class, \B is treated as a literal by default, but
13        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14        causes the error). The code is unchanged, but I tidied the documentation.
15    
16    4.  Inside a character class, PCRE always treated \R and \X as literals,
17        whereas Perl faults them if its -w option is set. I have changed PCRE so
18        that it faults them when PCRE_EXTRA is set.
19    
20    5.  Added support for \N, which always matches any character other than
21        newline. (It is the same as "." when PCRE_DOTALL is not set.)
22    
23    6.  When compiling pcregrep with newer versions of gcc which may have
24        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25        declared with attribute warn_unused_result" were given. Just casting the
26        result to (void) does not stop the warnings; a more elaborate fudge is
27        needed. I've used a macro to implement this.
28    
29    7.  Minor change to pcretest.c to avoid a compiler warning.
30    
31    8.  Added four artifical Unicode properties to help with an option to make
32        \s etc use properties (see next item). The new properties are: Xan
33        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34    
35    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36        use Unicode properties. (*UCP) at the start of a pattern can be used to set
37        this option. Modified pcretest to add /W to test this facility. Added
38        REG_UCP to make it available via the POSIX interface.
39    
40    10. Added --line-buffered to pcregrep.
41    
42    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43        studied, and the match started with a letter with a code point greater than
44        127 whose first byte was different to the first byte of the other case of
45        the letter, the other case of this starting letter was not recognized.
46    
47    12. pcreposix.c included pcre.h before including pcre_internal.h. This caused a
48        conflict in the definition of PCRE_EXP_DECL. I have removed the include of
49        pcre.h as pcre_internal.h includes pcre.h itself. (This may be a bit of
50        historical tidying that never got done.)
51    
52    13. If a pattern that was studied started with a repeated Unicode property
53        test, for example, \p{Nd}+, there was the theoretical possibility of
54        setting up an incorrect bitmap of starting bytes, but fortunately it could
55        not have actually happened in practice until change 8 above was made (it
56        added property types that matched character-matching opcodes).
57    
58    14. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
59        possible starting bytes for non-anchored patterns.
60    
61    15. The "auto-possessify" feature of pcre_compile() now recognizes \R.
62    
63    16. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
64        input, it could crash or give wrong results if characters with values
65        greater than 0xc0 were present in the subject string. (Detail: it assumed
66        UTF-8 input when processing these items.)
67    
68    
69    
70    Version 8.02 19-Mar-2010
71    ------------------------
72    
73    1.  The Unicode data tables have been updated to Unicode 5.2.0.
74    
75    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
76        configured.
77    
78    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
79        original author of that file, following a query about its status.
80    
81    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
82        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
83    
84    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
85        quantifier applied to a forward-referencing subroutine call, could compile
86        incorrect code or give the error "internal error: previously-checked
87        referenced subpattern not found".
88    
89    6.  Both MS Visual Studio and Symbian OS have problems with initializing
90        variables to point to external functions. For these systems, therefore,
91        pcre_malloc etc. are now initialized to local functions that call the
92        relevant global functions.
93    
94    7.  There were two entries missing in the vectors called coptable and poptable
95        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
96        I've fixed the data, and added a kludgy way of testing at compile time that
97        the lengths are correct (equal to the number of opcodes).
98    
99    8.  Following on from 7, I added a similar kludge to check the length of the
100        eint vector in pcreposix.c.
101    
102    9.  Error texts for pcre_compile() are held as one long string to avoid too
103        much relocation at load time. To find a text, the string is searched,
104        counting zeros. There was no check for running off the end of the string,
105        which could happen if a new error number was added without updating the
106        string.
107    
108    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
109    
110    11. \K was not working if it appeared in an atomic group or in a group that
111        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
112        \K is "not well defined" if used in an assertion. PCRE now accepts it if
113        the assertion is positive, but not if it is negative.
114    
115    12. Change 11 fortuitously reduced the size of the stack frame used in the
116        "match()" function of pcre_exec.c by one pointer. Forthcoming
117        implementation of support for (*MARK) will need an extra pointer on the
118        stack; I have reserved it now, so that the stack frame size does not
119        decrease.
120    
121    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
122        item in branch that calls a recursion is a subroutine call - as in the
123        second branch in the above example - was incorrectly given the compile-
124        time error "recursive call could loop indefinitely" because pcre_compile()
125        was not correctly checking the subroutine for matching a non-empty string.
126    
127    14. The checks for overrunning compiling workspace could trigger after an
128        overrun had occurred. This is a "should never occur" error, but it can be
129        triggered by pathological patterns such as hundreds of nested parentheses.
130        The checks now trigger 100 bytes before the end of the workspace.
131    
132    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
133    
134    
135    Version 8.01 19-Jan-2010
136    ------------------------
137    
138    1.  If a pattern contained a conditional subpattern with only one branch (in
139        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
140        computed the wrong minimum data length (which is of course zero for such
141        subpatterns). This could cause incorrect "no match" results.
142    
143    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
144        the pattern is reset in the first branch, pcre_compile() failed with
145        "internal error: code overflow at offset...". This happened only when
146        the reset was to the original external option setting. (An optimization
147        abstracts leading options settings into an external setting, which was the
148        cause of this.)
149    
150    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
151        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
152        assertion pattern did not match (meaning that the assertion was true), it
153        was incorrectly treated as false if the SKIP had been reached during the
154        matching. This also applied to assertions used as conditions.
155    
156    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
157        assertion subpattern, including such a pattern used as a condition,
158        unpredictable results occurred, instead of the error return
159        PCRE_ERROR_DFA_UITEM.
160    
161    5.  The C++ GlobalReplace function was not working like Perl for the special
162        situation when an empty string is matched. It now does the fancy magic
163        stuff that is necessary.
164    
165    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
166        removed. (These were left over from very, very early versions of PCRE.)
167    
168    7.  Some cosmetic changes to the code to make life easier when compiling it
169        as part of something else:
170    
171        (a) Change DEBUG to PCRE_DEBUG.
172    
173        (b) In pcre_compile(), rename the member of the "branch_chain" structure
174            called "current" as "current_branch", to prevent a collision with the
175            Linux macro when compiled as a kernel module.
176    
177        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
178            prevent a collision with the Linux macro when compiled as a kernel
179            module.
180    
181    8.  In pcre_compile() there are some checks for integer overflows that used to
182        cast potentially large values to (double). This has been changed to that
183        when building, a check for int64_t is made, and if it is found, it is used
184        instead, thus avoiding the use of floating point arithmetic. (There is no
185        other use of FP in PCRE.) If int64_t is not found, the fallback is to
186        double.
187    
188    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
189        2005 (difference between two addresses compared to an unsigned value).
190    
191    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
192        custom one, because of the following reported problem in Windows:
193    
194          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
195              under Win32.
196          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
197              therefore missing the function definition.
198          - The compiler thus generates a "C" signature for the test function.
199          - The linker fails to find the "C" function.
200          - PCRE fails to configure if asked to do so against libbz2.
201    
202    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
203        messages were output:
204    
205          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
206          rerunning libtoolize, to keep the correct libtool macros in-tree.
207          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
208    
209        I have done both of these things.
210    
211    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
212        most of the time, it *can* run out if it is given a pattern that contains a
213        runaway infinite recursion. I updated the discussion in the pcrestack man
214        page.
215    
216    13. Now that we have gone to the x.xx style of version numbers, the minor
217        version may start with zero. Using 08 or 09 is a bad idea because users
218        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
219        interpreted as invalid octal numbers. I've updated the previous comment in
220        configure.ac, and also added a check that gives an error if 08 or 09 are
221        used.
222    
223    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
224        causing partial matching to fail when the end of the subject matched \W
225        in a UTF-8 pattern where \W was quantified with a minimum of 3.
226    
227    15. There were some discrepancies between the declarations in pcre_internal.h
228        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
229        their definitions. The declarations used "const uschar *" and the
230        definitions used USPTR. Even though USPTR is normally defined as "const
231        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
232        reported that: "This difference in casting confuses some C++ compilers, for
233        example, SunCC recognizes above declarations as different functions and
234        generates broken code for hbpcre." I have changed the declarations to use
235        USPTR.
236    
237    16. GNU libtool is named differently on some systems. The autogen.sh script now
238        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
239        (FreeBSD).
240    
241    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
242        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
243        comment: "Figure out how to create a longlong from a string: strtoll and
244        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
245        instance, but it only takes 2 args instead of 3!"
246    
247    18. A subtle bug concerned with back references has been fixed by a change of
248        specification, with a corresponding code fix. A pattern such as
249        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
250        refers, was giving matches when it shouldn't. For example, xa=xaaa would
251        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
252        same bug. Such groups have to be quantified to be useful, or contained
253        inside another quantified group. (If there's no repetition, the reference
254        can never match.) The problem arises because, having left the group and
255        moved on to the rest of the pattern, a later failure that backtracks into
256        the group uses the captured value from the final iteration of the group
257        rather than the correct earlier one. I have fixed this in PCRE by forcing
258        any group that contains a reference to itself to be an atomic group; that
259        is, there cannot be any backtracking into it once it has completed. This is
260        similar to recursive and subroutine calls.
261    
262    
263    Version 8.00 19-Oct-09
264    ----------------------
265    
266    1.  The table for translating pcre_compile() error codes into POSIX error codes
267        was out-of-date, and there was no check on the pcre_compile() error code
268        being within the table. This could lead to an OK return being given in
269        error.
270    
271    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
272        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
273        in a Windows environment.
274    
275    3.  The pcregrep --count option prints the count for each file even when it is
276        zero, as does GNU grep. However, pcregrep was also printing all files when
277        --files-with-matches was added. Now, when both options are given, it prints
278        counts only for those files that have at least one match. (GNU grep just
279        prints the file name in this circumstance, but including the count seems
280        more useful - otherwise, why use --count?) Also ensured that the
281        combination -clh just lists non-zero counts, with no names.
282    
283    4.  The long form of the pcregrep -F option was incorrectly implemented as
284        --fixed_strings instead of --fixed-strings. This is an incompatible change,
285        but it seems right to fix it, and I didn't think it was worth preserving
286        the old behaviour.
287    
288    5.  The command line items --regex=pattern and --regexp=pattern were not
289        recognized by pcregrep, which required --regex pattern or --regexp pattern
290        (with a space rather than an '='). The man page documented the '=' forms,
291        which are compatible with GNU grep; these now work.
292    
293    6.  No libpcreposix.pc file was created for pkg-config; there was just
294        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
295    
296    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
297        when UCP support is not needed, by modifying the Python script that
298        generates it from Unicode data files. This should not matter if the module
299        is correctly used as a library, but I received one complaint about 50K of
300        unwanted data. My guess is that the person linked everything into his
301        program rather than using a library. Anyway, it does no harm.
302    
303    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
304        was a minimum greater than 1 for a wide character in a possessive
305        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
306        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
307        character. Chaos in the form of incorrect output or a compiling loop could
308        result.
309    
310    9.  The restrictions on what a pattern can contain when partial matching is
311        requested for pcre_exec() have been removed. All patterns can now be
312        partially matched by this function. In addition, if there are at least two
313        slots in the offset vector, the offset of the earliest inspected character
314        for the match and the offset of the end of the subject are set in them when
315        PCRE_ERROR_PARTIAL is returned.
316    
317    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
318        synonymous with PCRE_PARTIAL, for backwards compatibility, and
319        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
320        and may be more useful for multi-segment matching.
321    
322    11. Partial matching with pcre_exec() is now more intuitive. A partial match
323        used to be given if ever the end of the subject was reached; now it is
324        given only if matching could not proceed because another character was
325        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
326        string "Z", which now yields "no match" instead of "partial match". In the
327        case of pcre_dfa_exec(), "no match" is given if every matching path for the
328        final character ended with (*FAIL).
329    
330    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
331        if the pattern had a "must contain" character that was already found in the
332        earlier partial match, unless partial matching was again requested. For
333        example, with the pattern /dog.(body)?/, the "must contain" character is
334        "g". If the first part-match was for the string "dog", restarting with
335        "sbody" failed. This bug has been fixed.
336    
337    13. The string returned by pcre_dfa_exec() after a partial match has been
338        changed so that it starts at the first inspected character rather than the
339        first character of the match. This makes a difference only if the pattern
340        starts with a lookbehind assertion or \b or \B (\K is not supported by
341        pcre_dfa_exec()). It's an incompatible change, but it makes the two
342        matching functions compatible, and I think it's the right thing to do.
343    
344    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
345        so that the demonstration program is easily available in environments where
346        PCRE has not been installed from source.
347    
348    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
349        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
350        library.
351    
352    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
353        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
354        is not the first non-POSIX option to be added. Clearly some people find
355        these options useful.
356    
357    17. If a caller to the POSIX matching function regexec() passes a non-zero
358        value for nmatch with a NULL value for pmatch, the value of
359        nmatch is forced to zero.
360    
361    18. RunGrepTest did not have a test for the availability of the -u option of
362        the diff command, as RunTest does. It now checks in the same way as
363        RunTest, and also checks for the -b option.
364    
365    19. If an odd number of negated classes containing just a single character
366        interposed, within parentheses, between a forward reference to a named
367        subpattern and the definition of the subpattern, compilation crashed with
368        an internal error, complaining that it could not find the referenced
369        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
370        [The bug was that it was starting one character too far in when skipping
371        over the character class, thus treating the ] as data rather than
372        terminating the class. This meant it could skip too much.]
373    
374    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
375        /g option in pcretest when the pattern contains \K, which makes it possible
376        to have an empty string match not at the start, even when the pattern is
377        anchored. Updated pcretest and pcredemo to use this option.
378    
379    21. If the maximum number of capturing subpatterns in a recursion was greater
380        than the maximum at the outer level, the higher number was returned, but
381        with unset values at the outer level. The correct (outer level) value is
382        now given.
383    
384    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
385        PCRE did not set those parentheses (unlike Perl). I have now found a way to
386        make it do so. The string so far is captured, making this feature
387        compatible with Perl.
388    
389    23. The tests have been re-organized, adding tests 11 and 12, to make it
390        possible to check the Perl 5.10 features against Perl 5.10.
391    
392    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
393        pattern matches a fixed length string. PCRE did not allow this; now it
394        does. Neither allows recursion.
395    
396    25. I finally figured out how to implement a request to provide the minimum
397        length of subject string that was needed in order to match a given pattern.
398        (It was back references and recursion that I had previously got hung up
399        on.) This code has now been added to pcre_study(); it finds a lower bound
400        to the length of subject needed. It is not necessarily the greatest lower
401        bound, but using it to avoid searching strings that are too short does give
402        some useful speed-ups. The value is available to calling programs via
403        pcre_fullinfo().
404    
405    26. While implementing 25, I discovered to my embarrassment that pcretest had
406        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
407        study optimizations had never been tested with that matching function.
408        Oops. What is worse, even when it was passed study data, there was a bug in
409        pcre_dfa_exec() that meant it never actually used it. Double oops. There
410        were also very few tests of studied patterns with pcre_dfa_exec().
411    
412    27. If (?| is used to create subpatterns with duplicate numbers, they are now
413        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
414        on the other side of the coin, they are no longer allowed to have different
415        names, because these cannot be distinguished in PCRE, and this has caused
416        confusion. (This is a difference from Perl.)
417    
418    28. When duplicate subpattern names are present (necessarily with different
419        numbers, as required by 27 above), and a test is made by name in a
420        conditional pattern, either for a subpattern having been matched, or for
421        recursion in such a pattern, all the associated numbered subpatterns are
422        tested, and the overall condition is true if the condition is true for any
423        one of them. This is the way Perl works, and is also more like the way
424        testing by number works.
425    
426    
427    Version 7.9 11-Apr-09
428    ---------------------
429    
430    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
431        (pcretest), all targets were linked against these libraries. This included
432        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
433        libraries. This caused unwanted dependencies to be created. This problem
434        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
435        pcretest is linked with readline.
436    
437    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
438        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
439        moved outside it again, because FALSE and TRUE are already defined in AIX,
440        but BOOL is not.
441    
442    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
443        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
444    
445    4.  The pcregrep documentation said spaces were inserted as well as colons (or
446        hyphens) following file names and line numbers when outputting matching
447        lines. This is not true; no spaces are inserted. I have also clarified the
448        wording for the --colour (or --color) option.
449    
450    5.  In pcregrep, when --colour was used with -o, the list of matching strings
451        was not coloured; this is different to GNU grep, so I have changed it to be
452        the same.
453    
454    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
455        each matching line was coloured. Now it goes on to look for further matches
456        of any of the test patterns, which is the same behaviour as GNU grep.
457    
458    7.  A pattern that could match an empty string could cause pcregrep to loop; it
459        doesn't make sense to accept an empty string match in pcregrep, so I have
460        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
461        seems to be how GNU grep behaves.
462    
463    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
464        start or after a newline", because the conditional assertion was not being
465        correctly handled. The rule now is that both the assertion and what follows
466        in the first alternative must satisfy the test.
467    
468    9.  If auto-callout was enabled in a pattern with a conditional group whose
469        condition was an assertion, PCRE could crash during matching, both with
470        pcre_exec() and pcre_dfa_exec().
471    
472    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
473        used for matching.
474    
475    11. Unicode property support in character classes was not working for
476        characters (bytes) greater than 127 when not in UTF-8 mode.
477    
478    12. Added the -M command line option to pcretest.
479    
480    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
481    
482    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
483    
484    16. Added comments and documentation about mis-use of no_arg in the C++
485        wrapper.
486    
487    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
488        from Martin Jerabek that uses macro names for all relevant character and
489        string constants.
490    
491    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
492        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
493        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
494        these, but not everybody uses configure.
495    
496    19. A conditional group that had only one branch was not being correctly
497        recognized as an item that could match an empty string. This meant that an
498        enclosing group might also not be so recognized, causing infinite looping
499        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
500        with the subject "ab", where knowledge that the repeated group can match
501        nothing is needed in order to break the loop.
502    
503    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
504        exec(), but without supplying a callout function, matching went wrong.
505    
506    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
507        leak if the size of the offset vector was greater than 30. When the vector
508        is smaller, the saved offsets during recursion go onto a local stack
509        vector, but for larger vectors malloc() is used. It was failing to free
510        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
511        error, in fact).
512    
513    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
514        heapframe that is used only when UTF-8 support is enabled. This caused no
515        problem, but was untidy.
516    
517    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
518        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
519        included within another project.
520    
521    24. Steven Van Ingelgem's patches to add more options to the CMake support,
522        slightly modified by me:
523    
524          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
525              not building pcregrep.
526    
527          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
528              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
529    
530    25. Forward references, both numeric and by name, in patterns that made use of
531        duplicate group numbers, could behave incorrectly or give incorrect errors,
532        because when scanning forward to find the reference group, PCRE was not
533        taking into account the duplicate group numbers. A pattern such as
534        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
535    
536    26. Changed a few more instances of "const unsigned char *" to USPTR, making
537        the feature of a custom pointer more persuasive (as requested by a user).
538    
539    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
540        pcretest.c, inside #ifndefs, because it seems they are sometimes already
541        pre-defined.
542    
543    28. Added support for (*UTF8) at the start of a pattern.
544    
545    29. Arrange for flags added by the "release type" setting in CMake to be shown
546        in the configuration summary.
547    
548    
549    Version 7.8 05-Sep-08
550    ---------------------
551    
552    1.  Replaced UCP searching code with optimized version as implemented for Ad
553        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
554        stage table and inline lookup instead of a function, giving speed ups of 2
555        to 5 times on some simple patterns that I tested. Permission was given to
556        distribute the MultiStage2.py script that generates the tables (it's not in
557        the tarball, but is in the Subversion repository).
558    
559    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
560        scripts.
561    
562    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
563        a group with a zero qualifier. The result of the study could be incorrect,
564        or the function might crash, depending on the pattern.
565    
566    4.  Caseless matching was not working for non-ASCII characters in back
567        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
568        It now works when Unicode Property Support is available.
569    
570    5.  In pcretest, an escape such as \x{de} in the data was always generating
571        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
572        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
573        truncation.
574    
575    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
576    
577    7.  Added two (int) casts to pcregrep when printing the difference of two
578        pointers, in case they are 64-bit values.
579    
580    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
581        test 2 if it fails.
582    
583    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
584        and a #define of that name to empty if it is not externally set. This is to
585        allow users of MSVC to set it if necessary.
586    
587    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
588        the convenience functions in the pcre_get.c source file.
589    
590    11. An option change at the start of a pattern that had top-level alternatives
591        could cause overwriting and/or a crash. This command provoked a crash in
592        some environments:
593    
594          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
595    
596        This potential security problem was recorded as CVE-2008-2371.
597    
598    12. For a pattern where the match had to start at the beginning or immediately
599        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
600        pcre_dfa_exec() could read past the end of the passed subject if there was
601        no match. To help with detecting such bugs (e.g. with valgrind), I modified
602        pcretest so that it places the subject at the end of its malloc-ed buffer.
603    
604    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
605        exec() might read past the end of the data buffer in UTF-8 mode.
606    
607    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
608        the data contained the byte 0x85 as part of a UTF-8 character within its
609        first line. This applied both to normal and DFA matching.
610    
611    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
612        /^[^d]*?$/8 failed to match "abc".
613    
614    16. Added a missing copyright notice to pcrecpp_internal.h.
615    
616    17. Make it more clear in the documentation that values returned from
617        pcre_exec() in ovector are byte offsets, not character counts.
618    
619    18. Tidied a few places to stop certain compilers from issuing warnings.
620    
621    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
622        supplied by Stefan Weber. I made a further small update for 7.8 because
623        there is a change of source arrangements: the pcre_searchfuncs.c module is
624        replaced by pcre_ucd.c.
625    
626    
627    Version 7.7 07-May-08
628    ---------------------
629    
630    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
631        a string to a long long, pretend we don't even have a long long." This is
632        done by checking for the strtoq, strtoll, and _strtoi64 functions.
633    
634    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
635        pre-7.6 versions, which defined a global no_arg variable instead of putting
636        it in the RE class. (See also #8 below.)
637    
638    3.  Remove a line of dead code, identified by coverity and reported by Nuno
639        Lopes.
640    
641    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
642    
643        (1) The include/exclude patterns were being applied to the whole pathnames
644            of files, instead of just to the final components.
645    
646        (2) If there was more than one level of directory, the subdirectories were
647            skipped unless they satisfied the include/exclude conditions. This is
648            inconsistent with GNU grep (and could even be seen as contrary to the
649            pcregrep specification - which I improved to make it absolutely clear).
650            The action now is always to scan all levels of directory, and just
651            apply the include/exclude patterns to regular files.
652    
653    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
654        --exclude_dir in the tests to avoid scanning .svn directories.
655    
656    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
657        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
658        doesn't support NULs in patterns.
659    
660    7.  Added some missing "const"s to declarations of static tables in
661        pcre_compile.c and pcre_dfa_exec.c.
662    
663    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
664        caused by fix #2  above. (Subsequently also a second patch to fix the
665        first patch. And a third patch - this was a messy problem.)
666    
667    9.  Applied Craig's patch to remove the use of push_back().
668    
669    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
670        matching function regexec().
671    
672    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
673        which, however, unlike Perl's \g{...}, are subroutine calls, not back
674        references. PCRE supports relative numbers with this syntax (I don't think
675        Oniguruma does).
676    
677    12. Previously, a group with a zero repeat such as (...){0} was completely
678        omitted from the compiled regex. However, this means that if the group
679        was called as a subroutine from elsewhere in the pattern, things went wrong
680        (an internal error was given). Such groups are now left in the compiled
681        pattern, with a new opcode that causes them to be skipped at execution
682        time.
683    
684    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
685        to the way PCRE behaves:
686    
687        (a) A lone ] character is dis-allowed (Perl treats it as data).
688    
689        (b) A back reference to an unmatched subpattern matches an empty string
690            (Perl fails the current match path).
691    
692        (c) A data ] in a character class must be notated as \] because if the
693            first data character in a class is ], it defines an empty class. (In
694            Perl it is not possible to have an empty class.) The empty class []
695            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
696            The negative empty class [^] matches any one character, independently
697            of the DOTALL setting.
698    
699    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
700        non-existent subpattern following a character class starting with ']' and
701        containing () gave an internal compiling error instead of "reference to
702        non-existent subpattern". Fortunately, when the pattern did exist, the
703        compiled code was correct. (When scanning forwards to check for the
704        existencd of the subpattern, it was treating the data ']' as terminating
705        the class, so got the count wrong. When actually compiling, the reference
706        was subsequently set up correctly.)
707    
708    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
709        it was being rejected as not supported by pcre_dfa_exec(), even though
710        other assertions are supported. I have made pcre_dfa_exec() support
711        (*FAIL).
712    
713    16. The implementation of 13c above involved the invention of a new opcode,
714        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
715        cannot be changed at match time, I realized I could make a small
716        improvement to matching performance by compiling OP_ALLANY instead of
717        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
718        on the OP_ANY path.
719    
720    17. Compiling pcretest on Windows with readline support failed without the
721        following two fixes: (1) Make the unistd.h include conditional on
722        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
723    
724    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
725        ncurses library to be included for pcretest when ReadLine support is
726        requested, but also to allow for it to be overridden. This patch came from
727        Daniel Bergström.
728    
729    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
730        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
731        any errors with the current Unicode tables. Thanks to Peter Kankowski for
732        spotting this.
733    
734    
735    Version 7.6 28-Jan-08
736    ---------------------
737    
738    1.  A character class containing a very large number of characters with
739        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
740        overflow.
741    
742    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
743        HAVE_LONG_LONG is not defined.
744    
745    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
746        bring them up to date and include new features. This patch includes:
747    
748        - Fixed PH's badly added libz and libbz2 support.
749        - Fixed a problem with static linking.
750        - Added pcredemo. [But later removed - see 7 below.]
751        - Fixed dftables problem and added an option.
752        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
753            HAVE_LONG_LONG.
754        - Added readline support for pcretest.
755        - Added an listing of the option settings after cmake has run.
756    
757    4.  A user submitted a patch to Makefile that makes it easy to create
758        "pcre.dll" under mingw when using Configure/Make. I added stuff to
759        Makefile.am that cause it to include this special target, without
760        affecting anything else. Note that the same mingw target plus all
761        the other distribution libraries and programs are now supported
762        when configuring with CMake (see 6 below) instead of with
763        Configure/Make.
764    
765    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
766        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
767        exported in the Windows port". It has not yet been confirmed that the patch
768        solves the problem, but it does no harm.
769    
770    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
771        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
772        with CMake, and also correct the comment about stack recursion.
773    
774    7.  Remove the automatic building of pcredemo from the ./configure system and
775        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
776        of a program that users should build themselves after PCRE is installed, so
777        building it automatically is not really right. What is more, it gave
778        trouble in some build environments.
779    
780    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
781    
782    
783    Version 7.5 10-Jan-08
784    ---------------------
785    
786    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
787        values in parens when parsing an RE using the C++ wrapper."
788    
789    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
790        Characters greater than 255 were excluded from the class instead of being
791        included.
792    
793    3.  The same bug as (2) above applied to negated POSIX classes such as
794        [:^space:].
795    
796    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
797        defined or documented. It seems to have been a typo for PCRE_STATIC, so
798        I have changed it.
799    
800    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
801        first named subpattern) and a construct such as (?&a) would reference the
802        first named subpattern whose name started with "a" (in other words, the
803        length check was missing). Both these problems are fixed. "Subpattern name
804        expected" is now given for (?&) (a zero-length name), and this patch also
805        makes it give the same error for \k'' (previously it complained that that
806        was a reference to a non-existent subpattern).
807    
808    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
809        this is right because (?- can be followed by option settings as well as by
810        digits. I have, however, made the messages clearer.
811    
812    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
813        than the number used in the conditional) now cause a compile-time error.
814        This is actually not compatible with Perl, which accepts such patterns, but
815        treats the conditional as always being FALSE (as PCRE used to), but it
816        seems to me that giving a diagnostic is better.
817    
818    8.  Change "alphameric" to the more common word "alphanumeric" in comments
819        and messages.
820    
821    9.  Fix two occurrences of "backslash" in comments that should have been
822        "backspace".
823    
824    10. Remove two redundant lines of code that can never be obeyed (their function
825        was moved elsewhere).
826    
827    11. The program that makes PCRE's Unicode character property table had a bug
828        which caused it to generate incorrect table entries for sequences of
829        characters that have the same character type, but are in different scripts.
830        It amalgamated them into a single range, with the script of the first of
831        them. In other words, some characters were in the wrong script. There were
832        thirteen such cases, affecting characters in the following ranges:
833    
834          U+002b0 - U+002c1
835          U+0060c - U+0060d
836          U+0061e - U+00612
837          U+0064b - U+0065e
838          U+0074d - U+0076d
839          U+01800 - U+01805
840          U+01d00 - U+01d77
841          U+01d9b - U+01dbf
842          U+0200b - U+0200f
843          U+030fc - U+030fe
844          U+03260 - U+0327f
845          U+0fb46 - U+0fbb1
846          U+10450 - U+1049d
847    
848    12. The -o option (show only the matching part of a line) for pcregrep was not
849        compatible with GNU grep in that, if there was more than one match in a
850        line, it showed only the first of them. It now behaves in the same way as
851        GNU grep.
852    
853    13. If the -o and -v options were combined for pcregrep, it printed a blank
854        line for every non-matching line. GNU grep prints nothing, and pcregrep now
855        does the same. The return code can be used to tell if there were any
856        non-matching lines.
857    
858    14. Added --file-offsets and --line-offsets to pcregrep.
859    
860    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
861        infinitely looping recursion. The bug was that positive lookaheads were not
862        being skipped when checking for a possible empty match (negative lookaheads
863        and both kinds of lookbehind were skipped).
864    
865    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
866        inclusion of <windows.h> to before rather than after the definition of
867        INVALID_FILE_ATTRIBUTES (patch from David Byron).
868    
869    17. Specifying a possessive quantifier with a specific limit for a Unicode
870        character property caused pcre_compile() to compile bad code, which led at
871        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
872        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
873        caused the error; without that there was no problem.
874    
875    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
876    
877    19. Added --enable-pcretest-libreadline.
878    
879    20. In pcrecpp.cc, the variable 'count' was incremented twice in
880        RE::GlobalReplace(). As a result, the number of replacements returned was
881        double what it should be. I removed one of the increments, but Craig sent a
882        later patch that removed the other one (the right fix) and added unit tests
883        that check the return values (which was not done before).
884    
885    21. Several CMake things:
886    
887        (1) Arranged that, when cmake is used on Unix, the libraries end up with
888            the names libpcre and libpcreposix, not just pcre and pcreposix.
889    
890        (2) The above change means that pcretest and pcregrep are now correctly
891            linked with the newly-built libraries, not previously installed ones.
892    
893        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
894    
895    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
896        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
897        UTF-8 newline character). The key issue is that the pattern starts .*;
898        this means that the match must be either at the beginning, or after a
899        newline. The bug was in the code for advancing after a failed match and
900        checking that the new position followed a newline. It was not taking
901        account of UTF-8 characters correctly.
902    
903    23. PCRE was behaving differently from Perl in the way it recognized POSIX
904        character classes. PCRE was not treating the sequence [:...:] as a
905        character class unless the ... were all letters. Perl, however, seems to
906        allow any characters between [: and :], though of course it rejects as
907        unknown any "names" that contain non-letters, because all the known class
908        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
909        for example, whereas PCRE did not - it did not recognize a POSIX character
910        class. This seemed a bit dangerous, so the code has been changed to be
911        closer to Perl. The behaviour is not identical to Perl, because PCRE will
912        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
913        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
914        Perl does, and where it didn't before.
915    
916    24. Rewrite so as to remove the single use of %n from pcregrep because in some
917        Windows environments %n is disabled by default.
918    
919    
920    Version 7.4 21-Sep-07
921    ---------------------
922    
923    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
924        means that a class such as [\s] counted as "explicit reference to CR or
925        LF". That isn't really right - the whole point of the change was to try to
926        help when there was an actual mention of one of the two characters. So now
927        the change happens only if \r or \n (or a literal CR or LF) character is
928        encountered.
929    
930    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
931        of both had grown to the point where there were only 3 bits left.
932        Fortunately, there was spare space in the data structure, and so I have
933        moved the internal flags into a new 16-bit field to free up more option
934        bits.
935    
936    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
937        but did not set the internal JCHANGED flag - either of these is enough to
938        control the way the "get" function works - but the PCRE_INFO_JCHANGED
939        facility is supposed to tell if (?J) was ever used, so now (?J) at the
940        start sets both bits.
941    
942    4.  Added options (at build time, compile time, exec time) to change \R from
943        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
944    
945    5.  doc/pcresyntax.html was missing from the distribution.
946    
947    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
948        compatibility, even though it is no longer used.
949    
950    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
951        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
952        windows.h file is present (where different names are used). [This was
953        reversed later after testing - see 16 below.]
954    
955    8.  Changed all #include <config.h> to #include "config.h". There were also
956        some further <pcre.h> cases that I changed to "pcre.h".
957    
958    9.  When pcregrep was used with the --colour option, it missed the line ending
959        sequence off the lines that it output.
960    
961    10. It was pointed out to me that arrays of string pointers cause lots of
962        relocations when a shared library is dynamically loaded. A technique of
963        using a single long string with a table of offsets can drastically reduce
964        these. I have refactored PCRE in four places to do this. The result is
965        dramatic:
966    
967          Originally:                          290
968          After changing UCP table:            187
969          After changing error message table:   43
970          After changing table of "verbs"       36
971          After changing table of Posix names   22
972    
973        Thanks to the folks working on Gregex for glib for this insight.
974    
975    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
976        unicode-properties was also set.
977    
978    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
979    
980    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
981        checked only for CRLF.
982    
983    14. Added casts to pcretest.c to avoid compiler warnings.
984    
985    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
986    
987    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
988        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
989        entirely. This removes changes made in 7 above.
990    
991    17. The CMake files have been updated, and there is now more information about
992        building with CMake in the NON-UNIX-USE document.
993    
994    
995    Version 7.3 28-Aug-07
996    ---------------------
997    
998     1. In the rejigging of the build system that eventually resulted in 7.1, the
999        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1000        brackets there is not right, since it causes compilers to look for an
1001        installed pcre.h, not the version that is in the source that is being
1002        compiled (which of course may be different). I have changed it back to:
1003    
1004          #include "pcre.h"
1005    
1006        I have a vague recollection that the change was concerned with compiling in
1007        different directories, but in the new build system, that is taken care of
1008        by the VPATH setting the Makefile.
1009    
1010     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1011        when the subject happened to end in the byte 0x85 (e.g. if the last
1012        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1013        characters but of course it shouldn't be taken as a newline when it is part
1014        of another character. The bug was that, for an unlimited repeat of . in
1015        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1016        characters when looking for a newline.
1017    
1018     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1019    
1020     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1021        in debug output.
1022    
1023     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1024        long printing in the pcrecpp unittest when running under MinGW.
1025    
1026     6. ESC_K was left out of the EBCDIC table.
1027    
1028     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1029        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1030        limit also applies to "virtual nesting" when a pattern is recursive, and in
1031        this case 1000 isn't so big. I have been able to remove this limit at the
1032        expense of backing off one optimization in certain circumstances. Normally,
1033        when pcre_exec() would call its internal match() function recursively and
1034        immediately return the result unconditionally, it uses a "tail recursion"
1035        feature to save stack. However, when a subpattern that can match an empty
1036        string has an unlimited repetition quantifier, it no longer makes this
1037        optimization. That gives it a stack frame in which to save the data for
1038        checking that an empty string has been matched. Previously this was taken
1039        from the 1000-entry workspace that had been reserved. So now there is no
1040        explicit limit, but more stack is used.
1041    
1042     8. Applied Daniel's patches to solve problems with the import/export magic
1043        syntax that is required for Windows, and which was going wrong for the
1044        pcreposix and pcrecpp parts of the library. These were overlooked when this
1045        problem was solved for the main library.
1046    
1047     9. There were some crude static tests to avoid integer overflow when computing
1048        the size of patterns that contain repeated groups with explicit upper
1049        limits. As the maximum quantifier is 65535, the maximum group length was
1050        set at 30,000 so that the product of these two numbers did not overflow a
1051        32-bit integer. However, it turns out that people want to use groups that
1052        are longer than 30,000 bytes (though not repeat them that many times).
1053        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1054        made it possible to implement the integer overflow checks in a much more
1055        dynamic way, which I have now done. The artificial limitation on group
1056        length has been removed - we now have only the limit on the total length of
1057        the compiled pattern, which depends on the LINK_SIZE setting.
1058    
1059    10. Fixed a bug in the documentation for get/copy named substring when
1060        duplicate names are permitted. If none of the named substrings are set, the
1061        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1062        empty string.
1063    
1064    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1065        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1066        because the ] is interpreted as the first data character and the
1067        terminating ] is not found. PCRE has been made compatible with Perl in this
1068        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1069        cause memory overwriting.
1070    
1071    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1072        string has been matched (to stop an infinite loop). It was not recognizing
1073        a conditional subpattern that could match an empty string if that
1074        subpattern was within another subpattern. For example, it looped when
1075        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1076        condition was not nested. This bug has been fixed.
1077    
1078    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1079        past the start of the subject in the presence of bytes with the top bit
1080        set, for example "\x8aBCD".
1081    
1082    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1083        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1084    
1085    14. Optimized (?!) to (*FAIL).
1086    
1087    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1088        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1089        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1090        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1091        does: it's just the validity check that is more restrictive.
1092    
1093    16. Inserted checks for integer overflows during escape sequence (backslash)
1094        processing, and also fixed erroneous offset values for syntax errors during
1095        backslash processing.
1096    
1097    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1098        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1099    
1100    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1101        caused an overrun.
1102    
1103    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1104        something other than just ASCII characters) inside a group that had an
1105        unlimited repeat caused a loop at compile time (while checking to see
1106        whether the group could match an empty string).
1107    
1108    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1109        [\P{Any}] did so. (Error in the code for printing property names.)
1110    
1111    21. An orphan \E inside a character class could cause a crash.
1112    
1113    22. A repeated capturing bracket such as (A)? could cause a wild memory
1114        reference during compilation.
1115    
1116    23. There are several functions in pcre_compile() that scan along a compiled
1117        expression for various reasons (e.g. to see if it's fixed length for look
1118        behind). There were bugs in these functions when a repeated \p or \P was
1119        present in the pattern. These operators have additional parameters compared
1120        with \d, etc, and these were not being taken into account when moving along
1121        the compiled data. Specifically:
1122    
1123        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1124            length.
1125    
1126        (b) An item such as \pL+ within a repeated group could cause crashes or
1127            loops.
1128    
1129        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1130            "reference to non-existent subpattern" error.
1131    
1132        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1133    
1134    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1135        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1136    
1137    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1138    
1139    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1140        character were causing crashes (broken optimization).
1141    
1142    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1143        \p or \P) caused a compile-time loop.
1144    
1145    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1146        break. For example, the unstudied pattern [\r\n]A does not match the string
1147        "\r\nA" because change 7.0/46 below moves the current point on by two
1148        characters after failing to match at the start. However, the pattern \nA
1149        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1150        the same is true. There doesn't seem any very clean way out of this, but
1151        what I have chosen to do makes the common cases work: PCRE now takes note
1152        of whether there can be an explicit match for \r or \n anywhere in the
1153        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1154        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1155        pattern has explicit CR or LF references.
1156    
1157    29. Added (*CR) etc for changing newline setting at start of pattern.
1158    
1159    
1160    Version 7.2 19-Jun-07
1161    ---------------------
1162    
1163     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1164        which is apparently normally available under Windows.
1165    
1166     2. Re-jig the pcregrep tests with different newline settings in an attempt
1167        to make them independent of the local environment's newline setting.
1168    
1169     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1170    
1171     4. Some of the "internals" tests were previously cut out when the link size
1172        was not 2, because the output contained actual offsets. The recent new
1173        "Z" feature of pcretest means that these can be cut out, making the tests
1174        usable with all link sizes.
1175    
1176     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1177        stack recursion. This gives a massive performance boost under BSD, but just
1178        a small improvement under Linux. However, it saves one field in the frame
1179        in all cases.
1180    
1181     6. Added more features from the forthcoming Perl 5.10:
1182    
1183        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1184            recursion call. It refers to the nth most recently opened parentheses.
1185    
1186        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1187            to be opened parentheses.
1188    
1189        (c) Conditions that refer to capturing parentheses can be specified
1190            relatively, for example, (?(-2)... or (?(+3)...
1191    
1192        (d) \K resets the start of the current match so that everything before
1193            is not part of it.
1194    
1195        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1196    
1197        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1198            reference syntax.
1199    
1200        (g) (?| introduces a group in which the numbering of parentheses in each
1201            alternative starts with the same number.
1202    
1203        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1204    
1205     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1206        PCRE_INFO_JCHANGED.
1207    
1208     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1209        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1210        for detecting groups that can match an empty string.
1211    
1212     9. A pattern with a very large number of alternatives (more than several
1213        hundred) was running out of internal workspace during the pre-compile
1214        phase, where pcre_compile() figures out how much memory will be needed. A
1215        bit of new cunning has reduced the workspace needed for groups with
1216        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1217        workspace instead of running out of the 4096 that are available.
1218    
1219    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1220    
1221    11. Applied patch from Google to remove an optimization that didn't quite work.
1222        The report of the bug said:
1223    
1224          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1225          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1226          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1227    
1228    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1229        it matched the wrong number of bytes.
1230    
1231    
1232    Version 7.1 24-Apr-07
1233    ---------------------
1234    
1235     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1236        that is more "standard", making use of automake and other Autotools. There
1237        is some re-arrangement of the files and adjustment of comments consequent
1238        on this.
1239    
1240     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1241        for recursive directory scanning broke on some systems because the files
1242        are not scanned in any specific order and on different systems the order
1243        was different. A call to "sort" has been inserted into RunGrepTest for the
1244        approprate test as a short-term fix. In the longer term there may be an
1245        alternative.
1246    
1247     3. I had an email from Eric Raymond about problems translating some of PCRE's
1248        man pages to HTML (despite the fact that I distribute HTML pages, some
1249        people do their own conversions for various reasons). The problems
1250        concerned the use of low-level troff macros .br and .in. I have therefore
1251        removed all such uses from the man pages (some were redundant, some could
1252        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1253        HTML has been updated to handle .nf/.fi and to complain if it encounters
1254        .br or .in.
1255    
1256     4. Updated comments in configure.ac that get placed in config.h.in and also
1257        arranged for config.h to be included in the distribution, with the name
1258        config.h.generic, for the benefit of those who have to compile without
1259        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1260    
1261     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1262        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1263        makevp.bat for the current PCRE, using the additional files
1264        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1265    
1266     6. A Windows user reported a minor discrepancy with test 2, which turned out
1267        to be caused by a trailing space on an input line that had got lost in his
1268        copy. The trailing space was an accident, so I've just removed it.
1269    
1270     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1271        that is needed.
1272    
1273     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1274        as "const" (a) because they are and (b) because it helps the PHP
1275        maintainers who have recently made a script to detect big data structures
1276        in the php code that should be moved to the .rodata section. I remembered
1277        to update Builducptable as well, so it won't revert if ucptable.h is ever
1278        re-created.
1279    
1280     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1281        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1282        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1283        support is not required. This saves 1.5-2K of code, which is important in
1284        some applications.
1285    
1286        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1287        so as not to refer to the tables, even though these functions will never be
1288        called when UTF-8 support is disabled. Otherwise there are problems with a
1289        shared library.
1290    
1291    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1292    
1293        (a) It was defining its arguments as char * instead of void *.
1294    
1295        (b) It was assuming that all moves were upwards in memory; this was true
1296            a long time ago when I wrote it, but is no longer the case.
1297    
1298        The emulated memove() is provided for those environments that have neither
1299        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1300        is clearly not the case, as these two bugs were recently reported.
1301    
1302    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1303        and Detrail to create the HTML documentation, the .txt form of the man
1304        pages, and it removes trailing spaces from listed files. It also creates
1305        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1306        case, it wraps all the #defines with #ifndefs. This script should be run
1307        before "make dist".
1308    
1309    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1310        with Unicode property support.
1311    
1312        (a) For a maximizing quantifier, if the two different cases of the
1313            character were of different lengths in their UTF-8 codings (there are
1314            some cases like this - I found 11), and the matching function had to
1315            back up over a mixture of the two cases, it incorrectly assumed they
1316            were both the same length.
1317    
1318        (b) When PCRE was configured to use the heap rather than the stack for
1319            recursion during matching, it was not correctly preserving the data for
1320            the other case of a UTF-8 character when checking ahead for a match
1321            while processing a minimizing repeat. If the check also involved
1322            matching a wide character, but failed, corruption could cause an
1323            erroneous result when trying to check for a repeat of the original
1324            character.
1325    
1326    13. Some tidying changes to the testing mechanism:
1327    
1328        (a) The RunTest script now detects the internal link size and whether there
1329            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1330            values substituted by "configure". (The RunGrepTest script already did
1331            this for UTF-8.) The configure.ac script no longer substitutes the
1332            relevant variables.
1333    
1334        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1335            with length and offset values. This means that the output is different
1336            for different internal link sizes. Test 2 is skipped for link sizes
1337            other than 2 because of this, bypassing the problem. Unfortunately,
1338            there was also a test in test 3 (the locale tests) that used /B and
1339            failed for link sizes other than 2. Rather than cut the whole test out,
1340            I have added a new /Z option to pcretest that replaces the length and
1341            offset values with spaces. This is now used to make test 3 independent
1342            of link size. (Test 2 will be tidied up later.)
1343    
1344    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1345        segmentation fault instead of returning the appropriate error message.
1346    
1347    15. In multiline mode when the newline sequence was set to "any", the pattern
1348        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1349        This doesn't seem right; it now treats the CRLF combination as the line
1350        ending, and so does not match in that case. It's only a pattern such as ^$
1351        that would hit this one: something like ^ABC$ would have failed after \r
1352        and then tried again after \r\n.
1353    
1354    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1355        in an attempt to make files that differ only in their line terminators
1356        compare equal. This works on Linux.
1357    
1358    17. Under certain error circumstances pcregrep might try to free random memory
1359        as it exited. This is now fixed, thanks to valgrind.
1360    
1361    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1362        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1363        was because its rules for how to advance for /g after matching an empty
1364        string at the end of a line did not allow for this case. They now check for
1365        it specially.
1366    
1367    20. pcretest is supposed to handle patterns and data of any length, by
1368        extending its buffers when necessary. It was getting this wrong when the
1369        buffer for a data line had to be extended.
1370    
1371    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1372        CRLF as a newline sequence.
1373    
1374    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1375        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1376        I have nevertheless tidied it up.
1377    
1378    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1379    
1380    24. Added a man page for pcre-config.
1381    
1382    
1383    Version 7.0 19-Dec-06
1384    ---------------------
1385    
1386     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1387        moving to gcc 4.1.1.
1388    
1389     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1390        sys/time.h, which is documented as needed for this function. It doesn't
1391        seem to matter on Linux, but it showed up on some releases of OS X.
1392    
1393     3. It seems that there are systems where bytes whose values are greater than
1394        127 match isprint() in the "C" locale. The "C" locale should be the
1395        default when a C program starts up. In most systems, only ASCII printing
1396        characters match isprint(). This difference caused the output from pcretest
1397        to vary, making some of the tests fail. I have changed pcretest so that:
1398    
1399        (a) When it is outputting text in the compiled version of a pattern, bytes
1400            other than 32-126 are always shown as hex escapes.
1401    
1402        (b) When it is outputting text that is a matched part of a subject string,
1403            it does the same, unless a different locale has been set for the match
1404            (using the /L modifier). In this case, it uses isprint() to decide.
1405    
1406     4. Fixed a major bug that caused incorrect computation of the amount of memory
1407        required for a compiled pattern when options that changed within the
1408        pattern affected the logic of the preliminary scan that determines the
1409        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1410        that the computed length was too small. The symptoms of this bug were
1411        either the PCRE error "internal error: code overflow" from pcre_compile(),
1412        or a glibc crash with a message such as "pcretest: free(): invalid next
1413        size (fast)". Examples of patterns that provoked this bug (shown in
1414        pcretest format) are:
1415    
1416          /(?-x: )/x
1417          /(?x)(?-x: \s*#\s*)/
1418          /((?i)[\x{c0}])/8
1419          /(?i:[\x{c0}])/8
1420    
1421        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1422        is now done differently.
1423    
1424     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1425        wrapper classes; (b) implement a new function in the C++ scanner that is
1426        more efficient than the old way of doing things because it avoids levels of
1427        recursion in the regex matching; (c) add a paragraph to the documentation
1428        for the FullMatch() function.
1429    
1430     6. The escape sequence \n was being treated as whatever was defined as
1431        "newline". Not only was this contrary to the documentation, which states
1432        that \n is character 10 (hex 0A), but it also went horribly wrong when
1433        "newline" was defined as CRLF. This has been fixed.
1434    
1435     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1436        was being set to -1 for the "end of line" case (supposedly a value that no
1437        character can have). Though this value is never used (the check for end of
1438        line is "zero bytes in current character"), it caused compiler complaints.
1439        I've changed it to 0xffffffff.
1440    
1441     8. In pcre_version.c, the version string was being built by a sequence of
1442        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1443        string (as it is for production releases) called a macro with an empty
1444        argument. The C standard says the result of this is undefined. The gcc
1445        compiler treats it as an empty string (which was what was wanted) but it is
1446        reported that Visual C gives an error. The source has been hacked around to
1447        avoid this problem.
1448    
1449     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1450        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1451        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1452        of them did).
1453    
1454    10. Originally, pcretest opened its input and output without "b"; then I was
1455        told that "b" was needed in some environments, so it was added for release
1456        5.0 to both the input and output. (It makes no difference on Unix-like
1457        systems.) Later I was told that it is wrong for the input on Windows. I've
1458        now abstracted the modes into two macros, to make it easier to fiddle with
1459        them, and removed "b" from the input mode under Windows.
1460    
1461    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1462    
1463    12. Added -help and --help to pcretest as an official way of being reminded
1464        of the options.
1465    
1466    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1467        and pcrecpp.cc because they annoy compilers at high warning levels.
1468    
1469    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1470    
1471    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1472        scripts are not C programs :-) and which was not noticed because it works
1473        on Linux.
1474    
1475    16. pcretest is supposed to handle any length of pattern and data line (as one
1476        line or as a continued sequence of lines) by extending its input buffer if
1477        necessary. This feature was broken for very long pattern lines, leading to
1478        a string of junk being passed to pcre_compile() if the pattern was longer
1479        than about 50K.
1480    
1481    17. I have done a major re-factoring of the way pcre_compile() computes the
1482        amount of memory needed for a compiled pattern. Previously, there was code
1483        that made a preliminary scan of the pattern in order to do this. That was
1484        OK when PCRE was new, but as the facilities have expanded, it has become
1485        harder and harder to keep it in step with the real compile phase, and there
1486        have been a number of bugs (see for example, 4 above). I have now found a
1487        cunning way of running the real compile function in a "fake" mode that
1488        enables it to compute how much memory it would need, while actually only
1489        ever using a few hundred bytes of working memory and without too many
1490        tests of the mode. This should make future maintenance and development
1491        easier. A side effect of this work is that the limit of 200 on the nesting
1492        depth of parentheses has been removed (though this was never a serious
1493        limitation, I suspect). However, there is a downside: pcre_compile() now
1494        runs more slowly than before (30% or more, depending on the pattern). I
1495        hope this isn't a big issue. There is no effect on runtime performance.
1496    
1497    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1498        newline (only possible for the last line of a file) and it was a
1499        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1500    
1501    19. Added additional timing features to pcretest. (1) The -tm option now times
1502        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1503        separate command line item, by a number that specifies the number of
1504        repeats to use when timing. The default is 50000; this gives better
1505        precision, but takes uncomfortably long for very large patterns.
1506    
1507    20. Extended pcre_study() to be more clever in cases where a branch of a
1508        subpattern has no definite first character. For example, (a*|b*)[cd] would
1509        previously give no result from pcre_study(). Now it recognizes that the
1510        first character must be a, b, c, or d.
1511    
1512    21. There was an incorrect error "recursive call could loop indefinitely" if
1513        a subpattern (or the entire pattern) that was being tested for matching an
1514        empty string contained only one non-empty item after a nested subpattern.
1515        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1516        incorrectly, because the \d was being skipped in the check.
1517    
1518    22. The pcretest program now has a new pattern option /B and a command line
1519        option -b, which is equivalent to adding /B to every pattern. This causes
1520        it to show the compiled bytecode, without the additional information that
1521        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1522        is the same as /B/I).
1523    
1524    23. A new optimization is now able automatically to treat some sequences such
1525        as a*b as a*+b. More specifically, if something simple (such as a character
1526        or a simple class like \d) has an unlimited quantifier, and is followed by
1527        something that cannot possibly match the quantified thing, the quantifier
1528        is automatically "possessified".
1529    
1530    24. A recursive reference to a subpattern whose number was greater than 39
1531        went wrong under certain circumstances in UTF-8 mode. This bug could also
1532        have affected the operation of pcre_study().
1533    
1534    25. Realized that a little bit of performance could be had by replacing
1535        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1536    
1537    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1538    
1539    27. Possessive quantifiers such as a++ were previously implemented by turning
1540        them into atomic groups such as ($>a+). Now they have their own opcodes,
1541        which improves performance. This includes the automatically created ones
1542        from 23 above.
1543    
1544    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1545        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1546        the first matched character to be a colon. This applied both to named and
1547        numbered groups.
1548    
1549    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1550    
1551    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1552        building PCRE on Windows easier, so I have included it in the distribution.
1553    
1554    31. There is now a check in pcretest against a ridiculously large number being
1555        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1556        loop, the loop is abandoned.
1557    
1558    32. Forward references to subpatterns in conditions such as (?(2)...) where
1559        subpattern 2 is defined later cause pcre_compile() to search forwards in
1560        the pattern for the relevant set of parentheses. This search went wrong
1561        when there were unescaped parentheses in a character class, parentheses
1562        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1563    
1564    33. "Subroutine" calls and backreferences were previously restricted to
1565        referencing subpatterns earlier in the regex. This restriction has now
1566        been removed.
1567    
1568    34. Added a number of extra features that are going to be in Perl 5.10. On the
1569        whole, these are just syntactic alternatives for features that PCRE had
1570        previously implemented using the Python syntax or my own invention. The
1571        other formats are all retained for compatibility.
1572    
1573        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1574            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1575            also .NET compatible.
1576    
1577        (b) A recursion or subroutine call to a named group can now be defined as
1578            (?&name) as well as (?P>name).
1579    
1580        (c) A backreference to a named group can now be defined as \k<name> or
1581            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1582            5.10, are also .NET compatible.
1583    
1584        (d) A conditional reference to a named group can now use the syntax
1585            (?(<name>) or (?('name') as well as (?(name).
1586    
1587        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1588            groups (named and numbered) that are never evaluated inline, but can be
1589            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1590            is always false. There may be only one alternative in such a group.
1591    
1592        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1593            as the simple (?(R). The condition is true only if the most recent
1594            recursion is that of the given number or name. It does not search out
1595            through the entire recursion stack.
1596    
1597        (g) The escape \gN or \g{N} has been added, where N is a positive or
1598            negative number, specifying an absolute or relative reference.
1599    
1600    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1601        some "unreachable code" warnings.
1602    
1603    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1604        things, this adds five new scripts.
1605    
1606    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1607        There were also incompatibilities regarding the handling of \Q..\E inside
1608        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1609        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1610    
1611    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1612        matches an empty string, and forcibly breaks the loop. There were bugs in
1613        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1614        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1615        separate and independent bugs (that affected different cases) have been
1616        fixed.
1617    
1618    39. Refactored the code to abolish the use of different opcodes for small
1619        capturing bracket numbers. This is a tidy that I avoided doing when I
1620        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1621        The new approach is not only tidier, it makes it possible to reduce the
1622        memory needed to fix the previous bug (38).
1623    
1624    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1625        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1626        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1627        mode.
1628    
1629    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1630        report.
1631    
1632    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1633        copy construction and assignment in the C++ wrapper.
1634    
1635    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1636        couple of bugs that could have given wrong results in the "--newline=crlf"
1637        case.
1638    
1639    44. Added a number of casts and did some reorganization of signed/unsigned int
1640        variables following suggestions from Dair Grant. Also renamed the variable
1641        "this" as "item" because it is a C++ keyword.
1642    
1643    45. Arranged for dftables to add
1644    
1645          #include "pcre_internal.h"
1646    
1647        to pcre_chartables.c because without it, gcc 4.x may remove the array
1648        definition from the final binary if PCRE is built into a static library and
1649        dead code stripping is activated.
1650    
1651    46. For an unanchored pattern, if a match attempt fails at the start of a
1652        newline sequence, and the newline setting is CRLF or ANY, and the next two
1653        characters are CRLF, advance by two characters instead of one.
1654    
1655    
1656    Version 6.7 04-Jul-06
1657    ---------------------
1658    
1659     1. In order to handle tests when input lines are enormously long, pcretest has
1660        been re-factored so that it automatically extends its buffers when
1661        necessary. The code is crude, but this _is_ just a test program. The
1662        default size has been increased from 32K to 50K.
1663    
1664     2. The code in pcre_study() was using the value of the re argument before
1665        testing it for NULL. (Of course, in any sensible call of the function, it
1666        won't be NULL.)
1667    
1668     3. The memmove() emulation function in pcre_internal.h, which is used on
1669        systems that lack both memmove() and bcopy() - that is, hardly ever -
1670        was missing a "static" storage class specifier.
1671    
1672     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1673        containing an extended class (one that cannot be represented by a bitmap
1674        because it contains high-valued characters or Unicode property items, e.g.
1675        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1676        pattern, but PCRE should not loop if you do not (it no longer does).
1677        [Detail: two cases were found: (a) a repeated subpattern containing an
1678        extended class; (b) a recursive reference to a subpattern that followed a
1679        previous extended class. It wasn't skipping over the extended class
1680        correctly when UTF-8 mode was not set.]
1681    
1682     5. A negated single-character class was not being recognized as fixed-length
1683        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1684        compile error "lookbehind assertion is not fixed length".
1685    
1686     6. The RunPerlTest auxiliary script was showing an unexpected difference
1687        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1688        write a Perl script that can interpret lines of an input file either as
1689        byte characters or as UTF-8, which is what "perltest" was being required to
1690        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1691        can't do is switch easily at run time between having the "use utf8;" pragma
1692        or not. In the end, I fudged it by using the RunPerlTest script to insert
1693        "use utf8;" explicitly for the UTF-8 tests.
1694    
1695     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1696        the end of the subject string, contrary to the documentation and to what
1697        Perl does. This was true of both matching functions. Now it matches only at
1698        the start of the subject and immediately after *internal* newlines.
1699    
1700     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1701        a pointer to an int instead of a pointer to an unsigned long int. This
1702        caused problems on 64-bit systems.
1703    
1704     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1705        instance of the 'standard' template library not being so standard".
1706    
1707    10. There was no check on the number of named subpatterns nor the maximum
1708        length of a subpattern name. The product of these values is used to compute
1709        the size of the memory block for a compiled pattern. By supplying a very
1710        long subpattern name and a large number of named subpatterns, the size
1711        computation could be caused to overflow. This is now prevented by limiting
1712        the length of names to 32 characters, and the number of named subpatterns
1713        to 10,000.
1714    
1715    11. Subpatterns that are repeated with specific counts have to be replicated in
1716        the compiled pattern. The size of memory for this was computed from the
1717        length of the subpattern and the repeat count. The latter is limited to
1718        65535, but there was no limit on the former, meaning that integer overflow
1719        could in principle occur. The compiled length of a repeated subpattern is
1720        now limited to 30,000 bytes in order to prevent this.
1721    
1722    12. Added the optional facility to have named substrings with the same name.
1723    
1724    13. Added the ability to use a named substring as a condition, using the
1725        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1726        are numbers (not recommended). Forward references are permitted.
1727    
1728    14. Added forward references in named backreferences (if you see what I mean).
1729    
1730    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1731        pattern could run off the end of the subject. For example, the pattern
1732        "(?s)(.{1,5})"8 did this with the subject "ab".
1733    
1734    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1735        PCRE_CASELESS was set when matching characters that were quantified with ?
1736        or *.
1737    
1738    17. A character class other than a single negated character that had a minimum
1739        but no maximum quantifier - for example [ab]{6,} - was not handled
1740        correctly by pce_dfa_exec(). It would match only one character.
1741    
1742    18. A valid (though odd) pattern that looked like a POSIX character
1743        class but used an invalid character after [ (for example [[,abc,]]) caused
1744        pcre_compile() to give the error "Failed: internal error: code overflow" or
1745        in some cases to crash with a glibc free() error. This could even happen if
1746        the pattern terminated after [[ but there just happened to be a sequence of
1747        letters, a binary zero, and a closing ] in the memory that followed.
1748    
1749    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1750        over the years. Originally (before any Unicode support), just the bottom 8
1751        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1752        output from "man perlunicode" includes this:
1753    
1754          The regular expression compiler produces polymorphic opcodes.  That
1755          is, the pattern adapts to the data and automatically switches to
1756          the Unicode character scheme when presented with Unicode data--or
1757          instead uses a traditional byte scheme when presented with byte
1758          data.
1759    
1760        Sadly, a wide octal escape does not cause a switch, and in a string with
1761        no other multibyte characters, these octal escapes are treated as before.
1762        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1763        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1764        Unicode string.
1765    
1766        I have not perpetrated such confusion in PCRE. Up till now, it took just
1767        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1768        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1769        translate to the appropriate multibyte character.
1770    
1771    29. Applied some refactoring to reduce the number of warnings from Microsoft
1772        and Borland compilers. This has included removing the fudge introduced
1773        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1774        a warning about an unused variable.
1775    
1776    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1777        characters since release 4.0, because Perl (from release 5.004) does not.
1778        [Or at least, is documented not to: some releases seem to be in conflict
1779        with the documentation.] However, when a pattern was studied with
1780        pcre_study() and all its branches started with \s, PCRE still included VT
1781        as a possible starting character. Of course, this did no harm; it just
1782        caused an unnecessary match attempt.
1783    
1784    22. Removed a now-redundant internal flag bit that recorded the fact that case
1785        dependency changed within the pattern. This was once needed for "required
1786        byte" processing, but is no longer used. This recovers a now-scarce options
1787        bit. Also moved the least significant internal flag bit to the most-
1788        significant bit of the word, which was not previously used (hangover from
1789        the days when it was an int rather than a uint) to free up another bit for
1790        the future.
1791    
1792    23. Added support for CRLF line endings as well as CR and LF. As well as the
1793        default being selectable at build time, it can now be changed at runtime
1794        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1795        specify that it is scanning data with non-default line endings.
1796    
1797    24. Changed the definition of CXXLINK to make it agree with the definition of
1798        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1799    
1800    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1801        recursions. This makes a big different to stack usage for some patterns.
1802    
1803    26. If a subpattern containing a named recursion or subroutine reference such
1804        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1805        the space required for the compiled pattern went wrong and gave too small a
1806        value. Depending on the environment, this could lead to "Failed: internal
1807        error: code overflow at offset 49" or "glibc detected double free or
1808        corruption" errors.
1809    
1810    27. Applied patches from Google (a) to support the new newline modes and (b) to
1811        advance over multibyte UTF-8 characters in GlobalReplace.
1812    
1813    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1814        difference for some implementation of PCRE in some Windows version.
1815    
1816    29. Added some extra testing facilities to pcretest:
1817    
1818        \q<number>   in a data line sets the "match limit" value
1819        \Q<number>   in a data line sets the "match recursion limt" value
1820        -S <number>  sets the stack size, where <number> is in megabytes
1821    
1822        The -S option isn't available for Windows.
1823    
1824    
1825    Version 6.6 06-Feb-06
1826    ---------------------
1827    
1828     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1829        in pcreposix.h. I have copied the definition from pcre.h.
1830    
1831     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1832        because pcre.h is no longer a built file.
1833    
1834     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1835        not normally included in the compiled code.
1836    
1837    
1838    Version 6.5 01-Feb-06
1839    ---------------------
1840    
1841     1. When using the partial match feature with pcre_dfa_exec(), it was not
1842        anchoring the second and subsequent partial matches at the new starting
1843        point. This could lead to incorrect results. For example, with the pattern
1844        /1234/, partially matching against "123" and then "a4" gave a match.
1845    
1846     2. Changes to pcregrep:
1847    
1848        (a) All non-match returns from pcre_exec() were being treated as failures
1849            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1850            error message is output. Some extra information is given for the
1851            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1852            probably the only errors that are likely to be caused by users (by
1853            specifying a regex that has nested indefinite repeats, for instance).
1854            If there are more than 20 of these errors, pcregrep is abandoned.
1855    
1856        (b) A binary zero was treated as data while matching, but terminated the
1857            output line if it was written out. This has been fixed: binary zeroes
1858            are now no different to any other data bytes.
1859    
1860        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1861            used to set a locale for matching. The --locale=xxxx long option has
1862            been added (no short equivalent) to specify a locale explicitly on the
1863            pcregrep command, overriding the environment variables.
1864    
1865        (d) When -B was used with -n, some line numbers in the output were one less
1866            than they should have been.
1867    
1868        (e) Added the -o (--only-matching) option.
1869    
1870        (f) If -A or -C was used with -c (count only), some lines of context were
1871            accidentally printed for the final match.
1872    
1873        (g) Added the -H (--with-filename) option.
1874    
1875        (h) The combination of options -rh failed to suppress file names for files
1876            that were found from directory arguments.
1877    
1878        (i) Added the -D (--devices) and -d (--directories) options.
1879    
1880        (j) Added the -F (--fixed-strings) option.
1881    
1882        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1883    
1884        (l) Added the --colo(u)r option.
1885    
1886        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1887            is not present by default.
1888    
1889     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1890        items such as (?R) or (?1), when the recursion could match a number of
1891        alternatives. If it matched one of the alternatives, but subsequently,
1892        outside the recursion, there was a failure, the code tried to back up into
1893        the recursion. However, because of the way PCRE is implemented, this is not
1894        possible, and the result was an incorrect result from the match.
1895    
1896        In order to prevent this happening, the specification of recursion has
1897        been changed so that all such subpatterns are automatically treated as
1898        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1899    
1900     4. I had overlooked the fact that, in some locales, there are characters for
1901        which isalpha() is true but neither isupper() nor islower() are true. In
1902        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1903        and ordfeminine) are like this. This affected the treatment of \w and \W
1904        when they appeared in character classes, but not when they appeared outside
1905        a character class. The bit map for "word" characters is now created
1906        separately from the results of isalnum() instead of just taking it from the
1907        upper, lower, and digit maps. (Plus the underscore character, of course.)
1908    
1909     5. The above bug also affected the handling of POSIX character classes such as
1910        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1911        permanent tables. Instead, the bit maps for such a class were previously
1912        created as the appropriate unions of the upper, lower, and digit bitmaps.
1913        Now they are created by subtraction from the [[:word:]] class, which has
1914        its own bitmap.
1915    
1916     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1917        It is created by subtracting the vertical space characters (\x09, \x0a,
1918        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1919        subtraction was done in the overall bitmap for a character class, meaning
1920        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1921        be recognized. This bug has been fixed.
1922    
1923     7. Patches from the folks at Google:
1924    
1925          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1926          real life, but is still worth protecting against".
1927    
1928          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1929          regular expressions".
1930    
1931          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1932          have it.
1933    
1934          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1935          "configure" and the latter not, in order to fix a problem somebody had
1936          with compiling the Arg class on HP-UX.
1937    
1938          (e) Improve the error-handling of the C++ wrapper a little bit.
1939    
1940          (f) New tests for checking recursion limiting.
1941    
1942     8. The pcre_memmove() function, which is used only if the environment does not
1943        have a standard memmove() function (and is therefore rarely compiled),
1944        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1945        returning a result (though PCRE never actually uses the result).
1946    
1947     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1948        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1949        returned instead of calling malloc() with an overflowing number that would
1950        most likely cause subsequent chaos.
1951    
1952    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1953    
1954    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1955        with this option is matched, the nmatch and pmatch options of regexec() are
1956        ignored.
1957    
1958    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1959        provided in case anyone wants to the the POSIX interface with UTF-8
1960        strings.
1961    
1962    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1963        C++ linking (needed for some HP-UX environments).
1964    
1965    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1966        (unused parameter) and in the pcre_printint() function (omitted "default"
1967        switch label when the default is to do nothing).
1968    
1969    15. Added some code to make it possible, when PCRE is compiled as a C++
1970        library, to replace subject pointers for pcre_exec() with a smart pointer
1971        class, thus making it possible to process discontinuous strings.
1972    
1973    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1974        much the same function. They were added by different people who were trying
1975        to make PCRE easy to compile on non-Unix systems. It has been suggested
1976        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1977        for compiling on Windows systems. I have therefore replaced it with
1978        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1979        defaults to "extern" for C or "extern C" for C++, which works fine on
1980        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1981        SCOPE with something explicit in config.h. In addition:
1982    
1983        (a) pcreposix.h still had just "extern" instead of either of these macros;
1984            I have replaced it with PCRE_DATA_SCOPE.
1985    
1986        (b) Functions such as _pcre_xclass(), which are internal to the library,
1987            but external in the C sense, all had PCRE_EXPORT in their definitions.
1988            This is apparently wrong for the Windows case, so I have removed it.
1989            (It makes no difference on Unix-like systems.)
1990    
1991    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1992        of recursive calls to match(). This is different to MATCH_LIMIT because
1993        that limits the total number of calls to match(), not all of which increase
1994        the depth of recursion. Limiting the recursion depth limits the amount of
1995        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1996        when PCRE is compiled, and changed at run time. A patch from Google adds
1997        this functionality to the C++ interface.
1998    
1999    18. Changes to the handling of Unicode character properties:
2000    
2001        (a) Updated the table to Unicode 4.1.0.
2002    
2003        (b) Recognize characters that are not in the table as "Cn" (undefined).
2004    
2005        (c) I revised the way the table is implemented to a much improved format
2006            which includes recognition of ranges. It now supports the ranges that
2007            are defined in UnicodeData.txt, and it also amalgamates other
2008            characters into ranges. This has reduced the number of entries in the
2009            table from around 16,000 to around 3,000, thus reducing its size
2010            considerably. I realized I did not need to use a tree structure after
2011            all - a binary chop search is just as efficient. Having reduced the
2012            number of entries, I extended their size from 6 bytes to 8 bytes to
2013            allow for more data.
2014    
2015        (d) Added support for Unicode script names via properties such as \p{Han}.
2016    
2017    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2018        matching that character.
2019    
2020    20. When matching a repeated Unicode property with a minimum greater than zero,
2021        (for example \pL{2,}), PCRE could look past the end of the subject if it
2022        reached it while seeking the minimum number of characters. This could
2023        happen only if some of the characters were more than one byte long, because
2024        there is a check for at least the minimum number of bytes.
2025    
2026    21. Refactored the implementation of \p and \P so as to be more general, to
2027        allow for more different types of property in future. This has changed the
2028        compiled form incompatibly. Anybody with saved compiled patterns that use
2029        \p or \P will have to recompile them.
2030    
2031    22. Added "Any" and "L&" to the supported property types.
2032    
2033    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2034        but give a compile time error if the value is greater than 0xff.
2035    
2036    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2037        accidentally not being installed or uninstalled.
2038    
2039    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2040        made were to insert the current release number. This seemed silly, because
2041        it made things harder for people building PCRE on systems that don't run
2042        "configure". I have turned pcre.h into a distributed file, no longer built
2043        by "configure", with the version identification directly included. There is
2044        no longer a pcre.h.in file.
2045    
2046        However, this change necessitated a change to the pcre-config script as
2047        well. It is built from pcre-config.in, and one of the substitutions was the
2048        release number. I have updated configure.ac so that ./configure now finds
2049        the release number by grepping pcre.h.
2050    
2051    26. Added the ability to run the tests under valgrind.
2052    
2053    
2054    Version 6.4 05-Sep-05
2055    ---------------------
2056    
2057     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2058        "--" to be printed when multiple files were scanned, even when none of the
2059        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2060        consider it to be a bug, and have restored the previous behaviour.
2061    
2062     2. A couple of code tidies to get rid of compiler warnings.
2063    
2064     3. The pcretest program used to cheat by referring to symbols in the library
2065        whose names begin with _pcre_. These are internal symbols that are not
2066        really supposed to be visible externally, and in some environments it is
2067        possible to suppress them. The cheating is now confined to including
2068        certain files from the library's source, which is a bit cleaner.
2069    
2070     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2071        file's purpose clearer.
2072    
2073     5. Reorganized pcre_ucp_findchar().
2074    
2075    
2076    Version 6.3 15-Aug-05
2077    ---------------------
2078    
2079     1. The file libpcre.pc.in did not have general read permission in the tarball.
2080    
2081     2. There were some problems when building without C++ support:
2082    
2083        (a) If C++ support was not built, "make install" and "make test" still
2084            tried to test it.
2085    
2086        (b) There were problems when the value of CXX was explicitly set. Some
2087            changes have been made to try to fix these, and ...
2088    
2089        (c) --disable-cpp can now be used to explicitly disable C++ support.
2090    
2091        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2092            backslash in a target when C++ was disabled. This confuses some
2093            versions of "make", apparently. Using an intermediate variable solves
2094            this. (Same for CPP_LOBJ.)
2095    
2096     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2097        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2098        necessary on certain architectures.
2099    
2100     4. Added a setting of -export-symbols-regex to the link command to remove
2101        those symbols that are exported in the C sense, but actually are local
2102        within the library, and not documented. Their names all begin with
2103        "_pcre_". This is not a perfect job, because (a) we have to except some
2104        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2105        available (and never for static libraries). I have made a note to try to
2106        find a way round (a) in the future.
2107    
2108    
2109    Version 6.2 01-Aug-05
2110    ---------------------
2111    
2112     1. There was no test for integer overflow of quantifier values. A construction
2113        such as {1111111111111111} would give undefined results. What is worse, if
2114        a minimum quantifier for a parenthesized subpattern overflowed and became
2115        negative, the calculation of the memory size went wrong. This could have
2116        led to memory overwriting.
2117    
2118     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2119    
2120     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2121        operating environments where this matters.
2122    
2123     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2124        PCRE options from within the C++ wrapper.
2125    
2126     5. Named capturing subpatterns were not being correctly counted when a pattern
2127        was compiled. This caused two problems: (a) If there were more than 100
2128        such subpatterns, the calculation of the memory needed for the whole
2129        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2130        back references of the form \12, where the number was greater than 9, were
2131        not recognized as back references, even though there were sufficient
2132        previous subpatterns.
2133    
2134     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2135        versions of gcc, e.g. 2.95.4.
2136    
2137    
2138    Version 6.1 21-Jun-05
2139    ---------------------
2140    
2141     1. There was one reference to the variable "posix" in pcretest.c that was not
2142        surrounded by "#if !defined NOPOSIX".
2143    
2144     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2145        the cross-check on the old pcre_info() function, for the benefit of the
2146        cut-down version of PCRE that is currently imported into Exim.
2147    
2148     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2149        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2150        patterns that start (?i) but I don't think that matters. The use of (?i) is
2151        just an example; this all applies to the other options as well.
2152    
2153     4. Since libtool seems to echo the compile commands it is issuing, the output
2154        from "make" can be reduced a bit by putting "@" in front of each libtool
2155        compile command.
2156    
2157     5. Patch from the folks at Google for configure.in to be a bit more thorough
2158        in checking for a suitable C++ installation before trying to compile the
2159        C++ stuff. This should fix a reported problem when a compiler was present,
2160        but no suitable headers.
2161    
2162     6. The man pages all had just "PCRE" as their title. I have changed them to
2163        be the relevant file name. I have also arranged that these names are
2164        retained in the file doc/pcre.txt, which is a concatenation in text format
2165        of all the man pages except the little individual ones for each function.
2166    
2167     7. The NON-UNIX-USE file had not been updated for the different set of source
2168        files that come with release 6. I also added a few comments about the C++
2169        wrapper.
2170    
2171    
2172    Version 6.0 07-Jun-05
2173    ---------------------
2174    
2175     1. Some minor internal re-organization to help with my DFA experiments.
2176    
2177     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2178        didn't matter for the library itself when fully configured, but did matter
2179        when compiling without UCP support, or within Exim, where the ucp files are
2180        not imported.
2181    
2182     3. Refactoring of the library code to split up the various functions into
2183        different source modules. The addition of the new DFA matching code (see
2184        below) to a single monolithic source would have made it really too
2185        unwieldy, quite apart from causing all the code to be include in a
2186        statically linked application, when only some functions are used. This is
2187        relevant even without the DFA addition now that patterns can be compiled in
2188        one application and matched in another.
2189    
2190        The downside of splitting up is that there have to be some external
2191        functions and data tables that are used internally in different modules of
2192        the library but which are not part of the API. These have all had their
2193        names changed to start with "_pcre_" so that they are unlikely to clash
2194        with other external names.
2195    
2196     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2197        a different (DFA) algorithm. Although it is slower than the original
2198        function, it does have some advantages for certain types of matching
2199        problem.
2200    
2201     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2202        including restarting after a partial match.
2203    
2204     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2205        defined when compiling for Windows was sent to me. I have put it into the
2206        code, though I have no means of testing or verifying it.
2207    
2208     7. Added the pcre_refcount() auxiliary function.
2209    
2210     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2211        match before or at the first newline in the subject string. In pcretest,
2212        the /f option on a pattern can be used to set this.
2213    
2214     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2215        would behave wrongly. This has been present in PCRE since release 4.0.
2216    
2217    10. A number of changes to the pcregrep command:
2218    
2219        (a) Refactored how -x works; insert ^(...)$ instead of setting
2220            PCRE_ANCHORED and checking the length, in preparation for adding
2221            something similar for -w.
2222    
2223        (b) Added the -w (match as a word) option.
2224    
2225        (c) Refactored the way lines are read and buffered so as to have more
2226            than one at a time available.
2227    
2228        (d) Implemented a pcregrep test script.
2229    
2230        (e) Added the -M (multiline match) option. This allows patterns to match
2231            over several lines of the subject. The buffering ensures that at least
2232            8K, or the rest of the document (whichever is the shorter) is available
2233            for matching (and similarly the previous 8K for lookbehind assertions).
2234    
2235        (f) Changed the --help output so that it now says
2236    
2237              -w, --word-regex(p)
2238    
2239            instead of two lines, one with "regex" and the other with "regexp"
2240            because that confused at least one person since the short forms are the
2241            same. (This required a bit of code, as the output is generated
2242            automatically from a table. It wasn't just a text change.)
2243    
2244        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2245            option but starts with a hyphen. Could be a pattern or a path name
2246            starting with a hyphen, for instance.
2247    
2248        (h) "-" can be given as a file name to represent stdin.
2249    
2250        (i) When file names are being printed, "(standard input)" is used for
2251            the standard input, for compatibility with GNU grep. Previously
2252            "<stdin>" was used.
2253    
2254        (j) The option --label=xxx can be used to supply a name to be used for
2255            stdin when file names are being printed. There is no short form.
2256    
2257        (k) Re-factored the options decoding logic because we are going to add
2258            two more options that take data. Such options can now be given in four
2259            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2260    
2261        (l) Added the -A, -B, and -C options for requesting that lines of context
2262            around matches be printed.
2263    
2264        (m) Added the -L option to print the names of files that do not contain
2265            any matching lines, that is, the complement of -l.
2266    
2267        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2268            continue to scan other files.
2269    
2270        (o) The -s option was incorrectly implemented. For compatibility with other
2271            greps, it now suppresses the error message for a non-existent or non-
2272            accessible file (but not the return code). There is a new option called
2273            -q that suppresses the output of matching lines, which was what -s was
2274            previously doing.
2275    
2276        (p) Added --include and --exclude options to specify files for inclusion
2277            and exclusion when recursing.
2278    
2279    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2280        Hopefully, it now does.
2281    
2282    12. Missing cast in pcre_study().
2283    
2284    13. Added an "uninstall" target to the makefile.
2285    
2286    14. Replaced "extern" in the function prototypes in Makefile.in with
2287        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2288        world, but is set differently for Windows.
2289    
2290    15. Added a second compiling function called pcre_compile2(). The only
2291        difference is that it has an extra argument, which is a pointer to an
2292        integer error code. When there is a compile-time failure, this is set
2293        non-zero, in addition to the error test pointer being set to point to an
2294        error message. The new argument may be NULL if no error number is required
2295        (but then you may as well call pcre_compile(), which is now just a
2296        wrapper). This facility is provided because some applications need a
2297        numeric error indication, but it has also enabled me to tidy up the way
2298        compile-time errors are handled in the POSIX wrapper.
2299    
2300    16. Added VPATH=.libs to the makefile; this should help when building with one
2301        prefix path and installing with another. (Or so I'm told by someone who
2302        knows more about this stuff than I do.)
2303    
2304    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2305        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2306        match everything, including newlines. This is not POSIX-compatible, but
2307        somebody wanted the feature. From pcretest it can be activated by using
2308        both the P and the s flags.
2309    
2310    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2311    
2312    19. libpcre.pc was being incorrectly installed as executable.
2313    
2314    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2315        it now also looks for '\r' so that it will work unmodified on Windows.
2316    
2317    21. Added Google's contributed C++ wrapper to the distribution.
2318    
2319    22. Added some untidy missing memory free() calls in pcretest, to keep
2320        Electric Fence happy when testing.
2321    
2322    
2323    
2324    Version 5.0 13-Sep-04
2325    ---------------------
2326    
2327     1. Internal change: literal characters are no longer packed up into items
2328        containing multiple characters in a single byte-string. Each character
2329        is now matched using a separate opcode. However, there may be more than one
2330        byte in the character in UTF-8 mode.
2331    
2332     2. The pcre_callout_block structure has two new fields: pattern_position and
2333        next_item_length. These contain the offset in the pattern to the next match
2334        item, and its length, respectively.
2335    
2336     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2337        insertion of callouts before each pattern item. Added the /C option to
2338        pcretest to make use of this.
2339    
2340     4. On the advice of a Windows user, the lines
2341    
2342          #if defined(_WIN32) || defined(WIN32)
2343          _setmode( _fileno( stdout ), 0x8000 );
2344          #endif  /* defined(_WIN32) || defined(WIN32) */
2345    
2346        have been added to the source of pcretest. This apparently does useful
2347        magic in relation to line terminators.
2348    
2349     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2350        for the benefit of those environments where the "b" makes a difference.
2351    
2352     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2353        to know about it. I have put a hack into configure.in that adds in code
2354        to set GCC=yes if CC=icc. This seems to end up at a point in the
2355        generated configure script that is early enough to affect the setting of
2356        compiler options, which is what is needed, but I have no means of testing
2357        whether it really works. (The user who reported this had patched the
2358        generated configure script, which of course I cannot do.)
2359    
2360        LATER: After change 22 below (new libtool files), the configure script
2361        seems to know about icc (and also ecc). Therefore, I have commented out
2362        this hack in configure.in.
2363    
2364     7. Added support for pkg-config (2 patches were sent in).
2365    
2366     8. Negated POSIX character classes that used a combination of internal tables
2367        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2368        [[:^ascii]]. Typically, they would match almost any characters. The other
2369        POSIX classes were not broken in this way.
2370    
2371     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2372        to find the match, as PCRE was deluded into thinking that the match had to
2373        start at the start point or following a newline. The same bug applied to
2374        patterns with negative forward assertions or any backward assertions
2375        preceding ".*" at the start, unless the pattern required a fixed first
2376        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2377    
2378    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2379        starting at the last subject character, bytes beyond the end of the subject
2380        string were read.
2381    
2382    11. Renamed the variable "class" as "classbits" to make life easier for C++
2383        users. (Previously there was a macro definition, but it apparently wasn't
2384        enough.)
2385    
2386    12. Added the new field "tables" to the extra data so that tables can be passed
2387        in at exec time, or the internal tables can be re-selected. This allows
2388        a compiled regex to be saved and re-used at a later time by a different
2389        program that might have everything at different addresses.
2390    
2391    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2392        -R library as well as a -L library.
2393    
2394    14. The debugging options of pcretest (-d on the command line or D on a
2395        pattern) showed incorrect output for anything following an extended class
2396        that contained multibyte characters and which was followed by a quantifier.
2397    
2398    15. Added optional support for general category Unicode character properties
2399        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2400        support. It adds about 90K to the size of the library. The meanings of the
2401        inbuilt class escapes such as \d and \s have NOT been changed.
2402    
2403    16. Updated pcredemo.c to include calls to free() to release the memory for the
2404        compiled pattern.
2405    
2406    17. The generated file chartables.c was being created in the source directory
2407        instead of in the building directory. This caused the build to fail if the
2408        source directory was different from the building directory, and was
2409        read-only.
2410    
2411    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2412        file. No doubt somebody will tell me if they don't make sense... Also added
2413        Dan Mooney's comments about building on OpenVMS.
2414    
2415    19. Added support for partial matching via the PCRE_PARTIAL option for
2416        pcre_exec() and the \P data escape in pcretest.
2417    
2418    20. Extended pcretest with 3 new pattern features:
2419    
2420        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2421              write the compiled pattern to the file whose name is "rest-of-line".
2422              This is a straight binary dump of the data, with the saved pointer to
2423              the character tables forced to be NULL. The study data, if any, is
2424              written too. After writing, pcretest reads a new pattern.
2425    
2426        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2427              compiled pattern from the given file. There must not be any
2428              occurrences of "<" in the file name (pretty unlikely); if there are,
2429              pcretest will instead treat the initial "<" as a pattern delimiter.
2430              After reading in the pattern, pcretest goes on to read data lines as
2431              usual.
2432    
2433        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2434              and 16-bit fields in a compiled pattern, to simulate a pattern that
2435              was compiled on a host of opposite endianness.
2436    
2437    21. The pcre-exec() function can now cope with patterns that were compiled on
2438        hosts of opposite endianness, with this restriction:
2439    
2440          As for any compiled expression that is saved and used later, the tables
2441          pointer field cannot be preserved; the extra_data field in the arguments
2442          to pcre_exec() should be used to pass in a tables address if a value
2443          other than the default internal tables were used at compile time.
2444    
2445    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2446        now diagnosed as an error. Previously, most of the time, a negative number
2447        would have been treated as zero, but if in addition "ovector" was passed as
2448        NULL, a crash could occur.
2449    
2450    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2451        new versions from the libtool 1.5 distribution (the last one is a copy of
2452        a file called libtool.m4). This seems to have fixed the need to patch
2453        "configure" to support Darwin 1.3 (which I used to do). However, I still
2454        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2455        workstation).
2456    
2457    24. Changed the PCRE licence to be the more standard "BSD" licence.
2458    
2459    
2460    Version 4.5 01-Dec-03
2461    ---------------------
2462    
2463     1. There has been some re-arrangement of the code for the match() function so
2464        that it can be compiled in a version that does not call itself recursively.
2465        Instead, it keeps those local variables that need separate instances for
2466        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2467        needs to "recurse". Keeping track of where control must go is done by means
2468        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2469        hide most of the details from the main code, and operates only if
2470        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2471        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2472        operating.
2473    
2474        To make it easier for callers to provide specially tailored get/free
2475        functions for this usage, two new functions, pcre_stack_malloc, and
2476        pcre_stack_free, are used. They are always called in strict stacking order,
2477        and the size of block requested is always the same.
2478    
2479        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2480        PCRE has been compiled to use the stack or the heap for recursion. The
2481        -C option of pcretest uses this to show which version is compiled.
2482    
2483        A new data escape \S, is added to pcretest; it causes the amounts of store
2484        obtained and freed by both kinds of malloc/free at match time to be added
2485        to the output.
2486    
2487     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2488        what's available on my current Linux desktop machine.
2489    
2490     3. When matching a UTF-8 string, the test for a valid string at the start has
2491        been extended. If start_offset is not zero, PCRE now checks that it points
2492        to a byte that is the start of a UTF-8 character. If not, it returns
2493        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2494        this is necessary because there may be backward assertions in the pattern.
2495        When matching the same subject several times, it may save resources to use
2496        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2497    
2498     4. The code for checking the validity of UTF-8 strings has been tightened so
2499        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2500        containing "overlong sequences".
2501    
2502     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2503        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2504        should have been "&", but it just so happened that all the cases this let
2505        through by mistake were picked up later in the function.
2506    
2507     6. I had used a variable called "isblank" - this is a C99 function, causing
2508        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2509    
2510     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2511        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2512        so the version is shown; (c) stop "make test" echoing "./RunTest".
2513    
2514     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2515    
2516     9. The prototype for memmove() for systems that don't have it was using
2517        size_t, but the inclusion of the header that defines size_t was later. I've
2518        moved the #includes for the C headers earlier to avoid this.
2519    
2520    10. Added some adjustments to the code to make it easier to compiler on certain
2521        special systems:
2522    
2523          (a) Some "const" qualifiers were missing.
2524          (b) Added the macro EXPORT before all exported functions; by default this
2525              is defined to be empty.
2526          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2527              that it reads its output file name as an argument instead of writing
2528              to the standard output and assuming this can be redirected.
2529    
2530    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2531        class containing characters with values greater than 255, PCRE compilation
2532        went into a loop.
2533    
2534    12. A recursive reference to a subpattern that was within another subpattern
2535        that had a minimum quantifier of zero caused PCRE to crash. For example,
2536        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2537        recursion. If the recursively-called subpattern itself had a zero repeat,
2538        that was OK.
2539    
2540    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2541        buffer into which it was copied (for escape processing) was still set at
2542        1024, so long lines caused crashes.
2543    
2544    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2545        "internal error: code overflow...". This applied to any character class
2546        that was followed by a possessive quantifier.
2547    
2548    15. Modified the Makefile to add libpcre.la as a prerequisite for
2549        libpcreposix.la because I was told this is needed for a parallel build to
2550        work.
2551    
2552    16. If a pattern that contained .* following optional items at the start was
2553        studied, the wrong optimizing data was generated, leading to matching
2554        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2555        matching string must start with a or b or c. The correct conclusion for
2556        this pattern is that a match can start with any character.
2557    
2558    
2559    Version 4.4 13-Aug-03
2560    ---------------------
2561    
2562     1. In UTF-8 mode, a character class containing characters with values between
2563        127 and 255 was not handled correctly if the compiled pattern was studied.
2564        In fixing this, I have also improved the studying algorithm for such
2565        classes (slightly).
2566    
2567     2. Three internal functions had redundant arguments passed to them. Removal
2568        might give a very teeny performance improvement.
2569    
2570     3. Documentation bug: the value of the capture_top field in a callout is *one
2571        more than* the number of the hightest numbered captured substring.
2572    
2573     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2574        in incorrectly linking with a previously installed version. They now link
2575        explicitly with libpcre.la.
2576    
2577     5. configure.in no longer needs to recognize Cygwin specially.
2578    
2579     6. A problem in pcre.in for Windows platforms is fixed.
2580    
2581     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2582        pcretest, it used to include the size of the study block as part of its
2583        output. Unfortunately, the structure contains a field that has a different
2584        size on different hardware architectures. This meant that the tests that
2585        showed this size failed. As the block is currently always of a fixed size,
2586        this information isn't actually particularly useful in pcretest output, so
2587        I have just removed it.
2588    
2589     8. Three pre-processor statements accidentally did not start in column 1.
2590        Sadly, there are *still* compilers around that complain, even though
2591        standard C has not required this for well over a decade. Sigh.
2592    
2593     9. In pcretest, the code for checking callouts passed small integers in the
2594        callout_data field, which is a void * field. However, some picky compilers
2595        complained about the casts involved for this on 64-bit systems. Now
2596        pcretest passes the address of the small integer instead, which should get
2597        rid of the warnings.
2598    
2599    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2600        both compile and run time, and gives an error if an invalid UTF-8 sequence
2601        is found. There is a option for disabling this check in cases where the
2602        string is known to be correct and/or the maximum performance is wanted.
2603    
2604    11. In response to a bug report, I changed one line in Makefile.in from
2605    
2606            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2607        to
2608            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2609    
2610        to look similar to other lines, but I have no way of telling whether this
2611        is the right thing to do, as I do not use Windows. No doubt I'll get told
2612        if it's wrong...
2613    
2614    
2615    Version 4.3 21-May-03
2616    ---------------------
2617    
2618    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2619       Makefile.
2620    
2621    2. Some refactoring to improve the quality of the code:
2622    
2623       (i)   The utf8_table... variables are now declared "const".
2624    
2625       (ii)  The code for \cx, which used the "case flipping" table to upper case
2626             lower case letters, now just substracts 32. This is ASCII-specific,
2627             but the whole concept of \cx is ASCII-specific, so it seems
2628             reasonable.
2629    
2630       (iii) PCRE was using its character types table to recognize decimal and
2631             hexadecimal digits in the pattern. This is silly, because it handles
2632             only 0-9, a-f, and A-F, but the character types table is locale-
2633             specific, which means strange things might happen. A private
2634             table is now used for this - though it costs 256 bytes, a table is
2635             much faster than multiple explicit tests. Of course, the standard
2636             character types table is still used for matching digits in subject
2637             strings against \d.
2638    
2639       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2640             ending in _t are). So I've renamed it as ESC_tee.
2641    
2642    3. The first argument for regexec() in the POSIX wrapper should have been
2643       defined as "const".
2644    
2645    4. Changed pcretest to use malloc() for its buffers so that they can be
2646       Electric Fenced for debugging.
2647    
2648    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2649       to read one or more bytes before the start of the subject string. Often this
2650       had no effect on PCRE's behaviour, but in some circumstances it could
2651       provoke a segmentation fault.
2652    
2653    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2654       to try to read one or more bytes before the start of the subject string.
2655    
2656    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2657       UTF-8 support could misbehave in various ways if the subject string
2658       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2659       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2660       back over UTF-8 characters.)
2661    
2662    
2663    Version 4.2 14-Apr-03
2664    ---------------------
2665    
2666    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2667    
2668    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2669         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2670         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2671         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2672         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2673           and BUILD_EXEEXT
2674         Note: automatic setting of the BUILD variables is not yet working
2675         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2676           compile-time but not at link-time
2677         [LINK]: use for linking executables only
2678         make different versions for Windows and non-Windows
2679         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2680           libraries
2681         [LINK_FOR_BUILD]: new variable
2682         [OBJEXT]: use throughout
2683         [EXEEXT]: use throughout
2684         <winshared>: new target
2685         <wininstall>: new target
2686         <dftables.o>: use native compiler
2687         <dftables>: use native linker
2688         <install>: handle Windows platform correctly
2689         <clean>: ditto
2690         <check>: ditto
2691         copy DLL to top builddir before testing
2692    
2693       As part of these changes, -no-undefined was removed again. This was reported
2694       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2695       in any case.
2696    
2697    3. Some tidies to get rid of compiler warnings:
2698    
2699       . In the match_data structure, match_limit was an unsigned long int, whereas
2700         match_call_count was an int. I've made them both unsigned long ints.
2701    
2702       . In pcretest the fact that a const uschar * doesn't automatically cast to
2703         a void * provoked a warning.
2704    
2705       . Turning on some more compiler warnings threw up some "shadow" variables
2706         and a few more missing casts.
2707    
2708    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2709       option, a class that contained a single character with a value between 128
2710       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2711    
2712    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2713       option, a class that contained several characters, but with at least one
2714       whose value was between 128 and 255 caused PCRE to crash.
2715    
2716    
2717    Version 4.1 12-Mar-03
2718    ---------------------
2719    
2720    1. Compiling with gcc -pedantic found a couple of places where casts were
2721    needed, and a string in dftables.c that was longer than standard compilers are
2722    required to support.
2723    
2724    2. Compiling with Sun's compiler found a few more places where the code could
2725    be tidied up in order to avoid warnings.
2726    
2727    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2728    first of these names is deprecated in the latest Autoconf in favour of the name
2729    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2730    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2731    analogy I have changed it to CFLAGS_FOR_BUILD.
2732    
2733    4. Added -no-undefined to the linking command in the Makefile, because this is
2734    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2735    linking step for the pcreposix library.
2736    
2737    5. PCRE was failing to diagnose the case of two named groups with the same
2738    name.
2739    
2740    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2741    literal character that is needed in the subject for a match, and scans along to
2742    ensure that it is present before embarking on the full matching process. This
2743    saves time in cases of nested unlimited repeats that are never going to match.
2744    Problem: the scan can take a lot of time if the subject is very long (e.g.
2745    megabytes), thus penalizing straightforward matches. It is now done only if the
2746    amount of subject to be scanned is less than 1000 bytes.
2747    
2748    7. A lesser problem with the same optimization is that it was recording the
2749    first character of an anchored pattern as "needed", thus provoking a search
2750    right along the subject, even when the first match of the pattern was going to
2751    fail. The "needed" character is now not set for anchored patterns, unless it
2752    follows something in the pattern that is of non-fixed length. Thus, it still
2753    fulfils its original purpose of finding quick non-matches in cases of nested
2754    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2755    
2756    
2757    Version 4.0 17-Feb-03
2758    ---------------------
2759    
2760    1. If a comment in an extended regex that started immediately after a meta-item
2761    extended to the end of string, PCRE compiled incorrect data. This could lead to
2762    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2763    
2764    2. Moved to autoconf 2.53 and libtool 1.4.2.
2765    
2766    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2767    the special perltest8 script is no longer needed - all the tests can be run
2768    from a single perltest script.
2769    
2770    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2771    by \s. It has now been removed in PCRE. This means it isn't recognized as
2772    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2773    class [:space:] *does* include VT, thereby creating a mess.
2774    
2775    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2776    space and tab.
2777    
2778    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2779    its new features into the main test script, reducing the number of scripts.
2780    
2781    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2782    were backward compatible, and made the (?i) apply to the whole pattern, as if
2783    /i were given. Now it behaves more logically, and applies the option setting
2784    only to what follows. PCRE has been changed to follow suit. However, if it
2785    finds options settings right at the start of the pattern, it extracts them into
2786    the global options, as before. Thus, they show up in the info data.
2787    
2788    8. Added support for the \Q...\E escape sequence. Characters in between are
2789    treated as literals. This is slightly different from Perl in that $ and @ are
2790    also handled as literals inside the quotes. In Perl, they will cause variable
2791    interpolation. Note the following examples:
2792    
2793        Pattern            PCRE matches      Perl matches
2794    
2795        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2796        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2797        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2798    
2799    For compatibility with Perl, \Q...\E sequences are recognized inside character
2800    classes as well as outside them.
2801    
2802    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2803    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2804    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2805    signed/unsigned warnings.
2806    
2807    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2808    option for pcretest, so I've replaced it by a simple function that does just
2809    that job.
2810    
2811    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2812    "pcregrep -".
2813    
2814    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2815    Java package. This provides some syntactic sugar for simple cases of what my
2816    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2817    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2818    item, you can use this simplified notation. Note that only makes sense with
2819    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2820    greediness, whatever the setting of the PCRE_UNGREEDY option.
2821    
2822    13. A change of greediness default within a pattern was not taking effect at
2823    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2824    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2825    was abstracted outside.
2826    
2827    14. PCRE now supports the \G assertion. It is true when the current matching
2828    position is at the start point of the match. This differs from \A when the
2829    starting offset is non-zero. Used with the /g option of pcretest (or similar
2830    code), it works in the same way as it does for Perl's /g option. If all
2831    alternatives of a regex begin with \G, the expression is anchored to the start
2832    match position, and the "anchored" flag is set in the compiled expression.
2833    
2834    15. Some bugs concerning the handling of certain option changes within patterns
2835    have been fixed. These applied to options other than (?ims). For example,
2836    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2837    been the other way round. Some of this was related to change 7 above.
2838    
2839    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2840    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2841    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2842    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2843    
2844    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2845    mode. Unlike ".", it always matches newline, whatever the setting of
2846    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2847    assertions. Perl allows it, but it doesn't (in general) work because it can't
2848    calculate the length of the lookbehind. At least, that's the case for Perl
2849    5.8.0 - I've been told they are going to document that it doesn't work in
2850    future.
2851    
2852    18. Added an error diagnosis for escapes that PCRE does not support: these are
2853    \L, \l, \N, \P, \p, \U, \u, and \X.
2854    
2855    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2856    reading past the end of the pattern in cases such as /[abcd/.
2857    
2858    20. PCRE was getting more memory than necessary for patterns with classes that
2859    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2860    
2861    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2862    compiling PCRE for use with Virtual Pascal.
2863    
2864    22. Small fix to the Makefile to make it work properly if the build is done
2865    outside the source tree.
2866    
2867    23. Added a new extension: a condition to go with recursion. If a conditional
2868    subpattern starts with (?(R) the "true" branch is used if recursion has
2869    happened, whereas the "false" branch is used only at the top level.
2870    
2871    24. When there was a very long string of literal characters (over 255 bytes
2872    without UTF support, over 250 bytes with UTF support), the computation of how
2873    much memory was required could be incorrect, leading to segfaults or other
2874    strange effects.
2875    
2876    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2877    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2878    there was a subsequent back reference to those brackets. This meant that, for
2879    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2880    possible to check for precisely this case. All we can do is abandon the
2881    optimization if .* occurs inside capturing brackets when there are any back
2882    references whatsoever. (See below for a better fix that came later.)
2883    
2884    26. The handling of the optimization for finding the first character of a
2885    non-anchored pattern, and for finding a character that is required later in the
2886    match were failing in some cases. This didn't break the matching; it just
2887    failed to optimize when it could. The way this is done has been re-implemented.
2888    
2889    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2890    
2891    28. Added a new feature that provides some of the functionality that Perl
2892    provides with (?{...}). The facility is termed a "callout". The way it is done
2893    in PCRE is for the caller to provide an optional function, by setting
2894    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2895    global variable. By default it is unset, which disables all calling out. To get
2896    the function called, the regex must include (?C) at appropriate points. This
2897    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2898    This provides a means of identifying different callout points. When PCRE
2899    reaches such a point in the regex, if pcre_callout has been set, the external
2900    function is called. It is provided with data in a structure called
2901    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2902    matching continues; if it returns a non-zero value, the match at the current
2903    point fails. However, backtracking will occur if possible. [This was changed
2904    later and other features added - see item 49 below.]
2905    
2906    29. pcretest is upgraded to test the callout functionality. It provides a
2907    callout function that displays information. By default, it shows the start of
2908    the match and the current position in the text. There are some new data escapes
2909    to vary what happens:
2910    
2911        \C+         in addition, show current contents of captured substrings
2912        \C-         do not supply a callout function
2913        \C!n        return 1 when callout number n is reached
2914        \C!n!m      return 1 when callout number n is reached for the mth time
2915    
2916    30. If pcregrep was called with the -l option and just a single file name, it
2917    output "<stdin>" if a match was found, instead of the file name.
2918    
2919    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2920    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2921    pcre_exec(). This saves a malloc/free per call. The default value of
2922    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2923    when configuring.
2924    
2925    32. The default maximum size of a compiled pattern is 64K. There have been a
2926    few cases of people hitting this limit. The code now uses macros to handle the
2927    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2928    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2929    configuring. Tests 2 and 5 work only with 2-byte links because they output
2930    debugging information about compiled patterns.
2931    
2932    33. Internal code re-arrangements:
2933    
2934    (a) Moved the debugging function for printing out a compiled regex into
2935        its own source file (printint.c) and used #include to pull it into
2936        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2937        separate copies.
2938    
2939    (b) Defined the list of op-code names for debugging as a macro in
2940        internal.h so that it is next to the definition of the opcodes.
2941    
2942    (c) Defined a table of op-code lengths for simpler skipping along compiled
2943        code. This is again a macro in internal.h so that it is next to the
2944        definition of the opcodes.
2945    
2946    34. Added support for recursive calls to individual subpatterns, along the
2947    lines of Robin Houston's patch (but implemented somewhat differently).
2948    
2949    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2950    allow it to read and process whole directories in Win32. This code was
2951    contributed by Lionel Fourquaux; it has not been tested by me.
2952    
2953    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2954    used to name a group. Names consist of alphanumerics and underscores, and must
2955    be unique. Back references use the syntax (?P=name) and recursive calls use
2956    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2957    numbers. The function pcre_fullinfo() can be used after compilation to extract
2958    a name/number map. There are three relevant calls:
2959    
2960      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2961      PCRE_INFO_NAMECOUNT            yields the number of entries
2962      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2963    
2964    The map is a vector of fixed-size entries. The size of each entry depends on
2965    the length of the longest name used. The first two bytes of each entry are the
2966    group number, most significant byte first. There follows the corresponding
2967    name, zero terminated. The names are in alphabetical order.
2968    
2969    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2970    case instead of 255. Making it the same both with and without UTF-8 support
2971    means that the same test output works with both.
2972    
2973    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2974    calling malloc() with a zero argument.
2975    
2976    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2977    optimization. I've improved things by keeping a bitmap of backreferences with
2978    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2979    fact referenced, the optimization can be applied. It is unlikely that a
2980    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2981    the match to follow \n) will appear inside brackets with a number greater than
2982    31, but if it does, any back reference > 31 suppresses the optimization.
2983    
2984    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2985    of disabling numbered capturing parentheses. Any opening parenthesis that is
2986    not followed by ? behaves as if it were followed by ?: but named parentheses
2987    can still be used for capturing (and they will acquire numbers in the usual
2988    way).
2989    
2990    41. Redesigned the return codes from the match() function into yes/no/error so
2991    that errors can be passed back from deep inside the nested calls. A malloc
2992    failure while inside a recursive subpattern call now causes the
2993    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2994    
2995    42. It is now possible to set a limit on the number of times the match()
2996    function is called in a call to pcre_exec(). This facility makes it possible to
2997    limit the amount of recursion and backtracking, though not in a directly
2998    obvious way, because the match() function is used in a number of different
2999    circumstances. The count starts from zero for each position in the subject
3000    string (for non-anchored patterns). The default limit is, for compatibility, a
3001    large number, namely 10 000 000. You can change this in two ways:
3002    
3003    (a) When configuring PCRE before making, you can use --with-match-limit=n
3004        to set a default value for the compiled library.
3005    
3006    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3007        a different value is set. See 45 below.
3008    
3009    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3010    
3011    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3012    of things that can be changed at compile time. The first argument specifies
3013    what is wanted and the second points to where the information is to be placed.
3014    The current list of available information is:
3015    
3016      PCRE_CONFIG_UTF8
3017    
3018    The output is an integer that is set to one if UTF-8 support is available;
3019    otherwise it is set to zero.
3020    
3021      PCRE_CONFIG_NEWLINE
3022    
3023    The output is an integer that it set to the value of the code that is used for
3024    newline. It is either LF (10) or CR (13).
3025    
3026      PCRE_CONFIG_LINK_SIZE
3027    
3028    The output is an integer that contains the number of bytes used for internal
3029    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3030    
3031      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3032    
3033    The output is an integer that contains the threshold above which the POSIX
3034    interface uses malloc() for output vectors. See item 31 above.
3035    
3036      PCRE_CONFIG_MATCH_LIMIT
3037    
3038    The output is an unsigned integer that contains the default limit of the number
3039    of match() calls in a pcre_exec() execution. See 42 above.
3040    
3041    44. pcretest has been upgraded by the addition of the -C option. This causes it
3042    to extract all the available output from the new pcre_config() function, and to
3043    output it. The program then exits immediately.
3044    
3045    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3046    order to support additional features. One way would have been to define
3047    pcre_exec2() (for example) with extra arguments, but this would not have been
3048    extensible, and would also have required all calls to the original function to
3049    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3050    is used for passing in "extra" data from pcre_study().
3051    
3052    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3053    contains the following fields:
3054    
3055      flags         a bitmap indicating which of the following fields are set
3056      study_data    opaque data from pcre_study()
3057      match_limit   a way of specifying a limit on match() calls for a specific
3058                      call to pcre_exec()
3059      callout_data  data for callouts (see 49 below)
3060    
3061    The flag bits are also defined in pcre.h, and are
3062    
3063      PCRE_EXTRA_STUDY_DATA
3064      PCRE_EXTRA_MATCH_LIMIT
3065      PCRE_EXTRA_CALLOUT_DATA
3066    
3067    The pcre_study() function now returns one of these new pcre_extra blocks, with
3068    the actual study data pointed to by the study_data field, and the
3069    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3070    before. That is, this change is entirely upwards-compatible and requires no
3071    change to existing code.
3072    
3073    If you want to pass in additional data to pcre_exec(), you can either place it
3074    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3075    block.
3076    
3077    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3078    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3079    times with different match limits, until it finds the minimum value needed for
3080    pcre_exec() to complete. The value is then output. This can be instructive; for
3081    most simple matches the number is quite small, but for pathological cases it
3082    gets very large very quickly.
3083    
3084    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3085    returns the size of the data block pointed to by the study_data field in a
3086    pcre_extra block, that is, the value that was passed as the argument to
3087    pcre_malloc() when PCRE was getting memory in which to place the information
3088    created by pcre_study(). The fourth argument should point to a size_t variable.
3089    pcretest has been extended so that this information is shown after a successful
3090    pcre_study() call when information about the compiled regex is being displayed.
3091    
3092    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3093    because what follows is always an absolute path. (Later: it turns out that this
3094    is more than cosmetic for MinGW, because it doesn't like empty path
3095    components.)
3096    
3097    49. Some changes have been made to the callout feature (see 28 above):
3098    
3099    (i)  A callout function now has three choices for what it returns:
3100    
3101           0  =>  success, carry on matching
3102         > 0  =>  failure at this point, but backtrack if possible
3103         < 0  =>  serious error, return this value from pcre_exec()
3104    
3105         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3106         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3107         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3108         use by callout functions. It will never be used by PCRE itself.
3109    
3110    (ii) The pcre_extra structure (see 45 above) has a void * field called
3111         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3112         pcre_callout_block structure has a field of the same name. The contents of
3113         the field passed in the pcre_extra structure are passed to the callout
3114         function in the corresponding field in the callout block. This makes it
3115         easier to use the same callout-containing regex from multiple threads. For
3116         testing, the pcretest program has a new data escape
3117    
3118           \C*n        pass the number n (may be negative) as callout_data
3119    
3120         If the callout function in pcretest receives a non-zero value as
3121         callout_data, it returns that value.
3122    
3123    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3124    there were some redundant $(CFLAGS) in commands that are now specified as
3125    $(LINK), which already includes $(CFLAGS).
3126    
3127    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3128    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3129    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3130    one-byte characters throughout. Note that case-insensitive matching applies
3131    only to characters whose values are less than 256. PCRE doesn't support the
3132    notion of cases for higher-valued characters.
3133    
3134    (i)   A character class whose characters are all within 0-255 is handled as
3135          a bit map, and the map is inverted for negative classes. Previously, a
3136          character > 255 always failed to match such a class; however it should
3137          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3138    
3139    (ii)  A negated character class with a single character < 255 is coded as
3140          "not this character" (OP_NOT). This wasn't working properly when the test
3141          character was multibyte, either singly or repeated.
3142    
3143    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3144          mode, for example: \x{100}{2,3}.
3145    
3146    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3147          singly or repeated) now correctly test multibyte characters. However,
3148          PCRE doesn't recognize any characters with values greater than 255 as
3149          digits, spaces, or word characters. Such characters always match \D, \S,
3150          and \W, and never match \d, \s, or \w.
3151    
3152    (v)   Classes may now contain characters and character ranges with values
3153          greater than 255. For example: [ab\x{100}-\x{400}].
3154    
3155    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3156          PCRE in UTF-8 mode.
3157    
3158    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3159    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3160    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3161    value.)
3162    
3163    53. The single man page has become too large. I have therefore split it up into
3164    a number of separate man pages. These also give rise to individual HTML pages;
3165    these are now put in a separate directory, and there is an index.html page that
3166    lists them all. Some hyperlinking between the pages has been installed.
3167    
3168    54. Added convenience functions for handling named capturing parentheses.
3169    
3170    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3171    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3172    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3173    are faulted.
3174    
3175    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3176    calling configure. These values are used when compiling the dftables.c program
3177    which is run to generate the source of the default character tables. They
3178    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3179    you will need to set these values.
3180    
3181    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3182    
3183    
3184    Version 3.9 02-Jan-02
3185    ---------------------
3186    
3187    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3188    
3189    2. If --disable-static was given, the building process failed when trying to
3190    build pcretest and pcregrep. (For some reason it was using libtool to compile
3191    them, which is not right, as they aren't part of the library.)
3192    
3193    
3194    Version 3.8 18-Dec-01
3195    ---------------------
3196    
3197    1. The experimental UTF-8 code was completely screwed up. It was packing the
3198    bytes in the wrong order. How dumb can you get?
3199    
3200    
3201    Version 3.7 29-Oct-01
3202    ---------------------
3203    
3204    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3205    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3206    this didn't happen under Solaris 8, where I normally test things.
3207    
3208    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3209    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3210    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3211    this fix an hour or so after the initial 3.7 release.)
3212    
3213    
3214    Version 3.6 23-Oct-01
3215    ---------------------
3216    
3217    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3218    offsets passed as NULL with zero offset count.
3219    
3220    2. The config.guess and config.sub files had not been updated when I moved to
3221    the latest autoconf.
3222    
3223    
3224    Version 3.5 15-Aug-01
3225    ---------------------
3226    
3227    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3228    had been forgotten.
3229    
3230    2. By using declared but undefined structures, we can avoid using "void"
3231    definitions in pcre.h while keeping the internal definitions of the structures
3232    private.
3233    
3234    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3235    user point of view, this means that both static and shared libraries are built
3236    by default, but this can be individually controlled. More of the work of
3237    handling this static/shared cases is now inside libtool instead of PCRE's make
3238    file.
3239    
3240    4. The pcretest utility is now installed along with pcregrep because it is
3241    useful for users (to test regexs) and by doing this, it automatically gets
3242    relinked by libtool. The documentation has been turned into a man page, so
3243    there are now .1, .txt, and .html versions in /doc.
3244    
3245    5. Upgrades to pcregrep:
3246       (i)   Added long-form option names like gnu grep.
3247       (ii)  Added --help to list all options with an explanatory phrase.
3248       (iii) Added -r, --recursive to recurse into sub-directories.
3249       (iv)  Added -f, --file to read patterns from a file.
3250    
3251    6. pcre_exec() was referring to its "code" argument before testing that
3252    argument for NULL (and giving an error if it was NULL).
3253    
3254    7. Upgraded Makefile.in to allow for compiling in a different directory from
3255    the source directory.
3256    
3257    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3258    options bits, the pointer it was passed was to an int instead of to an unsigned
3259    long int. This mattered only on 64-bit systems.
3260    
3261    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3262    generated) instead of pcre.in, which it its source. Also made the same change
3263    in several of the .c files.
3264    
3265    10. A new release of gcc defines printf() as a macro, which broke pcretest
3266    because it had an ifdef in the middle of a string argument for printf(). Fixed
3267    by using separate calls to printf().
3268    
3269    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3270    script, to force use of CR or LF instead of \n in the source. On non-Unix
3271    systems, the value can be set in config.h.
3272    
3273    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3274    absolute limit. Changed the text of the error message to make this clear, and
3275    likewise updated the man page.
3276    
3277    13. The limit of 99 on the number of capturing subpatterns has been removed.
3278    The new limit is 65535, which I hope will not be a "real" limit.
3279    
3280    
3281    Version 3.4 22-Aug-00
3282    ---------------------
3283    
3284    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3285    
3286    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3287    
3288    
3289    Version 3.3 01-Aug-00
3290    ---------------------
3291    
3292    1. If an octal character was given, but the value was greater than \377, it
3293    was not getting masked to the least significant bits, as documented. This could
3294    lead to crashes in some systems.
3295    
3296    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3297    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3298    
3299    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3300    These just pass their arguments on to (pcre_free)(), but they are provided
3301    because some uses of PCRE bind it to non-C systems that can call its functions,
3302    but cannot call free() or pcre_free() directly.
3303    
3304    4. Add "make test" as a synonym for "make check". Corrected some comments in
3305    the Makefile.
3306    
3307    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3308    Makefile.
3309    
3310    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3311    command called pgrep for grepping around the active processes.
3312    
3313    7. Added the beginnings of support for UTF-8 character strings.
3314    
3315    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3316    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3317    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3318    out for the ar command.)
3319    
3320    
3321    Version 3.2 12-May-00
3322    ---------------------
3323    
3324    This is purely a bug fixing release.
3325    
3326    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3327    of ZA. This was just one example of several cases that could provoke this bug,
3328    which was introduced by change 9 of version 2.00. The code for breaking
3329    infinite loops after an iteration that matches an empty string was't working
3330    correctly.
3331    
3332    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3333    when matched against abbab (for example). After matching an empty string, it
3334    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3335    caused it to match further down the string than it should.
3336    
3337    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3338    was there because it doesn't seem to be needed, and it causes trouble on some
3339    systems, as it is not a Standard C header. It has been removed.
3340    
3341    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3342    were reported on the Macintosh. The changes were from
3343    
3344      while ((c = *(++ptr)) != 0 && c != '\n');
3345    to
3346      while ((c = *(++ptr)) != 0 && c != '\n') ;
3347    
3348    Totally extraordinary, but if that's what it takes...
3349    
3350    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3351    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3352    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3353    assumes the way PCRE uses memmove() (always moving upwards).
3354    
3355    6. PCRE is being used in one environment where strchr() is not available. There
3356    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3357    faster code anyway.
3358    
3359    
3360    Version 3.1 09-Feb-00
3361    ---------------------
3362    
3363    The only change in this release is the fixing of some bugs in Makefile.in for
3364    the "install" target:
3365    
3366    (1) It was failing to install pcreposix.h.
3367    
3368    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3369    
3370    
3371    Version 3.0 01-Feb-00
3372    ---------------------
3373    
3374    1. Add support for the /+ modifier to perltest (to output $` like it does in
3375    pcretest).
3376    
3377    2. Add support for the /g modifier to perltest.
3378    
3379    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3380    matches null strings.
3381    
3382    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3383    pattern. Perl treats empty patterns specially - it reuses the most recent
3384    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3385    effect.
3386    
3387    5. The POSIX interface was broken in that it was just handing over the POSIX
3388    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3389    required a bigger vector, with some working space on the end. This means that
3390    the POSIX wrapper now has to get and free some memory, and copy the results.
3391    
3392    6. Added some simple autoconf support, placing the test data and the
3393    documentation in separate directories, re-organizing some of the
3394    information files, and making it build pcre-config (a GNU standard). Also added
3395    libtool support for building PCRE as a shared library, which is now the
3396    default.
3397    
3398    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3399    09 are not valid octal constants. Single digits will be used for minor values
3400    less than 10.
3401    
3402    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3403    existing programs that set these in the POSIX interface can use PCRE without
3404    modification.
3405    
3406    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3407    return all that pcre_info() returns, plus additional data. The pcre_info()
3408    function is retained for compatibility, but is considered to be obsolete.
3409    
3410    10. Added experimental recursion feature (?R) to handle one common case that
3411    Perl 5.6 will be able to do with (?p{...}).
3412    
3413    11. Added support for POSIX character classes like [:alpha:], which Perl is
3414    adopting.
3415    
3416    
3417    Version 2.08 31-Aug-99
3418    ----------------------
3419    
3420    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3421    trying to match at the startoffset position, but instead was moving forward to
3422    the next newline as if a previous match had failed.
3423    
3424    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3425    and could get into a loop if a null string was matched other than at the start
3426    of the subject.
3427    
3428    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3429    be distinguished at compile time, and for completeness also added PCRE_DATE.
3430    
3431    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3432    in GnuWin32 environments.
3433    
3434    
3435    Version 2.07 29-Jul-99
3436    ----------------------
3437    
3438    1. The documentation is now supplied in plain text form and HTML as well as in
3439    the form of man page sources.
3440    
3441    2. C++ compilers don't like assigning (void *) values to other pointer types.
3442    In particular this affects malloc(). Although there is no problem in Standard
3443    C, I've put in casts to keep C++ compilers happy.
3444    
3445    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3446    should be (const char *).
3447    
3448    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3449    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3450    However, I haven't made this a standard facility. The documentation doesn't
3451    mention it, and the Makefile doesn't support it.
3452    
3453    5. The Makefile now contains an "install" target, with editable destinations at
3454    the top of the file. The pcretest program is not installed.
3455    
3456    6. pgrep -V now gives the PCRE version number and date.
3457    
3458    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3459    causing the entire string to be ignored, instead of just the last character.
3460    
3461    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3462    non-matching string, it can take a very, very long time, even for strings of
3463    quite modest length, because of the nested recursion. PCRE now does better in
3464    some of these cases. It does this by remembering the last required literal
3465    character in the pattern, and pre-searching the subject to ensure it is present
3466    before running the real match. In other words, it applies a heuristic to detect
3467    some types of certain failure quickly, and in the above example, if presented
3468    with a string that has no trailing " it gives "no match" very quickly.
3469    
3470    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3471    other alternatives are tried instead.
3472    
3473    
3474    Version 2.06 09-Jun-99
3475    ----------------------
3476    
3477    1. Change pcretest's output for amount of store used to show just the code
3478    space, because the remainder (the data block) varies in size between 32-bit and
3479    64-bit systems.
3480    
3481    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3482    start matching at. This allows lookbehinds to work when searching for multiple
3483    occurrences in a string.
3484    
3485    3. Added additional options to pcretest for testing multiple occurrences:
3486    
3487       /+   outputs the rest of the string that follows a match
3488       /g   loops for multiple occurrences, using the new startoffset argument
3489       /G   loops for multiple occurrences by passing an incremented pointer
3490    
3491    4. PCRE wasn't doing the "first character" optimization for patterns starting
3492    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3493    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3494    the letter 'x'. On long subject strings, this gives a significant speed-up.
3495    
3496    
3497    Version 2.05 21-Apr-99
3498    ----------------------
3499    
3500    1. Changed the type of magic_number from int to long int so that it works
3501    properly on 16-bit systems.
3502    
3503    2. Fixed a bug which caused patterns starting with .* not to work correctly
3504    when the subject string contained newline characters. PCRE was assuming
3505    anchoring for such patterns in all cases, which is not correct because .* will
3506    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3507    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3508    must be retried after every newline in the subject.
3509    
3510    
3511    Version 2.04 18-Feb-99
3512    ----------------------
3513    
3514    1. For parenthesized subpatterns with repeats whose minimum was zero, the