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

Diff of /code/trunk/ChangeLog

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

revision 83 by nigel, Sat Feb 24 21:41:06 2007 UTC revision 550 by ph10, Sun Oct 10 16:24:11 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.11 10-Oct-2010
5    ------------------------
6    
7    1.  (*THEN) was not working properly if there were untried alternatives prior
8        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
9        backtracked to try for "b" instead of moving to the next alternative branch
10        at the same level (in this case, to look for "c"). The Perl documentation
11        is clear that when (*THEN) is backtracked onto, it goes to the "next
12        alternative in the innermost enclosing group".
13    
14    
15    Version 8.10 25-Jun-2010
16    ------------------------
17    
18    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
19        THEN.
20    
21    2.  (*ACCEPT) was not working when inside an atomic group.
22    
23    3.  Inside a character class, \B is treated as a literal by default, but
24        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
25        causes the error). The code is unchanged, but I tidied the documentation.
26    
27    4.  Inside a character class, PCRE always treated \R and \X as literals,
28        whereas Perl faults them if its -w option is set. I have changed PCRE so
29        that it faults them when PCRE_EXTRA is set.
30    
31    5.  Added support for \N, which always matches any character other than
32        newline. (It is the same as "." when PCRE_DOTALL is not set.)
33    
34    6.  When compiling pcregrep with newer versions of gcc which may have
35        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
36        declared with attribute warn_unused_result" were given. Just casting the
37        result to (void) does not stop the warnings; a more elaborate fudge is
38        needed. I've used a macro to implement this.
39    
40    7.  Minor change to pcretest.c to avoid a compiler warning.
41    
42    8.  Added four artifical Unicode properties to help with an option to make
43        \s etc use properties (see next item). The new properties are: Xan
44        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
45    
46    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
47        use Unicode properties. (*UCP) at the start of a pattern can be used to set
48        this option. Modified pcretest to add /W to test this facility. Added
49        REG_UCP to make it available via the POSIX interface.
50    
51    10. Added --line-buffered to pcregrep.
52    
53    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
54        studied, and the match started with a letter with a code point greater than
55        127 whose first byte was different to the first byte of the other case of
56        the letter, the other case of this starting letter was not recognized
57        (#976).
58    
59    12. If a pattern that was studied started with a repeated Unicode property
60        test, for example, \p{Nd}+, there was the theoretical possibility of
61        setting up an incorrect bitmap of starting bytes, but fortunately it could
62        not have actually happened in practice until change 8 above was made (it
63        added property types that matched character-matching opcodes).
64    
65    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
66        possible starting bytes for non-anchored patterns.
67    
68    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
69        \R, and also a number of cases that involve Unicode properties, both
70        explicit and implicit when PCRE_UCP is set.
71    
72    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
73        input, it could crash or give wrong results if characters with values
74        greater than 0xc0 were present in the subject string. (Detail: it assumed
75        UTF-8 input when processing these items.)
76    
77    16. Added a lot of (int) casts to avoid compiler warnings in systems where
78        size_t is 64-bit (#991).
79    
80    17. Added a check for running out of memory when PCRE is compiled with
81        --disable-stack-for-recursion (#990).
82    
83    18. If the last data line in a file for pcretest does not have a newline on
84        the end, a newline was missing in the output.
85    
86    19. The default pcre_chartables.c file recognizes only ASCII characters (values
87        less than 128) in its various bitmaps. However, there is a facility for
88        generating tables according to the current locale when PCRE is compiled. It
89        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
90        characters, are recognized by isspace() and therefore were getting set in
91        these tables, and indeed these tables seem to approximate to ISO 8859. This
92        caused a problem in UTF-8 mode when pcre_study() was used to create a list
93        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
94        which of course cannot start UTF-8 characters. I have changed the code so
95        that only real ASCII characters (less than 128) and the correct starting
96        bytes for UTF-8 encodings are set for characters greater than 127 when in
97        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
98        altogether.)
99    
100    20. Added the /T option to pcretest so as to be able to run tests with non-
101        standard character tables, thus making it possible to include the tests
102        used for 19 above in the standard set of tests.
103    
104    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
105        reference to a subpattern the other side of a comment that contains an
106        opening parenthesis caused either an internal compiling error, or a
107        reference to the wrong subpattern.
108    
109    
110    Version 8.02 19-Mar-2010
111    ------------------------
112    
113    1.  The Unicode data tables have been updated to Unicode 5.2.0.
114    
115    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
116        configured.
117    
118    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
119        original author of that file, following a query about its status.
120    
121    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
122        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
123    
124    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
125        quantifier applied to a forward-referencing subroutine call, could compile
126        incorrect code or give the error "internal error: previously-checked
127        referenced subpattern not found".
128    
129    6.  Both MS Visual Studio and Symbian OS have problems with initializing
130        variables to point to external functions. For these systems, therefore,
131        pcre_malloc etc. are now initialized to local functions that call the
132        relevant global functions.
133    
134    7.  There were two entries missing in the vectors called coptable and poptable
135        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
136        I've fixed the data, and added a kludgy way of testing at compile time that
137        the lengths are correct (equal to the number of opcodes).
138    
139    8.  Following on from 7, I added a similar kludge to check the length of the
140        eint vector in pcreposix.c.
141    
142    9.  Error texts for pcre_compile() are held as one long string to avoid too
143        much relocation at load time. To find a text, the string is searched,
144        counting zeros. There was no check for running off the end of the string,
145        which could happen if a new error number was added without updating the
146        string.
147    
148    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
149    
150    11. \K was not working if it appeared in an atomic group or in a group that
151        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
152        \K is "not well defined" if used in an assertion. PCRE now accepts it if
153        the assertion is positive, but not if it is negative.
154    
155    12. Change 11 fortuitously reduced the size of the stack frame used in the
156        "match()" function of pcre_exec.c by one pointer. Forthcoming
157        implementation of support for (*MARK) will need an extra pointer on the
158        stack; I have reserved it now, so that the stack frame size does not
159        decrease.
160    
161    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
162        item in branch that calls a recursion is a subroutine call - as in the
163        second branch in the above example - was incorrectly given the compile-
164        time error "recursive call could loop indefinitely" because pcre_compile()
165        was not correctly checking the subroutine for matching a non-empty string.
166    
167    14. The checks for overrunning compiling workspace could trigger after an
168        overrun had occurred. This is a "should never occur" error, but it can be
169        triggered by pathological patterns such as hundreds of nested parentheses.
170        The checks now trigger 100 bytes before the end of the workspace.
171    
172    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
173    
174    
175    Version 8.01 19-Jan-2010
176    ------------------------
177    
178    1.  If a pattern contained a conditional subpattern with only one branch (in
179        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
180        computed the wrong minimum data length (which is of course zero for such
181        subpatterns). This could cause incorrect "no match" results.
182    
183    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
184        the pattern is reset in the first branch, pcre_compile() failed with
185        "internal error: code overflow at offset...". This happened only when
186        the reset was to the original external option setting. (An optimization
187        abstracts leading options settings into an external setting, which was the
188        cause of this.)
189    
190    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
191        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
192        assertion pattern did not match (meaning that the assertion was true), it
193        was incorrectly treated as false if the SKIP had been reached during the
194        matching. This also applied to assertions used as conditions.
195    
196    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
197        assertion subpattern, including such a pattern used as a condition,
198        unpredictable results occurred, instead of the error return
199        PCRE_ERROR_DFA_UITEM.
200    
201    5.  The C++ GlobalReplace function was not working like Perl for the special
202        situation when an empty string is matched. It now does the fancy magic
203        stuff that is necessary.
204    
205    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
206        removed. (These were left over from very, very early versions of PCRE.)
207    
208    7.  Some cosmetic changes to the code to make life easier when compiling it
209        as part of something else:
210    
211        (a) Change DEBUG to PCRE_DEBUG.
212    
213        (b) In pcre_compile(), rename the member of the "branch_chain" structure
214            called "current" as "current_branch", to prevent a collision with the
215            Linux macro when compiled as a kernel module.
216    
217        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
218            prevent a collision with the Linux macro when compiled as a kernel
219            module.
220    
221    8.  In pcre_compile() there are some checks for integer overflows that used to
222        cast potentially large values to (double). This has been changed to that
223        when building, a check for int64_t is made, and if it is found, it is used
224        instead, thus avoiding the use of floating point arithmetic. (There is no
225        other use of FP in PCRE.) If int64_t is not found, the fallback is to
226        double.
227    
228    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
229        2005 (difference between two addresses compared to an unsigned value).
230    
231    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
232        custom one, because of the following reported problem in Windows:
233    
234          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
235              under Win32.
236          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
237              therefore missing the function definition.
238          - The compiler thus generates a "C" signature for the test function.
239          - The linker fails to find the "C" function.
240          - PCRE fails to configure if asked to do so against libbz2.
241    
242    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
243        messages were output:
244    
245          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
246          rerunning libtoolize, to keep the correct libtool macros in-tree.
247          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
248    
249        I have done both of these things.
250    
251    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
252        most of the time, it *can* run out if it is given a pattern that contains a
253        runaway infinite recursion. I updated the discussion in the pcrestack man
254        page.
255    
256    13. Now that we have gone to the x.xx style of version numbers, the minor
257        version may start with zero. Using 08 or 09 is a bad idea because users
258        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
259        interpreted as invalid octal numbers. I've updated the previous comment in
260        configure.ac, and also added a check that gives an error if 08 or 09 are
261        used.
262    
263    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
264        causing partial matching to fail when the end of the subject matched \W
265        in a UTF-8 pattern where \W was quantified with a minimum of 3.
266    
267    15. There were some discrepancies between the declarations in pcre_internal.h
268        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
269        their definitions. The declarations used "const uschar *" and the
270        definitions used USPTR. Even though USPTR is normally defined as "const
271        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
272        reported that: "This difference in casting confuses some C++ compilers, for
273        example, SunCC recognizes above declarations as different functions and
274        generates broken code for hbpcre." I have changed the declarations to use
275        USPTR.
276    
277    16. GNU libtool is named differently on some systems. The autogen.sh script now
278        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
279        (FreeBSD).
280    
281    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
282        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
283        comment: "Figure out how to create a longlong from a string: strtoll and
284        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
285        instance, but it only takes 2 args instead of 3!"
286    
287    18. A subtle bug concerned with back references has been fixed by a change of
288        specification, with a corresponding code fix. A pattern such as
289        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
290        refers, was giving matches when it shouldn't. For example, xa=xaaa would
291        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
292        same bug. Such groups have to be quantified to be useful, or contained
293        inside another quantified group. (If there's no repetition, the reference
294        can never match.) The problem arises because, having left the group and
295        moved on to the rest of the pattern, a later failure that backtracks into
296        the group uses the captured value from the final iteration of the group
297        rather than the correct earlier one. I have fixed this in PCRE by forcing
298        any group that contains a reference to itself to be an atomic group; that
299        is, there cannot be any backtracking into it once it has completed. This is
300        similar to recursive and subroutine calls.
301    
302    
303    Version 8.00 19-Oct-09
304    ----------------------
305    
306    1.  The table for translating pcre_compile() error codes into POSIX error codes
307        was out-of-date, and there was no check on the pcre_compile() error code
308        being within the table. This could lead to an OK return being given in
309        error.
310    
311    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
312        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
313        in a Windows environment.
314    
315    3.  The pcregrep --count option prints the count for each file even when it is
316        zero, as does GNU grep. However, pcregrep was also printing all files when
317        --files-with-matches was added. Now, when both options are given, it prints
318        counts only for those files that have at least one match. (GNU grep just
319        prints the file name in this circumstance, but including the count seems
320        more useful - otherwise, why use --count?) Also ensured that the
321        combination -clh just lists non-zero counts, with no names.
322    
323    4.  The long form of the pcregrep -F option was incorrectly implemented as
324        --fixed_strings instead of --fixed-strings. This is an incompatible change,
325        but it seems right to fix it, and I didn't think it was worth preserving
326        the old behaviour.
327    
328    5.  The command line items --regex=pattern and --regexp=pattern were not
329        recognized by pcregrep, which required --regex pattern or --regexp pattern
330        (with a space rather than an '='). The man page documented the '=' forms,
331        which are compatible with GNU grep; these now work.
332    
333    6.  No libpcreposix.pc file was created for pkg-config; there was just
334        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
335    
336    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
337        when UCP support is not needed, by modifying the Python script that
338        generates it from Unicode data files. This should not matter if the module
339        is correctly used as a library, but I received one complaint about 50K of
340        unwanted data. My guess is that the person linked everything into his
341        program rather than using a library. Anyway, it does no harm.
342    
343    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
344        was a minimum greater than 1 for a wide character in a possessive
345        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
346        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
347        character. Chaos in the form of incorrect output or a compiling loop could
348        result.
349    
350    9.  The restrictions on what a pattern can contain when partial matching is
351        requested for pcre_exec() have been removed. All patterns can now be
352        partially matched by this function. In addition, if there are at least two
353        slots in the offset vector, the offset of the earliest inspected character
354        for the match and the offset of the end of the subject are set in them when
355        PCRE_ERROR_PARTIAL is returned.
356    
357    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
358        synonymous with PCRE_PARTIAL, for backwards compatibility, and
359        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
360        and may be more useful for multi-segment matching.
361    
362    11. Partial matching with pcre_exec() is now more intuitive. A partial match
363        used to be given if ever the end of the subject was reached; now it is
364        given only if matching could not proceed because another character was
365        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
366        string "Z", which now yields "no match" instead of "partial match". In the
367        case of pcre_dfa_exec(), "no match" is given if every matching path for the
368        final character ended with (*FAIL).
369    
370    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
371        if the pattern had a "must contain" character that was already found in the
372        earlier partial match, unless partial matching was again requested. For
373        example, with the pattern /dog.(body)?/, the "must contain" character is
374        "g". If the first part-match was for the string "dog", restarting with
375        "sbody" failed. This bug has been fixed.
376    
377    13. The string returned by pcre_dfa_exec() after a partial match has been
378        changed so that it starts at the first inspected character rather than the
379        first character of the match. This makes a difference only if the pattern
380        starts with a lookbehind assertion or \b or \B (\K is not supported by
381        pcre_dfa_exec()). It's an incompatible change, but it makes the two
382        matching functions compatible, and I think it's the right thing to do.
383    
384    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
385        so that the demonstration program is easily available in environments where
386        PCRE has not been installed from source.
387    
388    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
389        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
390        library.
391    
392    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
393        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
394        is not the first non-POSIX option to be added. Clearly some people find
395        these options useful.
396    
397    17. If a caller to the POSIX matching function regexec() passes a non-zero
398        value for nmatch with a NULL value for pmatch, the value of
399        nmatch is forced to zero.
400    
401    18. RunGrepTest did not have a test for the availability of the -u option of
402        the diff command, as RunTest does. It now checks in the same way as
403        RunTest, and also checks for the -b option.
404    
405    19. If an odd number of negated classes containing just a single character
406        interposed, within parentheses, between a forward reference to a named
407        subpattern and the definition of the subpattern, compilation crashed with
408        an internal error, complaining that it could not find the referenced
409        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
410        [The bug was that it was starting one character too far in when skipping
411        over the character class, thus treating the ] as data rather than
412        terminating the class. This meant it could skip too much.]
413    
414    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
415        /g option in pcretest when the pattern contains \K, which makes it possible
416        to have an empty string match not at the start, even when the pattern is
417        anchored. Updated pcretest and pcredemo to use this option.
418    
419    21. If the maximum number of capturing subpatterns in a recursion was greater
420        than the maximum at the outer level, the higher number was returned, but
421        with unset values at the outer level. The correct (outer level) value is
422        now given.
423    
424    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
425        PCRE did not set those parentheses (unlike Perl). I have now found a way to
426        make it do so. The string so far is captured, making this feature
427        compatible with Perl.
428    
429    23. The tests have been re-organized, adding tests 11 and 12, to make it
430        possible to check the Perl 5.10 features against Perl 5.10.
431    
432    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
433        pattern matches a fixed length string. PCRE did not allow this; now it
434        does. Neither allows recursion.
435    
436    25. I finally figured out how to implement a request to provide the minimum
437        length of subject string that was needed in order to match a given pattern.
438        (It was back references and recursion that I had previously got hung up
439        on.) This code has now been added to pcre_study(); it finds a lower bound
440        to the length of subject needed. It is not necessarily the greatest lower
441        bound, but using it to avoid searching strings that are too short does give
442        some useful speed-ups. The value is available to calling programs via
443        pcre_fullinfo().
444    
445    26. While implementing 25, I discovered to my embarrassment that pcretest had
446        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
447        study optimizations had never been tested with that matching function.
448        Oops. What is worse, even when it was passed study data, there was a bug in
449        pcre_dfa_exec() that meant it never actually used it. Double oops. There
450        were also very few tests of studied patterns with pcre_dfa_exec().
451    
452    27. If (?| is used to create subpatterns with duplicate numbers, they are now
453        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
454        on the other side of the coin, they are no longer allowed to have different
455        names, because these cannot be distinguished in PCRE, and this has caused
456        confusion. (This is a difference from Perl.)
457    
458    28. When duplicate subpattern names are present (necessarily with different
459        numbers, as required by 27 above), and a test is made by name in a
460        conditional pattern, either for a subpattern having been matched, or for
461        recursion in such a pattern, all the associated numbered subpatterns are
462        tested, and the overall condition is true if the condition is true for any
463        one of them. This is the way Perl works, and is also more like the way
464        testing by number works.
465    
466    
467    Version 7.9 11-Apr-09
468    ---------------------
469    
470    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
471        (pcretest), all targets were linked against these libraries. This included
472        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
473        libraries. This caused unwanted dependencies to be created. This problem
474        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
475        pcretest is linked with readline.
476    
477    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
478        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
479        moved outside it again, because FALSE and TRUE are already defined in AIX,
480        but BOOL is not.
481    
482    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
483        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
484    
485    4.  The pcregrep documentation said spaces were inserted as well as colons (or
486        hyphens) following file names and line numbers when outputting matching
487        lines. This is not true; no spaces are inserted. I have also clarified the
488        wording for the --colour (or --color) option.
489    
490    5.  In pcregrep, when --colour was used with -o, the list of matching strings
491        was not coloured; this is different to GNU grep, so I have changed it to be
492        the same.
493    
494    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
495        each matching line was coloured. Now it goes on to look for further matches
496        of any of the test patterns, which is the same behaviour as GNU grep.
497    
498    7.  A pattern that could match an empty string could cause pcregrep to loop; it
499        doesn't make sense to accept an empty string match in pcregrep, so I have
500        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
501        seems to be how GNU grep behaves.
502    
503    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
504        start or after a newline", because the conditional assertion was not being
505        correctly handled. The rule now is that both the assertion and what follows
506        in the first alternative must satisfy the test.
507    
508    9.  If auto-callout was enabled in a pattern with a conditional group whose
509        condition was an assertion, PCRE could crash during matching, both with
510        pcre_exec() and pcre_dfa_exec().
511    
512    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
513        used for matching.
514    
515    11. Unicode property support in character classes was not working for
516        characters (bytes) greater than 127 when not in UTF-8 mode.
517    
518    12. Added the -M command line option to pcretest.
519    
520    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
521    
522    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
523    
524    16. Added comments and documentation about mis-use of no_arg in the C++
525        wrapper.
526    
527    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
528        from Martin Jerabek that uses macro names for all relevant character and
529        string constants.
530    
531    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
532        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
533        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
534        these, but not everybody uses configure.
535    
536    19. A conditional group that had only one branch was not being correctly
537        recognized as an item that could match an empty string. This meant that an
538        enclosing group might also not be so recognized, causing infinite looping
539        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
540        with the subject "ab", where knowledge that the repeated group can match
541        nothing is needed in order to break the loop.
542    
543    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
544        exec(), but without supplying a callout function, matching went wrong.
545    
546    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
547        leak if the size of the offset vector was greater than 30. When the vector
548        is smaller, the saved offsets during recursion go onto a local stack
549        vector, but for larger vectors malloc() is used. It was failing to free
550        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
551        error, in fact).
552    
553    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
554        heapframe that is used only when UTF-8 support is enabled. This caused no
555        problem, but was untidy.
556    
557    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
558        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
559        included within another project.
560    
561    24. Steven Van Ingelgem's patches to add more options to the CMake support,
562        slightly modified by me:
563    
564          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
565              not building pcregrep.
566    
567          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
568              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
569    
570    25. Forward references, both numeric and by name, in patterns that made use of
571        duplicate group numbers, could behave incorrectly or give incorrect errors,
572        because when scanning forward to find the reference group, PCRE was not
573        taking into account the duplicate group numbers. A pattern such as
574        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
575    
576    26. Changed a few more instances of "const unsigned char *" to USPTR, making
577        the feature of a custom pointer more persuasive (as requested by a user).
578    
579    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
580        pcretest.c, inside #ifndefs, because it seems they are sometimes already
581        pre-defined.
582    
583    28. Added support for (*UTF8) at the start of a pattern.
584    
585    29. Arrange for flags added by the "release type" setting in CMake to be shown
586        in the configuration summary.
587    
588    
589    Version 7.8 05-Sep-08
590    ---------------------
591    
592    1.  Replaced UCP searching code with optimized version as implemented for Ad
593        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
594        stage table and inline lookup instead of a function, giving speed ups of 2
595        to 5 times on some simple patterns that I tested. Permission was given to
596        distribute the MultiStage2.py script that generates the tables (it's not in
597        the tarball, but is in the Subversion repository).
598    
599    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
600        scripts.
601    
602    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
603        a group with a zero qualifier. The result of the study could be incorrect,
604        or the function might crash, depending on the pattern.
605    
606    4.  Caseless matching was not working for non-ASCII characters in back
607        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
608        It now works when Unicode Property Support is available.
609    
610    5.  In pcretest, an escape such as \x{de} in the data was always generating
611        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
612        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
613        truncation.
614    
615    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
616    
617    7.  Added two (int) casts to pcregrep when printing the difference of two
618        pointers, in case they are 64-bit values.
619    
620    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
621        test 2 if it fails.
622    
623    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
624        and a #define of that name to empty if it is not externally set. This is to
625        allow users of MSVC to set it if necessary.
626    
627    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
628        the convenience functions in the pcre_get.c source file.
629    
630    11. An option change at the start of a pattern that had top-level alternatives
631        could cause overwriting and/or a crash. This command provoked a crash in
632        some environments:
633    
634          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
635    
636        This potential security problem was recorded as CVE-2008-2371.
637    
638    12. For a pattern where the match had to start at the beginning or immediately
639        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
640        pcre_dfa_exec() could read past the end of the passed subject if there was
641        no match. To help with detecting such bugs (e.g. with valgrind), I modified
642        pcretest so that it places the subject at the end of its malloc-ed buffer.
643    
644    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
645        exec() might read past the end of the data buffer in UTF-8 mode.
646    
647    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
648        the data contained the byte 0x85 as part of a UTF-8 character within its
649        first line. This applied both to normal and DFA matching.
650    
651    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
652        /^[^d]*?$/8 failed to match "abc".
653    
654    16. Added a missing copyright notice to pcrecpp_internal.h.
655    
656    17. Make it more clear in the documentation that values returned from
657        pcre_exec() in ovector are byte offsets, not character counts.
658    
659    18. Tidied a few places to stop certain compilers from issuing warnings.
660    
661    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
662        supplied by Stefan Weber. I made a further small update for 7.8 because
663        there is a change of source arrangements: the pcre_searchfuncs.c module is
664        replaced by pcre_ucd.c.
665    
666    
667    Version 7.7 07-May-08
668    ---------------------
669    
670    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
671        a string to a long long, pretend we don't even have a long long." This is
672        done by checking for the strtoq, strtoll, and _strtoi64 functions.
673    
674    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
675        pre-7.6 versions, which defined a global no_arg variable instead of putting
676        it in the RE class. (See also #8 below.)
677    
678    3.  Remove a line of dead code, identified by coverity and reported by Nuno
679        Lopes.
680    
681    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
682    
683        (1) The include/exclude patterns were being applied to the whole pathnames
684            of files, instead of just to the final components.
685    
686        (2) If there was more than one level of directory, the subdirectories were
687            skipped unless they satisfied the include/exclude conditions. This is
688            inconsistent with GNU grep (and could even be seen as contrary to the
689            pcregrep specification - which I improved to make it absolutely clear).
690            The action now is always to scan all levels of directory, and just
691            apply the include/exclude patterns to regular files.
692    
693    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
694        --exclude_dir in the tests to avoid scanning .svn directories.
695    
696    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
697        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
698        doesn't support NULs in patterns.
699    
700    7.  Added some missing "const"s to declarations of static tables in
701        pcre_compile.c and pcre_dfa_exec.c.
702    
703    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
704        caused by fix #2  above. (Subsequently also a second patch to fix the
705        first patch. And a third patch - this was a messy problem.)
706    
707    9.  Applied Craig's patch to remove the use of push_back().
708    
709    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
710        matching function regexec().
711    
712    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
713        which, however, unlike Perl's \g{...}, are subroutine calls, not back
714        references. PCRE supports relative numbers with this syntax (I don't think
715        Oniguruma does).
716    
717    12. Previously, a group with a zero repeat such as (...){0} was completely
718        omitted from the compiled regex. However, this means that if the group
719        was called as a subroutine from elsewhere in the pattern, things went wrong
720        (an internal error was given). Such groups are now left in the compiled
721        pattern, with a new opcode that causes them to be skipped at execution
722        time.
723    
724    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
725        to the way PCRE behaves:
726    
727        (a) A lone ] character is dis-allowed (Perl treats it as data).
728    
729        (b) A back reference to an unmatched subpattern matches an empty string
730            (Perl fails the current match path).
731    
732        (c) A data ] in a character class must be notated as \] because if the
733            first data character in a class is ], it defines an empty class. (In
734            Perl it is not possible to have an empty class.) The empty class []
735            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
736            The negative empty class [^] matches any one character, independently
737            of the DOTALL setting.
738    
739    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
740        non-existent subpattern following a character class starting with ']' and
741        containing () gave an internal compiling error instead of "reference to
742        non-existent subpattern". Fortunately, when the pattern did exist, the
743        compiled code was correct. (When scanning forwards to check for the
744        existencd of the subpattern, it was treating the data ']' as terminating
745        the class, so got the count wrong. When actually compiling, the reference
746        was subsequently set up correctly.)
747    
748    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
749        it was being rejected as not supported by pcre_dfa_exec(), even though
750        other assertions are supported. I have made pcre_dfa_exec() support
751        (*FAIL).
752    
753    16. The implementation of 13c above involved the invention of a new opcode,
754        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
755        cannot be changed at match time, I realized I could make a small
756        improvement to matching performance by compiling OP_ALLANY instead of
757        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
758        on the OP_ANY path.
759    
760    17. Compiling pcretest on Windows with readline support failed without the
761        following two fixes: (1) Make the unistd.h include conditional on
762        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
763    
764    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
765        ncurses library to be included for pcretest when ReadLine support is
766        requested, but also to allow for it to be overridden. This patch came from
767        Daniel Bergström.
768    
769    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
770        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
771        any errors with the current Unicode tables. Thanks to Peter Kankowski for
772        spotting this.
773    
774    
775    Version 7.6 28-Jan-08
776    ---------------------
777    
778    1.  A character class containing a very large number of characters with
779        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
780        overflow.
781    
782    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
783        HAVE_LONG_LONG is not defined.
784    
785    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
786        bring them up to date and include new features. This patch includes:
787    
788        - Fixed PH's badly added libz and libbz2 support.
789        - Fixed a problem with static linking.
790        - Added pcredemo. [But later removed - see 7 below.]
791        - Fixed dftables problem and added an option.
792        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
793            HAVE_LONG_LONG.
794        - Added readline support for pcretest.
795        - Added an listing of the option settings after cmake has run.
796    
797    4.  A user submitted a patch to Makefile that makes it easy to create
798        "pcre.dll" under mingw when using Configure/Make. I added stuff to
799        Makefile.am that cause it to include this special target, without
800        affecting anything else. Note that the same mingw target plus all
801        the other distribution libraries and programs are now supported
802        when configuring with CMake (see 6 below) instead of with
803        Configure/Make.
804    
805    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
806        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
807        exported in the Windows port". It has not yet been confirmed that the patch
808        solves the problem, but it does no harm.
809    
810    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
811        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
812        with CMake, and also correct the comment about stack recursion.
813    
814    7.  Remove the automatic building of pcredemo from the ./configure system and
815        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
816        of a program that users should build themselves after PCRE is installed, so
817        building it automatically is not really right. What is more, it gave
818        trouble in some build environments.
819    
820    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
821    
822    
823    Version 7.5 10-Jan-08
824    ---------------------
825    
826    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
827        values in parens when parsing an RE using the C++ wrapper."
828    
829    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
830        Characters greater than 255 were excluded from the class instead of being
831        included.
832    
833    3.  The same bug as (2) above applied to negated POSIX classes such as
834        [:^space:].
835    
836    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
837        defined or documented. It seems to have been a typo for PCRE_STATIC, so
838        I have changed it.
839    
840    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
841        first named subpattern) and a construct such as (?&a) would reference the
842        first named subpattern whose name started with "a" (in other words, the
843        length check was missing). Both these problems are fixed. "Subpattern name
844        expected" is now given for (?&) (a zero-length name), and this patch also
845        makes it give the same error for \k'' (previously it complained that that
846        was a reference to a non-existent subpattern).
847    
848    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
849        this is right because (?- can be followed by option settings as well as by
850        digits. I have, however, made the messages clearer.
851    
852    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
853        than the number used in the conditional) now cause a compile-time error.
854        This is actually not compatible with Perl, which accepts such patterns, but
855        treats the conditional as always being FALSE (as PCRE used to), but it
856        seems to me that giving a diagnostic is better.
857    
858    8.  Change "alphameric" to the more common word "alphanumeric" in comments
859        and messages.
860    
861    9.  Fix two occurrences of "backslash" in comments that should have been
862        "backspace".
863    
864    10. Remove two redundant lines of code that can never be obeyed (their function
865        was moved elsewhere).
866    
867    11. The program that makes PCRE's Unicode character property table had a bug
868        which caused it to generate incorrect table entries for sequences of
869        characters that have the same character type, but are in different scripts.
870        It amalgamated them into a single range, with the script of the first of
871        them. In other words, some characters were in the wrong script. There were
872        thirteen such cases, affecting characters in the following ranges:
873    
874          U+002b0 - U+002c1
875          U+0060c - U+0060d
876          U+0061e - U+00612
877          U+0064b - U+0065e
878          U+0074d - U+0076d
879          U+01800 - U+01805
880          U+01d00 - U+01d77
881          U+01d9b - U+01dbf
882          U+0200b - U+0200f
883          U+030fc - U+030fe
884          U+03260 - U+0327f
885          U+0fb46 - U+0fbb1
886          U+10450 - U+1049d
887    
888    12. The -o option (show only the matching part of a line) for pcregrep was not
889        compatible with GNU grep in that, if there was more than one match in a
890        line, it showed only the first of them. It now behaves in the same way as
891        GNU grep.
892    
893    13. If the -o and -v options were combined for pcregrep, it printed a blank
894        line for every non-matching line. GNU grep prints nothing, and pcregrep now
895        does the same. The return code can be used to tell if there were any
896        non-matching lines.
897    
898    14. Added --file-offsets and --line-offsets to pcregrep.
899    
900    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
901        infinitely looping recursion. The bug was that positive lookaheads were not
902        being skipped when checking for a possible empty match (negative lookaheads
903        and both kinds of lookbehind were skipped).
904    
905    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
906        inclusion of <windows.h> to before rather than after the definition of
907        INVALID_FILE_ATTRIBUTES (patch from David Byron).
908    
909    17. Specifying a possessive quantifier with a specific limit for a Unicode
910        character property caused pcre_compile() to compile bad code, which led at
911        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
912        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
913        caused the error; without that there was no problem.
914    
915    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
916    
917    19. Added --enable-pcretest-libreadline.
918    
919    20. In pcrecpp.cc, the variable 'count' was incremented twice in
920        RE::GlobalReplace(). As a result, the number of replacements returned was
921        double what it should be. I removed one of the increments, but Craig sent a
922        later patch that removed the other one (the right fix) and added unit tests
923        that check the return values (which was not done before).
924    
925    21. Several CMake things:
926    
927        (1) Arranged that, when cmake is used on Unix, the libraries end up with
928            the names libpcre and libpcreposix, not just pcre and pcreposix.
929    
930        (2) The above change means that pcretest and pcregrep are now correctly
931            linked with the newly-built libraries, not previously installed ones.
932    
933        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
934    
935    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
936        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
937        UTF-8 newline character). The key issue is that the pattern starts .*;
938        this means that the match must be either at the beginning, or after a
939        newline. The bug was in the code for advancing after a failed match and
940        checking that the new position followed a newline. It was not taking
941        account of UTF-8 characters correctly.
942    
943    23. PCRE was behaving differently from Perl in the way it recognized POSIX
944        character classes. PCRE was not treating the sequence [:...:] as a
945        character class unless the ... were all letters. Perl, however, seems to
946        allow any characters between [: and :], though of course it rejects as
947        unknown any "names" that contain non-letters, because all the known class
948        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
949        for example, whereas PCRE did not - it did not recognize a POSIX character
950        class. This seemed a bit dangerous, so the code has been changed to be
951        closer to Perl. The behaviour is not identical to Perl, because PCRE will
952        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
953        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
954        Perl does, and where it didn't before.
955    
956    24. Rewrite so as to remove the single use of %n from pcregrep because in some
957        Windows environments %n is disabled by default.
958    
959    
960    Version 7.4 21-Sep-07
961    ---------------------
962    
963    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
964        means that a class such as [\s] counted as "explicit reference to CR or
965        LF". That isn't really right - the whole point of the change was to try to
966        help when there was an actual mention of one of the two characters. So now
967        the change happens only if \r or \n (or a literal CR or LF) character is
968        encountered.
969    
970    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
971        of both had grown to the point where there were only 3 bits left.
972        Fortunately, there was spare space in the data structure, and so I have
973        moved the internal flags into a new 16-bit field to free up more option
974        bits.
975    
976    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
977        but did not set the internal JCHANGED flag - either of these is enough to
978        control the way the "get" function works - but the PCRE_INFO_JCHANGED
979        facility is supposed to tell if (?J) was ever used, so now (?J) at the
980        start sets both bits.
981    
982    4.  Added options (at build time, compile time, exec time) to change \R from
983        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
984    
985    5.  doc/pcresyntax.html was missing from the distribution.
986    
987    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
988        compatibility, even though it is no longer used.
989    
990    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
991        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
992        windows.h file is present (where different names are used). [This was
993        reversed later after testing - see 16 below.]
994    
995    8.  Changed all #include <config.h> to #include "config.h". There were also
996        some further <pcre.h> cases that I changed to "pcre.h".
997    
998    9.  When pcregrep was used with the --colour option, it missed the line ending
999        sequence off the lines that it output.
1000    
1001    10. It was pointed out to me that arrays of string pointers cause lots of
1002        relocations when a shared library is dynamically loaded. A technique of
1003        using a single long string with a table of offsets can drastically reduce
1004        these. I have refactored PCRE in four places to do this. The result is
1005        dramatic:
1006    
1007          Originally:                          290
1008          After changing UCP table:            187
1009          After changing error message table:   43
1010          After changing table of "verbs"       36
1011          After changing table of Posix names   22
1012    
1013        Thanks to the folks working on Gregex for glib for this insight.
1014    
1015    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1016        unicode-properties was also set.
1017    
1018    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1019    
1020    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1021        checked only for CRLF.
1022    
1023    14. Added casts to pcretest.c to avoid compiler warnings.
1024    
1025    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1026    
1027    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1028        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1029        entirely. This removes changes made in 7 above.
1030    
1031    17. The CMake files have been updated, and there is now more information about
1032        building with CMake in the NON-UNIX-USE document.
1033    
1034    
1035    Version 7.3 28-Aug-07
1036    ---------------------
1037    
1038     1. In the rejigging of the build system that eventually resulted in 7.1, the
1039        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1040        brackets there is not right, since it causes compilers to look for an
1041        installed pcre.h, not the version that is in the source that is being
1042        compiled (which of course may be different). I have changed it back to:
1043    
1044          #include "pcre.h"
1045    
1046        I have a vague recollection that the change was concerned with compiling in
1047        different directories, but in the new build system, that is taken care of
1048        by the VPATH setting the Makefile.
1049    
1050     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1051        when the subject happened to end in the byte 0x85 (e.g. if the last
1052        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1053        characters but of course it shouldn't be taken as a newline when it is part
1054        of another character. The bug was that, for an unlimited repeat of . in
1055        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1056        characters when looking for a newline.
1057    
1058     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1059    
1060     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1061        in debug output.
1062    
1063     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1064        long printing in the pcrecpp unittest when running under MinGW.
1065    
1066     6. ESC_K was left out of the EBCDIC table.
1067    
1068     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1069        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1070        limit also applies to "virtual nesting" when a pattern is recursive, and in
1071        this case 1000 isn't so big. I have been able to remove this limit at the
1072        expense of backing off one optimization in certain circumstances. Normally,
1073        when pcre_exec() would call its internal match() function recursively and
1074        immediately return the result unconditionally, it uses a "tail recursion"
1075        feature to save stack. However, when a subpattern that can match an empty
1076        string has an unlimited repetition quantifier, it no longer makes this
1077        optimization. That gives it a stack frame in which to save the data for
1078        checking that an empty string has been matched. Previously this was taken
1079        from the 1000-entry workspace that had been reserved. So now there is no
1080        explicit limit, but more stack is used.
1081    
1082     8. Applied Daniel's patches to solve problems with the import/export magic
1083        syntax that is required for Windows, and which was going wrong for the
1084        pcreposix and pcrecpp parts of the library. These were overlooked when this
1085        problem was solved for the main library.
1086    
1087     9. There were some crude static tests to avoid integer overflow when computing
1088        the size of patterns that contain repeated groups with explicit upper
1089        limits. As the maximum quantifier is 65535, the maximum group length was
1090        set at 30,000 so that the product of these two numbers did not overflow a
1091        32-bit integer. However, it turns out that people want to use groups that
1092        are longer than 30,000 bytes (though not repeat them that many times).
1093        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1094        made it possible to implement the integer overflow checks in a much more
1095        dynamic way, which I have now done. The artificial limitation on group
1096        length has been removed - we now have only the limit on the total length of
1097        the compiled pattern, which depends on the LINK_SIZE setting.
1098    
1099    10. Fixed a bug in the documentation for get/copy named substring when
1100        duplicate names are permitted. If none of the named substrings are set, the
1101        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1102        empty string.
1103    
1104    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1105        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1106        because the ] is interpreted as the first data character and the
1107        terminating ] is not found. PCRE has been made compatible with Perl in this
1108        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1109        cause memory overwriting.
1110    
1111    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1112        string has been matched (to stop an infinite loop). It was not recognizing
1113        a conditional subpattern that could match an empty string if that
1114        subpattern was within another subpattern. For example, it looped when
1115        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1116        condition was not nested. This bug has been fixed.
1117    
1118    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1119        past the start of the subject in the presence of bytes with the top bit
1120        set, for example "\x8aBCD".
1121    
1122    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1123        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1124    
1125    14. Optimized (?!) to (*FAIL).
1126    
1127    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1128        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1129        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1130        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1131        does: it's just the validity check that is more restrictive.
1132    
1133    16. Inserted checks for integer overflows during escape sequence (backslash)
1134        processing, and also fixed erroneous offset values for syntax errors during
1135        backslash processing.
1136    
1137    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1138        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1139    
1140    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1141        caused an overrun.
1142    
1143    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1144        something other than just ASCII characters) inside a group that had an
1145        unlimited repeat caused a loop at compile time (while checking to see
1146        whether the group could match an empty string).
1147    
1148    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1149        [\P{Any}] did so. (Error in the code for printing property names.)
1150    
1151    21. An orphan \E inside a character class could cause a crash.
1152    
1153    22. A repeated capturing bracket such as (A)? could cause a wild memory
1154        reference during compilation.
1155    
1156    23. There are several functions in pcre_compile() that scan along a compiled
1157        expression for various reasons (e.g. to see if it's fixed length for look
1158        behind). There were bugs in these functions when a repeated \p or \P was
1159        present in the pattern. These operators have additional parameters compared
1160        with \d, etc, and these were not being taken into account when moving along
1161        the compiled data. Specifically:
1162    
1163        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1164            length.
1165    
1166        (b) An item such as \pL+ within a repeated group could cause crashes or
1167            loops.
1168    
1169        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1170            "reference to non-existent subpattern" error.
1171    
1172        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1173    
1174    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1175        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1176    
1177    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1178    
1179    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1180        character were causing crashes (broken optimization).
1181    
1182    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1183        \p or \P) caused a compile-time loop.
1184    
1185    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1186        break. For example, the unstudied pattern [\r\n]A does not match the string
1187        "\r\nA" because change 7.0/46 below moves the current point on by two
1188        characters after failing to match at the start. However, the pattern \nA
1189        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1190        the same is true. There doesn't seem any very clean way out of this, but
1191        what I have chosen to do makes the common cases work: PCRE now takes note
1192        of whether there can be an explicit match for \r or \n anywhere in the
1193        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1194        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1195        pattern has explicit CR or LF references.
1196    
1197    29. Added (*CR) etc for changing newline setting at start of pattern.
1198    
1199    
1200    Version 7.2 19-Jun-07
1201    ---------------------
1202    
1203     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1204        which is apparently normally available under Windows.
1205    
1206     2. Re-jig the pcregrep tests with different newline settings in an attempt
1207        to make them independent of the local environment's newline setting.
1208    
1209     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1210    
1211     4. Some of the "internals" tests were previously cut out when the link size
1212        was not 2, because the output contained actual offsets. The recent new
1213        "Z" feature of pcretest means that these can be cut out, making the tests
1214        usable with all link sizes.
1215    
1216     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1217        stack recursion. This gives a massive performance boost under BSD, but just
1218        a small improvement under Linux. However, it saves one field in the frame
1219        in all cases.
1220    
1221     6. Added more features from the forthcoming Perl 5.10:
1222    
1223        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1224            recursion call. It refers to the nth most recently opened parentheses.
1225    
1226        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1227            to be opened parentheses.
1228    
1229        (c) Conditions that refer to capturing parentheses can be specified
1230            relatively, for example, (?(-2)... or (?(+3)...
1231    
1232        (d) \K resets the start of the current match so that everything before
1233            is not part of it.
1234    
1235        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1236    
1237        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1238            reference syntax.
1239    
1240        (g) (?| introduces a group in which the numbering of parentheses in each
1241            alternative starts with the same number.
1242    
1243        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1244    
1245     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1246        PCRE_INFO_JCHANGED.
1247    
1248     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1249        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1250        for detecting groups that can match an empty string.
1251    
1252     9. A pattern with a very large number of alternatives (more than several
1253        hundred) was running out of internal workspace during the pre-compile
1254        phase, where pcre_compile() figures out how much memory will be needed. A
1255        bit of new cunning has reduced the workspace needed for groups with
1256        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1257        workspace instead of running out of the 4096 that are available.
1258    
1259    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1260    
1261    11. Applied patch from Google to remove an optimization that didn't quite work.
1262        The report of the bug said:
1263    
1264          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1265          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1266          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1267    
1268    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1269        it matched the wrong number of bytes.
1270    
1271    
1272    Version 7.1 24-Apr-07
1273    ---------------------
1274    
1275     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1276        that is more "standard", making use of automake and other Autotools. There
1277        is some re-arrangement of the files and adjustment of comments consequent
1278        on this.
1279    
1280     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1281        for recursive directory scanning broke on some systems because the files
1282        are not scanned in any specific order and on different systems the order
1283        was different. A call to "sort" has been inserted into RunGrepTest for the
1284        approprate test as a short-term fix. In the longer term there may be an
1285        alternative.
1286    
1287     3. I had an email from Eric Raymond about problems translating some of PCRE's
1288        man pages to HTML (despite the fact that I distribute HTML pages, some
1289        people do their own conversions for various reasons). The problems
1290        concerned the use of low-level troff macros .br and .in. I have therefore
1291        removed all such uses from the man pages (some were redundant, some could
1292        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1293        HTML has been updated to handle .nf/.fi and to complain if it encounters
1294        .br or .in.
1295    
1296     4. Updated comments in configure.ac that get placed in config.h.in and also
1297        arranged for config.h to be included in the distribution, with the name
1298        config.h.generic, for the benefit of those who have to compile without
1299        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1300    
1301     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1302        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1303        makevp.bat for the current PCRE, using the additional files
1304        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1305    
1306     6. A Windows user reported a minor discrepancy with test 2, which turned out
1307        to be caused by a trailing space on an input line that had got lost in his
1308        copy. The trailing space was an accident, so I've just removed it.
1309    
1310     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1311        that is needed.
1312    
1313     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1314        as "const" (a) because they are and (b) because it helps the PHP
1315        maintainers who have recently made a script to detect big data structures
1316        in the php code that should be moved to the .rodata section. I remembered
1317        to update Builducptable as well, so it won't revert if ucptable.h is ever
1318        re-created.
1319    
1320     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1321        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1322        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1323        support is not required. This saves 1.5-2K of code, which is important in
1324        some applications.
1325    
1326        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1327        so as not to refer to the tables, even though these functions will never be
1328        called when UTF-8 support is disabled. Otherwise there are problems with a
1329        shared library.
1330    
1331    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1332    
1333        (a) It was defining its arguments as char * instead of void *.
1334    
1335        (b) It was assuming that all moves were upwards in memory; this was true
1336            a long time ago when I wrote it, but is no longer the case.
1337    
1338        The emulated memove() is provided for those environments that have neither
1339        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1340        is clearly not the case, as these two bugs were recently reported.
1341    
1342    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1343        and Detrail to create the HTML documentation, the .txt form of the man
1344        pages, and it removes trailing spaces from listed files. It also creates
1345        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1346        case, it wraps all the #defines with #ifndefs. This script should be run
1347        before "make dist".
1348    
1349    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1350        with Unicode property support.
1351    
1352        (a) For a maximizing quantifier, if the two different cases of the
1353            character were of different lengths in their UTF-8 codings (there are
1354            some cases like this - I found 11), and the matching function had to
1355            back up over a mixture of the two cases, it incorrectly assumed they
1356            were both the same length.
1357    
1358        (b) When PCRE was configured to use the heap rather than the stack for
1359            recursion during matching, it was not correctly preserving the data for
1360            the other case of a UTF-8 character when checking ahead for a match
1361            while processing a minimizing repeat. If the check also involved
1362            matching a wide character, but failed, corruption could cause an
1363            erroneous result when trying to check for a repeat of the original
1364            character.
1365    
1366    13. Some tidying changes to the testing mechanism:
1367    
1368        (a) The RunTest script now detects the internal link size and whether there
1369            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1370            values substituted by "configure". (The RunGrepTest script already did
1371            this for UTF-8.) The configure.ac script no longer substitutes the
1372            relevant variables.
1373    
1374        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1375            with length and offset values. This means that the output is different
1376            for different internal link sizes. Test 2 is skipped for link sizes
1377            other than 2 because of this, bypassing the problem. Unfortunately,
1378            there was also a test in test 3 (the locale tests) that used /B and
1379            failed for link sizes other than 2. Rather than cut the whole test out,
1380            I have added a new /Z option to pcretest that replaces the length and
1381            offset values with spaces. This is now used to make test 3 independent
1382            of link size. (Test 2 will be tidied up later.)
1383    
1384    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1385        segmentation fault instead of returning the appropriate error message.
1386    
1387    15. In multiline mode when the newline sequence was set to "any", the pattern
1388        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1389        This doesn't seem right; it now treats the CRLF combination as the line
1390        ending, and so does not match in that case. It's only a pattern such as ^$
1391        that would hit this one: something like ^ABC$ would have failed after \r
1392        and then tried again after \r\n.
1393    
1394    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1395        in an attempt to make files that differ only in their line terminators
1396        compare equal. This works on Linux.
1397    
1398    17. Under certain error circumstances pcregrep might try to free random memory
1399        as it exited. This is now fixed, thanks to valgrind.
1400    
1401    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1402        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1403        was because its rules for how to advance for /g after matching an empty
1404        string at the end of a line did not allow for this case. They now check for
1405        it specially.
1406    
1407    20. pcretest is supposed to handle patterns and data of any length, by
1408        extending its buffers when necessary. It was getting this wrong when the
1409        buffer for a data line had to be extended.
1410    
1411    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1412        CRLF as a newline sequence.
1413    
1414    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1415        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1416        I have nevertheless tidied it up.
1417    
1418    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1419    
1420    24. Added a man page for pcre-config.
1421    
1422    
1423    Version 7.0 19-Dec-06
1424    ---------------------
1425    
1426     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1427        moving to gcc 4.1.1.
1428    
1429     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1430        sys/time.h, which is documented as needed for this function. It doesn't
1431        seem to matter on Linux, but it showed up on some releases of OS X.
1432    
1433     3. It seems that there are systems where bytes whose values are greater than
1434        127 match isprint() in the "C" locale. The "C" locale should be the
1435        default when a C program starts up. In most systems, only ASCII printing
1436        characters match isprint(). This difference caused the output from pcretest
1437        to vary, making some of the tests fail. I have changed pcretest so that:
1438    
1439        (a) When it is outputting text in the compiled version of a pattern, bytes
1440            other than 32-126 are always shown as hex escapes.
1441    
1442        (b) When it is outputting text that is a matched part of a subject string,
1443            it does the same, unless a different locale has been set for the match
1444            (using the /L modifier). In this case, it uses isprint() to decide.
1445    
1446     4. Fixed a major bug that caused incorrect computation of the amount of memory
1447        required for a compiled pattern when options that changed within the
1448        pattern affected the logic of the preliminary scan that determines the
1449        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1450        that the computed length was too small. The symptoms of this bug were
1451        either the PCRE error "internal error: code overflow" from pcre_compile(),
1452        or a glibc crash with a message such as "pcretest: free(): invalid next
1453        size (fast)". Examples of patterns that provoked this bug (shown in
1454        pcretest format) are:
1455    
1456          /(?-x: )/x
1457          /(?x)(?-x: \s*#\s*)/
1458          /((?i)[\x{c0}])/8
1459          /(?i:[\x{c0}])/8
1460    
1461        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1462        is now done differently.
1463    
1464     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1465        wrapper classes; (b) implement a new function in the C++ scanner that is
1466        more efficient than the old way of doing things because it avoids levels of
1467        recursion in the regex matching; (c) add a paragraph to the documentation
1468        for the FullMatch() function.
1469    
1470     6. The escape sequence \n was being treated as whatever was defined as
1471        "newline". Not only was this contrary to the documentation, which states
1472        that \n is character 10 (hex 0A), but it also went horribly wrong when
1473        "newline" was defined as CRLF. This has been fixed.
1474    
1475     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1476        was being set to -1 for the "end of line" case (supposedly a value that no
1477        character can have). Though this value is never used (the check for end of
1478        line is "zero bytes in current character"), it caused compiler complaints.
1479        I've changed it to 0xffffffff.
1480    
1481     8. In pcre_version.c, the version string was being built by a sequence of
1482        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1483        string (as it is for production releases) called a macro with an empty
1484        argument. The C standard says the result of this is undefined. The gcc
1485        compiler treats it as an empty string (which was what was wanted) but it is
1486        reported that Visual C gives an error. The source has been hacked around to
1487        avoid this problem.
1488    
1489     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1490        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1491        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1492        of them did).
1493    
1494    10. Originally, pcretest opened its input and output without "b"; then I was
1495        told that "b" was needed in some environments, so it was added for release
1496        5.0 to both the input and output. (It makes no difference on Unix-like
1497        systems.) Later I was told that it is wrong for the input on Windows. I've
1498        now abstracted the modes into two macros, to make it easier to fiddle with
1499        them, and removed "b" from the input mode under Windows.
1500    
1501    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1502    
1503    12. Added -help and --help to pcretest as an official way of being reminded
1504        of the options.
1505    
1506    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1507        and pcrecpp.cc because they annoy compilers at high warning levels.
1508    
1509    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1510    
1511    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1512        scripts are not C programs :-) and which was not noticed because it works
1513        on Linux.
1514    
1515    16. pcretest is supposed to handle any length of pattern and data line (as one
1516        line or as a continued sequence of lines) by extending its input buffer if
1517        necessary. This feature was broken for very long pattern lines, leading to
1518        a string of junk being passed to pcre_compile() if the pattern was longer
1519        than about 50K.
1520    
1521    17. I have done a major re-factoring of the way pcre_compile() computes the
1522        amount of memory needed for a compiled pattern. Previously, there was code
1523        that made a preliminary scan of the pattern in order to do this. That was
1524        OK when PCRE was new, but as the facilities have expanded, it has become
1525        harder and harder to keep it in step with the real compile phase, and there
1526        have been a number of bugs (see for example, 4 above). I have now found a
1527        cunning way of running the real compile function in a "fake" mode that
1528        enables it to compute how much memory it would need, while actually only
1529        ever using a few hundred bytes of working memory and without too many
1530        tests of the mode. This should make future maintenance and development
1531        easier. A side effect of this work is that the limit of 200 on the nesting
1532        depth of parentheses has been removed (though this was never a serious
1533        limitation, I suspect). However, there is a downside: pcre_compile() now
1534        runs more slowly than before (30% or more, depending on the pattern). I
1535        hope this isn't a big issue. There is no effect on runtime performance.
1536    
1537    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1538        newline (only possible for the last line of a file) and it was a
1539        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1540    
1541    19. Added additional timing features to pcretest. (1) The -tm option now times
1542        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1543        separate command line item, by a number that specifies the number of
1544        repeats to use when timing. The default is 50000; this gives better
1545        precision, but takes uncomfortably long for very large patterns.
1546    
1547    20. Extended pcre_study() to be more clever in cases where a branch of a
1548        subpattern has no definite first character. For example, (a*|b*)[cd] would
1549        previously give no result from pcre_study(). Now it recognizes that the
1550        first character must be a, b, c, or d.
1551    
1552    21. There was an incorrect error "recursive call could loop indefinitely" if
1553        a subpattern (or the entire pattern) that was being tested for matching an
1554        empty string contained only one non-empty item after a nested subpattern.
1555        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1556        incorrectly, because the \d was being skipped in the check.
1557    
1558    22. The pcretest program now has a new pattern option /B and a command line
1559        option -b, which is equivalent to adding /B to every pattern. This causes
1560        it to show the compiled bytecode, without the additional information that
1561        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1562        is the same as /B/I).
1563    
1564    23. A new optimization is now able automatically to treat some sequences such
1565        as a*b as a*+b. More specifically, if something simple (such as a character
1566        or a simple class like \d) has an unlimited quantifier, and is followed by
1567        something that cannot possibly match the quantified thing, the quantifier
1568        is automatically "possessified".
1569    
1570    24. A recursive reference to a subpattern whose number was greater than 39
1571        went wrong under certain circumstances in UTF-8 mode. This bug could also
1572        have affected the operation of pcre_study().
1573    
1574    25. Realized that a little bit of performance could be had by replacing
1575        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1576    
1577    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1578    
1579    27. Possessive quantifiers such as a++ were previously implemented by turning
1580        them into atomic groups such as ($>a+). Now they have their own opcodes,
1581        which improves performance. This includes the automatically created ones
1582        from 23 above.
1583    
1584    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1585        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1586        the first matched character to be a colon. This applied both to named and
1587        numbered groups.
1588    
1589    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1590    
1591    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1592        building PCRE on Windows easier, so I have included it in the distribution.
1593    
1594    31. There is now a check in pcretest against a ridiculously large number being
1595        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1596        loop, the loop is abandoned.
1597    
1598    32. Forward references to subpatterns in conditions such as (?(2)...) where
1599        subpattern 2 is defined later cause pcre_compile() to search forwards in
1600        the pattern for the relevant set of parentheses. This search went wrong
1601        when there were unescaped parentheses in a character class, parentheses
1602        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1603    
1604    33. "Subroutine" calls and backreferences were previously restricted to
1605        referencing subpatterns earlier in the regex. This restriction has now
1606        been removed.
1607    
1608    34. Added a number of extra features that are going to be in Perl 5.10. On the
1609        whole, these are just syntactic alternatives for features that PCRE had
1610        previously implemented using the Python syntax or my own invention. The
1611        other formats are all retained for compatibility.
1612    
1613        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1614            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1615            also .NET compatible.
1616    
1617        (b) A recursion or subroutine call to a named group can now be defined as
1618            (?&name) as well as (?P>name).
1619    
1620        (c) A backreference to a named group can now be defined as \k<name> or
1621            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1622            5.10, are also .NET compatible.
1623    
1624        (d) A conditional reference to a named group can now use the syntax
1625            (?(<name>) or (?('name') as well as (?(name).
1626    
1627        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1628            groups (named and numbered) that are never evaluated inline, but can be
1629            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1630            is always false. There may be only one alternative in such a group.
1631    
1632        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1633            as the simple (?(R). The condition is true only if the most recent
1634            recursion is that of the given number or name. It does not search out
1635            through the entire recursion stack.
1636    
1637        (g) The escape \gN or \g{N} has been added, where N is a positive or
1638            negative number, specifying an absolute or relative reference.
1639    
1640    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1641        some "unreachable code" warnings.
1642    
1643    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1644        things, this adds five new scripts.
1645    
1646    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1647        There were also incompatibilities regarding the handling of \Q..\E inside
1648        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1649        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1650    
1651    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1652        matches an empty string, and forcibly breaks the loop. There were bugs in
1653        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1654        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1655        separate and independent bugs (that affected different cases) have been
1656        fixed.
1657    
1658    39. Refactored the code to abolish the use of different opcodes for small
1659        capturing bracket numbers. This is a tidy that I avoided doing when I
1660        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1661        The new approach is not only tidier, it makes it possible to reduce the
1662        memory needed to fix the previous bug (38).
1663    
1664    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1665        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1666        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1667        mode.
1668    
1669    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1670        report.
1671    
1672    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1673        copy construction and assignment in the C++ wrapper.
1674    
1675    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1676        couple of bugs that could have given wrong results in the "--newline=crlf"
1677        case.
1678    
1679    44. Added a number of casts and did some reorganization of signed/unsigned int
1680        variables following suggestions from Dair Grant. Also renamed the variable
1681        "this" as "item" because it is a C++ keyword.
1682    
1683    45. Arranged for dftables to add
1684    
1685          #include "pcre_internal.h"
1686    
1687        to pcre_chartables.c because without it, gcc 4.x may remove the array
1688        definition from the final binary if PCRE is built into a static library and
1689        dead code stripping is activated.
1690    
1691    46. For an unanchored pattern, if a match attempt fails at the start of a
1692        newline sequence, and the newline setting is CRLF or ANY, and the next two
1693        characters are CRLF, advance by two characters instead of one.
1694    
1695    
1696    Version 6.7 04-Jul-06
1697    ---------------------
1698    
1699     1. In order to handle tests when input lines are enormously long, pcretest has
1700        been re-factored so that it automatically extends its buffers when
1701        necessary. The code is crude, but this _is_ just a test program. The
1702        default size has been increased from 32K to 50K.
1703    
1704     2. The code in pcre_study() was using the value of the re argument before
1705        testing it for NULL. (Of course, in any sensible call of the function, it
1706        won't be NULL.)
1707    
1708     3. The memmove() emulation function in pcre_internal.h, which is used on
1709        systems that lack both memmove() and bcopy() - that is, hardly ever -
1710        was missing a "static" storage class specifier.
1711    
1712     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1713        containing an extended class (one that cannot be represented by a bitmap
1714        because it contains high-valued characters or Unicode property items, e.g.
1715        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1716        pattern, but PCRE should not loop if you do not (it no longer does).
1717        [Detail: two cases were found: (a) a repeated subpattern containing an
1718        extended class; (b) a recursive reference to a subpattern that followed a
1719        previous extended class. It wasn't skipping over the extended class
1720        correctly when UTF-8 mode was not set.]
1721    
1722     5. A negated single-character class was not being recognized as fixed-length
1723        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1724        compile error "lookbehind assertion is not fixed length".
1725    
1726     6. The RunPerlTest auxiliary script was showing an unexpected difference
1727        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1728        write a Perl script that can interpret lines of an input file either as
1729        byte characters or as UTF-8, which is what "perltest" was being required to
1730        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1731        can't do is switch easily at run time between having the "use utf8;" pragma
1732        or not. In the end, I fudged it by using the RunPerlTest script to insert
1733        "use utf8;" explicitly for the UTF-8 tests.
1734    
1735     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1736        the end of the subject string, contrary to the documentation and to what
1737        Perl does. This was true of both matching functions. Now it matches only at
1738        the start of the subject and immediately after *internal* newlines.
1739    
1740     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1741        a pointer to an int instead of a pointer to an unsigned long int. This
1742        caused problems on 64-bit systems.
1743    
1744     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1745        instance of the 'standard' template library not being so standard".
1746    
1747    10. There was no check on the number of named subpatterns nor the maximum
1748        length of a subpattern name. The product of these values is used to compute
1749        the size of the memory block for a compiled pattern. By supplying a very
1750        long subpattern name and a large number of named subpatterns, the size
1751        computation could be caused to overflow. This is now prevented by limiting
1752        the length of names to 32 characters, and the number of named subpatterns
1753        to 10,000.
1754    
1755    11. Subpatterns that are repeated with specific counts have to be replicated in
1756        the compiled pattern. The size of memory for this was computed from the
1757        length of the subpattern and the repeat count. The latter is limited to
1758        65535, but there was no limit on the former, meaning that integer overflow
1759        could in principle occur. The compiled length of a repeated subpattern is
1760        now limited to 30,000 bytes in order to prevent this.
1761    
1762    12. Added the optional facility to have named substrings with the same name.
1763    
1764    13. Added the ability to use a named substring as a condition, using the
1765        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1766        are numbers (not recommended). Forward references are permitted.
1767    
1768    14. Added forward references in named backreferences (if you see what I mean).
1769    
1770    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1771        pattern could run off the end of the subject. For example, the pattern
1772        "(?s)(.{1,5})"8 did this with the subject "ab".
1773    
1774    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1775        PCRE_CASELESS was set when matching characters that were quantified with ?
1776        or *.
1777    
1778    17. A character class other than a single negated character that had a minimum
1779        but no maximum quantifier - for example [ab]{6,} - was not handled
1780        correctly by pce_dfa_exec(). It would match only one character.
1781    
1782    18. A valid (though odd) pattern that looked like a POSIX character
1783        class but used an invalid character after [ (for example [[,abc,]]) caused
1784        pcre_compile() to give the error "Failed: internal error: code overflow" or
1785        in some cases to crash with a glibc free() error. This could even happen if
1786        the pattern terminated after [[ but there just happened to be a sequence of
1787        letters, a binary zero, and a closing ] in the memory that followed.
1788    
1789    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1790        over the years. Originally (before any Unicode support), just the bottom 8
1791        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1792        output from "man perlunicode" includes this:
1793    
1794          The regular expression compiler produces polymorphic opcodes.  That
1795          is, the pattern adapts to the data and automatically switches to
1796          the Unicode character scheme when presented with Unicode data--or
1797          instead uses a traditional byte scheme when presented with byte
1798          data.
1799    
1800        Sadly, a wide octal escape does not cause a switch, and in a string with
1801        no other multibyte characters, these octal escapes are treated as before.
1802        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1803        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1804        Unicode string.
1805    
1806        I have not perpetrated such confusion in PCRE. Up till now, it took just
1807        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1808        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1809        translate to the appropriate multibyte character.
1810    
1811    29. Applied some refactoring to reduce the number of warnings from Microsoft
1812        and Borland compilers. This has included removing the fudge introduced
1813        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1814        a warning about an unused variable.
1815    
1816    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1817        characters since release 4.0, because Perl (from release 5.004) does not.
1818        [Or at least, is documented not to: some releases seem to be in conflict
1819        with the documentation.] However, when a pattern was studied with
1820        pcre_study() and all its branches started with \s, PCRE still included VT
1821        as a possible starting character. Of course, this did no harm; it just
1822        caused an unnecessary match attempt.
1823    
1824    22. Removed a now-redundant internal flag bit that recorded the fact that case
1825        dependency changed within the pattern. This was once needed for "required
1826        byte" processing, but is no longer used. This recovers a now-scarce options
1827        bit. Also moved the least significant internal flag bit to the most-
1828        significant bit of the word, which was not previously used (hangover from
1829        the days when it was an int rather than a uint) to free up another bit for
1830        the future.
1831    
1832    23. Added support for CRLF line endings as well as CR and LF. As well as the
1833        default being selectable at build time, it can now be changed at runtime
1834        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1835        specify that it is scanning data with non-default line endings.
1836    
1837    24. Changed the definition of CXXLINK to make it agree with the definition of
1838        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1839    
1840    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1841        recursions. This makes a big different to stack usage for some patterns.
1842    
1843    26. If a subpattern containing a named recursion or subroutine reference such
1844        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1845        the space required for the compiled pattern went wrong and gave too small a
1846        value. Depending on the environment, this could lead to "Failed: internal
1847        error: code overflow at offset 49" or "glibc detected double free or
1848        corruption" errors.
1849    
1850    27. Applied patches from Google (a) to support the new newline modes and (b) to
1851        advance over multibyte UTF-8 characters in GlobalReplace.
1852    
1853    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1854        difference for some implementation of PCRE in some Windows version.
1855    
1856    29. Added some extra testing facilities to pcretest:
1857    
1858        \q<number>   in a data line sets the "match limit" value
1859        \Q<number>   in a data line sets the "match recursion limt" value
1860        -S <number>  sets the stack size, where <number> is in megabytes
1861    
1862        The -S option isn't available for Windows.
1863    
1864    
1865    Version 6.6 06-Feb-06
1866    ---------------------
1867    
1868     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1869        in pcreposix.h. I have copied the definition from pcre.h.
1870    
1871     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1872        because pcre.h is no longer a built file.
1873    
1874     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1875        not normally included in the compiled code.
1876    
1877    
1878    Version 6.5 01-Feb-06
1879    ---------------------
1880    
1881     1. When using the partial match feature with pcre_dfa_exec(), it was not
1882        anchoring the second and subsequent partial matches at the new starting
1883        point. This could lead to incorrect results. For example, with the pattern
1884        /1234/, partially matching against "123" and then "a4" gave a match.
1885    
1886     2. Changes to pcregrep:
1887    
1888        (a) All non-match returns from pcre_exec() were being treated as failures
1889            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1890            error message is output. Some extra information is given for the
1891            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1892            probably the only errors that are likely to be caused by users (by
1893            specifying a regex that has nested indefinite repeats, for instance).
1894            If there are more than 20 of these errors, pcregrep is abandoned.
1895    
1896        (b) A binary zero was treated as data while matching, but terminated the
1897            output line if it was written out. This has been fixed: binary zeroes
1898            are now no different to any other data bytes.
1899    
1900        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1901            used to set a locale for matching. The --locale=xxxx long option has
1902            been added (no short equivalent) to specify a locale explicitly on the
1903            pcregrep command, overriding the environment variables.
1904    
1905        (d) When -B was used with -n, some line numbers in the output were one less
1906            than they should have been.
1907    
1908        (e) Added the -o (--only-matching) option.
1909    
1910        (f) If -A or -C was used with -c (count only), some lines of context were
1911            accidentally printed for the final match.
1912    
1913        (g) Added the -H (--with-filename) option.
1914    
1915        (h) The combination of options -rh failed to suppress file names for files
1916            that were found from directory arguments.
1917    
1918        (i) Added the -D (--devices) and -d (--directories) options.
1919    
1920        (j) Added the -F (--fixed-strings) option.
1921    
1922        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1923    
1924        (l) Added the --colo(u)r option.
1925    
1926        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1927            is not present by default.
1928    
1929     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1930        items such as (?R) or (?1), when the recursion could match a number of
1931        alternatives. If it matched one of the alternatives, but subsequently,
1932        outside the recursion, there was a failure, the code tried to back up into
1933        the recursion. However, because of the way PCRE is implemented, this is not
1934        possible, and the result was an incorrect result from the match.
1935    
1936        In order to prevent this happening, the specification of recursion has
1937        been changed so that all such subpatterns are automatically treated as
1938        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1939    
1940     4. I had overlooked the fact that, in some locales, there are characters for
1941        which isalpha() is true but neither isupper() nor islower() are true. In
1942        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1943        and ordfeminine) are like this. This affected the treatment of \w and \W
1944        when they appeared in character classes, but not when they appeared outside
1945        a character class. The bit map for "word" characters is now created
1946        separately from the results of isalnum() instead of just taking it from the
1947        upper, lower, and digit maps. (Plus the underscore character, of course.)
1948    
1949     5. The above bug also affected the handling of POSIX character classes such as
1950        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1951        permanent tables. Instead, the bit maps for such a class were previously
1952        created as the appropriate unions of the upper, lower, and digit bitmaps.
1953        Now they are created by subtraction from the [[:word:]] class, which has
1954        its own bitmap.
1955    
1956     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1957        It is created by subtracting the vertical space characters (\x09, \x0a,
1958        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1959        subtraction was done in the overall bitmap for a character class, meaning
1960        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1961        be recognized. This bug has been fixed.
1962    
1963     7. Patches from the folks at Google:
1964    
1965          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1966          real life, but is still worth protecting against".
1967    
1968          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1969          regular expressions".
1970    
1971          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1972          have it.
1973    
1974          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1975          "configure" and the latter not, in order to fix a problem somebody had
1976          with compiling the Arg class on HP-UX.
1977    
1978          (e) Improve the error-handling of the C++ wrapper a little bit.
1979    
1980          (f) New tests for checking recursion limiting.
1981    
1982     8. The pcre_memmove() function, which is used only if the environment does not
1983        have a standard memmove() function (and is therefore rarely compiled),
1984        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1985        returning a result (though PCRE never actually uses the result).
1986    
1987     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1988        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1989        returned instead of calling malloc() with an overflowing number that would
1990        most likely cause subsequent chaos.
1991    
1992    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1993    
1994    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1995        with this option is matched, the nmatch and pmatch options of regexec() are
1996        ignored.
1997    
1998    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1999        provided in case anyone wants to the the POSIX interface with UTF-8
2000        strings.
2001    
2002    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2003        C++ linking (needed for some HP-UX environments).
2004    
2005    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2006        (unused parameter) and in the pcre_printint() function (omitted "default"
2007        switch label when the default is to do nothing).
2008    
2009    15. Added some code to make it possible, when PCRE is compiled as a C++
2010        library, to replace subject pointers for pcre_exec() with a smart pointer
2011        class, thus making it possible to process discontinuous strings.
2012    
2013    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2014        much the same function. They were added by different people who were trying
2015        to make PCRE easy to compile on non-Unix systems. It has been suggested
2016        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2017        for compiling on Windows systems. I have therefore replaced it with
2018        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2019        defaults to "extern" for C or "extern C" for C++, which works fine on
2020        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2021        SCOPE with something explicit in config.h. In addition:
2022    
2023        (a) pcreposix.h still had just "extern" instead of either of these macros;
2024            I have replaced it with PCRE_DATA_SCOPE.
2025    
2026        (b) Functions such as _pcre_xclass(), which are internal to the library,
2027            but external in the C sense, all had PCRE_EXPORT in their definitions.
2028            This is apparently wrong for the Windows case, so I have removed it.
2029            (It makes no difference on Unix-like systems.)
2030    
2031    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2032        of recursive calls to match(). This is different to MATCH_LIMIT because
2033        that limits the total number of calls to match(), not all of which increase
2034        the depth of recursion. Limiting the recursion depth limits the amount of
2035        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2036        when PCRE is compiled, and changed at run time. A patch from Google adds
2037        this functionality to the C++ interface.
2038    
2039    18. Changes to the handling of Unicode character properties:
2040    
2041        (a) Updated the table to Unicode 4.1.0.
2042    
2043        (b) Recognize characters that are not in the table as "Cn" (undefined).
2044    
2045        (c) I revised the way the table is implemented to a much improved format
2046            which includes recognition of ranges. It now supports the ranges that
2047            are defined in UnicodeData.txt, and it also amalgamates other
2048            characters into ranges. This has reduced the number of entries in the
2049            table from around 16,000 to around 3,000, thus reducing its size
2050            considerably. I realized I did not need to use a tree structure after
2051            all - a binary chop search is just as efficient. Having reduced the
2052            number of entries, I extended their size from 6 bytes to 8 bytes to
2053            allow for more data.
2054    
2055        (d) Added support for Unicode script names via properties such as \p{Han}.
2056    
2057    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2058        matching that character.
2059    
2060    20. When matching a repeated Unicode property with a minimum greater than zero,
2061        (for example \pL{2,}), PCRE could look past the end of the subject if it
2062        reached it while seeking the minimum number of characters. This could
2063        happen only if some of the characters were more than one byte long, because
2064        there is a check for at least the minimum number of bytes.
2065    
2066    21. Refactored the implementation of \p and \P so as to be more general, to
2067        allow for more different types of property in future. This has changed the
2068        compiled form incompatibly. Anybody with saved compiled patterns that use
2069        \p or \P will have to recompile them.
2070    
2071    22. Added "Any" and "L&" to the supported property types.
2072    
2073    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2074        but give a compile time error if the value is greater than 0xff.
2075    
2076    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2077        accidentally not being installed or uninstalled.
2078    
2079    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2080        made were to insert the current release number. This seemed silly, because
2081        it made things harder for people building PCRE on systems that don't run
2082        "configure". I have turned pcre.h into a distributed file, no longer built
2083        by "configure", with the version identification directly included. There is
2084        no longer a pcre.h.in file.
2085    
2086        However, this change necessitated a change to the pcre-config script as
2087        well. It is built from pcre-config.in, and one of the substitutions was the
2088        release number. I have updated configure.ac so that ./configure now finds
2089        the release number by grepping pcre.h.
2090    
2091    26. Added the ability to run the tests under valgrind.
2092    
2093    
2094    Version 6.4 05-Sep-05
2095    ---------------------
2096    
2097     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2098        "--" to be printed when multiple files were scanned, even when none of the
2099        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2100        consider it to be a bug, and have restored the previous behaviour.
2101    
2102     2. A couple of code tidies to get rid of compiler warnings.
2103    
2104     3. The pcretest program used to cheat by referring to symbols in the library
2105        whose names begin with _pcre_. These are internal symbols that are not
2106        really supposed to be visible externally, and in some environments it is
2107        possible to suppress them. The cheating is now confined to including
2108        certain files from the library's source, which is a bit cleaner.
2109    
2110     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2111        file's purpose clearer.
2112    
2113     5. Reorganized pcre_ucp_findchar().
2114    
2115    
2116  Version 6.3 15-Aug-05  Version 6.3 15-Aug-05
2117  ---------------------  ---------------------
2118    

Legend:
Removed from v.83  
changed lines
  Added in v.550

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12