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

Diff of /code/trunk/ChangeLog

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

revision 13 by nigel, Sat Feb 24 21:38:21 2007 UTC revision 536 by ph10, Mon Jun 7 16:39:32 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.10 03-Jun-2010
5    ------------------------
6    
7    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8        THEN.
9    
10    2.  (*ACCEPT) was not working when inside an atomic group.
11    
12    3.  Inside a character class, \B is treated as a literal by default, but
13        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14        causes the error). The code is unchanged, but I tidied the documentation.
15    
16    4.  Inside a character class, PCRE always treated \R and \X as literals,
17        whereas Perl faults them if its -w option is set. I have changed PCRE so
18        that it faults them when PCRE_EXTRA is set.
19    
20    5.  Added support for \N, which always matches any character other than
21        newline. (It is the same as "." when PCRE_DOTALL is not set.)
22    
23    6.  When compiling pcregrep with newer versions of gcc which may have
24        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25        declared with attribute warn_unused_result" were given. Just casting the
26        result to (void) does not stop the warnings; a more elaborate fudge is
27        needed. I've used a macro to implement this.
28    
29    7.  Minor change to pcretest.c to avoid a compiler warning.
30    
31    8.  Added four artifical Unicode properties to help with an option to make
32        \s etc use properties (see next item). The new properties are: Xan
33        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34    
35    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36        use Unicode properties. (*UCP) at the start of a pattern can be used to set
37        this option. Modified pcretest to add /W to test this facility. Added
38        REG_UCP to make it available via the POSIX interface.
39    
40    10. Added --line-buffered to pcregrep.
41    
42    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43        studied, and the match started with a letter with a code point greater than
44        127 whose first byte was different to the first byte of the other case of
45        the letter, the other case of this starting letter was not recognized
46        (#976).
47    
48    12. If a pattern that was studied started with a repeated Unicode property
49        test, for example, \p{Nd}+, there was the theoretical possibility of
50        setting up an incorrect bitmap of starting bytes, but fortunately it could
51        not have actually happened in practice until change 8 above was made (it
52        added property types that matched character-matching opcodes).
53    
54    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
55        possible starting bytes for non-anchored patterns.
56    
57    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
58        \R, and also a number of cases that involve Unicode properties, both
59        explicit and implicit when PCRE_UCP is set.
60    
61    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
62        input, it could crash or give wrong results if characters with values
63        greater than 0xc0 were present in the subject string. (Detail: it assumed
64        UTF-8 input when processing these items.)
65    
66    16. Added a lot of (int) casts to avoid compiler warnings in systems where
67        size_t is 64-bit (#991).
68    
69    17. Added a check for running out of memory when PCRE is compiled with
70        --disable-stack-for-recursion (#990).
71    
72    
73    
74    Version 8.02 19-Mar-2010
75    ------------------------
76    
77    1.  The Unicode data tables have been updated to Unicode 5.2.0.
78    
79    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
80        configured.
81    
82    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
83        original author of that file, following a query about its status.
84    
85    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
86        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
87    
88    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
89        quantifier applied to a forward-referencing subroutine call, could compile
90        incorrect code or give the error "internal error: previously-checked
91        referenced subpattern not found".
92    
93    6.  Both MS Visual Studio and Symbian OS have problems with initializing
94        variables to point to external functions. For these systems, therefore,
95        pcre_malloc etc. are now initialized to local functions that call the
96        relevant global functions.
97    
98    7.  There were two entries missing in the vectors called coptable and poptable
99        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
100        I've fixed the data, and added a kludgy way of testing at compile time that
101        the lengths are correct (equal to the number of opcodes).
102    
103    8.  Following on from 7, I added a similar kludge to check the length of the
104        eint vector in pcreposix.c.
105    
106    9.  Error texts for pcre_compile() are held as one long string to avoid too
107        much relocation at load time. To find a text, the string is searched,
108        counting zeros. There was no check for running off the end of the string,
109        which could happen if a new error number was added without updating the
110        string.
111    
112    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
113    
114    11. \K was not working if it appeared in an atomic group or in a group that
115        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
116        \K is "not well defined" if used in an assertion. PCRE now accepts it if
117        the assertion is positive, but not if it is negative.
118    
119    12. Change 11 fortuitously reduced the size of the stack frame used in the
120        "match()" function of pcre_exec.c by one pointer. Forthcoming
121        implementation of support for (*MARK) will need an extra pointer on the
122        stack; I have reserved it now, so that the stack frame size does not
123        decrease.
124    
125    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
126        item in branch that calls a recursion is a subroutine call - as in the
127        second branch in the above example - was incorrectly given the compile-
128        time error "recursive call could loop indefinitely" because pcre_compile()
129        was not correctly checking the subroutine for matching a non-empty string.
130    
131    14. The checks for overrunning compiling workspace could trigger after an
132        overrun had occurred. This is a "should never occur" error, but it can be
133        triggered by pathological patterns such as hundreds of nested parentheses.
134        The checks now trigger 100 bytes before the end of the workspace.
135    
136    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
137    
138    
139    Version 8.01 19-Jan-2010
140    ------------------------
141    
142    1.  If a pattern contained a conditional subpattern with only one branch (in
143        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
144        computed the wrong minimum data length (which is of course zero for such
145        subpatterns). This could cause incorrect "no match" results.
146    
147    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
148        the pattern is reset in the first branch, pcre_compile() failed with
149        "internal error: code overflow at offset...". This happened only when
150        the reset was to the original external option setting. (An optimization
151        abstracts leading options settings into an external setting, which was the
152        cause of this.)
153    
154    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
155        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
156        assertion pattern did not match (meaning that the assertion was true), it
157        was incorrectly treated as false if the SKIP had been reached during the
158        matching. This also applied to assertions used as conditions.
159    
160    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
161        assertion subpattern, including such a pattern used as a condition,
162        unpredictable results occurred, instead of the error return
163        PCRE_ERROR_DFA_UITEM.
164    
165    5.  The C++ GlobalReplace function was not working like Perl for the special
166        situation when an empty string is matched. It now does the fancy magic
167        stuff that is necessary.
168    
169    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
170        removed. (These were left over from very, very early versions of PCRE.)
171    
172    7.  Some cosmetic changes to the code to make life easier when compiling it
173        as part of something else:
174    
175        (a) Change DEBUG to PCRE_DEBUG.
176    
177        (b) In pcre_compile(), rename the member of the "branch_chain" structure
178            called "current" as "current_branch", to prevent a collision with the
179            Linux macro when compiled as a kernel module.
180    
181        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
182            prevent a collision with the Linux macro when compiled as a kernel
183            module.
184    
185    8.  In pcre_compile() there are some checks for integer overflows that used to
186        cast potentially large values to (double). This has been changed to that
187        when building, a check for int64_t is made, and if it is found, it is used
188        instead, thus avoiding the use of floating point arithmetic. (There is no
189        other use of FP in PCRE.) If int64_t is not found, the fallback is to
190        double.
191    
192    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
193        2005 (difference between two addresses compared to an unsigned value).
194    
195    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
196        custom one, because of the following reported problem in Windows:
197    
198          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
199              under Win32.
200          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
201              therefore missing the function definition.
202          - The compiler thus generates a "C" signature for the test function.
203          - The linker fails to find the "C" function.
204          - PCRE fails to configure if asked to do so against libbz2.
205    
206    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
207        messages were output:
208    
209          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
210          rerunning libtoolize, to keep the correct libtool macros in-tree.
211          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
212    
213        I have done both of these things.
214    
215    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
216        most of the time, it *can* run out if it is given a pattern that contains a
217        runaway infinite recursion. I updated the discussion in the pcrestack man
218        page.
219    
220    13. Now that we have gone to the x.xx style of version numbers, the minor
221        version may start with zero. Using 08 or 09 is a bad idea because users
222        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
223        interpreted as invalid octal numbers. I've updated the previous comment in
224        configure.ac, and also added a check that gives an error if 08 or 09 are
225        used.
226    
227    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
228        causing partial matching to fail when the end of the subject matched \W
229        in a UTF-8 pattern where \W was quantified with a minimum of 3.
230    
231    15. There were some discrepancies between the declarations in pcre_internal.h
232        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
233        their definitions. The declarations used "const uschar *" and the
234        definitions used USPTR. Even though USPTR is normally defined as "const
235        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
236        reported that: "This difference in casting confuses some C++ compilers, for
237        example, SunCC recognizes above declarations as different functions and
238        generates broken code for hbpcre." I have changed the declarations to use
239        USPTR.
240    
241    16. GNU libtool is named differently on some systems. The autogen.sh script now
242        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
243        (FreeBSD).
244    
245    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
246        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
247        comment: "Figure out how to create a longlong from a string: strtoll and
248        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
249        instance, but it only takes 2 args instead of 3!"
250    
251    18. A subtle bug concerned with back references has been fixed by a change of
252        specification, with a corresponding code fix. A pattern such as
253        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
254        refers, was giving matches when it shouldn't. For example, xa=xaaa would
255        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
256        same bug. Such groups have to be quantified to be useful, or contained
257        inside another quantified group. (If there's no repetition, the reference
258        can never match.) The problem arises because, having left the group and
259        moved on to the rest of the pattern, a later failure that backtracks into
260        the group uses the captured value from the final iteration of the group
261        rather than the correct earlier one. I have fixed this in PCRE by forcing
262        any group that contains a reference to itself to be an atomic group; that
263        is, there cannot be any backtracking into it once it has completed. This is
264        similar to recursive and subroutine calls.
265    
266    
267    Version 8.00 19-Oct-09
268    ----------------------
269    
270    1.  The table for translating pcre_compile() error codes into POSIX error codes
271        was out-of-date, and there was no check on the pcre_compile() error code
272        being within the table. This could lead to an OK return being given in
273        error.
274    
275    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
276        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
277        in a Windows environment.
278    
279    3.  The pcregrep --count option prints the count for each file even when it is
280        zero, as does GNU grep. However, pcregrep was also printing all files when
281        --files-with-matches was added. Now, when both options are given, it prints
282        counts only for those files that have at least one match. (GNU grep just
283        prints the file name in this circumstance, but including the count seems
284        more useful - otherwise, why use --count?) Also ensured that the
285        combination -clh just lists non-zero counts, with no names.
286    
287    4.  The long form of the pcregrep -F option was incorrectly implemented as
288        --fixed_strings instead of --fixed-strings. This is an incompatible change,
289        but it seems right to fix it, and I didn't think it was worth preserving
290        the old behaviour.
291    
292    5.  The command line items --regex=pattern and --regexp=pattern were not
293        recognized by pcregrep, which required --regex pattern or --regexp pattern
294        (with a space rather than an '='). The man page documented the '=' forms,
295        which are compatible with GNU grep; these now work.
296    
297    6.  No libpcreposix.pc file was created for pkg-config; there was just
298        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
299    
300    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
301        when UCP support is not needed, by modifying the Python script that
302        generates it from Unicode data files. This should not matter if the module
303        is correctly used as a library, but I received one complaint about 50K of
304        unwanted data. My guess is that the person linked everything into his
305        program rather than using a library. Anyway, it does no harm.
306    
307    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
308        was a minimum greater than 1 for a wide character in a possessive
309        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
310        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
311        character. Chaos in the form of incorrect output or a compiling loop could
312        result.
313    
314    9.  The restrictions on what a pattern can contain when partial matching is
315        requested for pcre_exec() have been removed. All patterns can now be
316        partially matched by this function. In addition, if there are at least two
317        slots in the offset vector, the offset of the earliest inspected character
318        for the match and the offset of the end of the subject are set in them when
319        PCRE_ERROR_PARTIAL is returned.
320    
321    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
322        synonymous with PCRE_PARTIAL, for backwards compatibility, and
323        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
324        and may be more useful for multi-segment matching.
325    
326    11. Partial matching with pcre_exec() is now more intuitive. A partial match
327        used to be given if ever the end of the subject was reached; now it is
328        given only if matching could not proceed because another character was
329        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
330        string "Z", which now yields "no match" instead of "partial match". In the
331        case of pcre_dfa_exec(), "no match" is given if every matching path for the
332        final character ended with (*FAIL).
333    
334    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
335        if the pattern had a "must contain" character that was already found in the
336        earlier partial match, unless partial matching was again requested. For
337        example, with the pattern /dog.(body)?/, the "must contain" character is
338        "g". If the first part-match was for the string "dog", restarting with
339        "sbody" failed. This bug has been fixed.
340    
341    13. The string returned by pcre_dfa_exec() after a partial match has been
342        changed so that it starts at the first inspected character rather than the
343        first character of the match. This makes a difference only if the pattern
344        starts with a lookbehind assertion or \b or \B (\K is not supported by
345        pcre_dfa_exec()). It's an incompatible change, but it makes the two
346        matching functions compatible, and I think it's the right thing to do.
347    
348    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
349        so that the demonstration program is easily available in environments where
350        PCRE has not been installed from source.
351    
352    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
353        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
354        library.
355    
356    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
357        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
358        is not the first non-POSIX option to be added. Clearly some people find
359        these options useful.
360    
361    17. If a caller to the POSIX matching function regexec() passes a non-zero
362        value for nmatch with a NULL value for pmatch, the value of
363        nmatch is forced to zero.
364    
365    18. RunGrepTest did not have a test for the availability of the -u option of
366        the diff command, as RunTest does. It now checks in the same way as
367        RunTest, and also checks for the -b option.
368    
369    19. If an odd number of negated classes containing just a single character
370        interposed, within parentheses, between a forward reference to a named
371        subpattern and the definition of the subpattern, compilation crashed with
372        an internal error, complaining that it could not find the referenced
373        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
374        [The bug was that it was starting one character too far in when skipping
375        over the character class, thus treating the ] as data rather than
376        terminating the class. This meant it could skip too much.]
377    
378    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
379        /g option in pcretest when the pattern contains \K, which makes it possible
380        to have an empty string match not at the start, even when the pattern is
381        anchored. Updated pcretest and pcredemo to use this option.
382    
383    21. If the maximum number of capturing subpatterns in a recursion was greater
384        than the maximum at the outer level, the higher number was returned, but
385        with unset values at the outer level. The correct (outer level) value is
386        now given.
387    
388    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
389        PCRE did not set those parentheses (unlike Perl). I have now found a way to
390        make it do so. The string so far is captured, making this feature
391        compatible with Perl.
392    
393    23. The tests have been re-organized, adding tests 11 and 12, to make it
394        possible to check the Perl 5.10 features against Perl 5.10.
395    
396    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
397        pattern matches a fixed length string. PCRE did not allow this; now it
398        does. Neither allows recursion.
399    
400    25. I finally figured out how to implement a request to provide the minimum
401        length of subject string that was needed in order to match a given pattern.
402        (It was back references and recursion that I had previously got hung up
403        on.) This code has now been added to pcre_study(); it finds a lower bound
404        to the length of subject needed. It is not necessarily the greatest lower
405        bound, but using it to avoid searching strings that are too short does give
406        some useful speed-ups. The value is available to calling programs via
407        pcre_fullinfo().
408    
409    26. While implementing 25, I discovered to my embarrassment that pcretest had
410        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
411        study optimizations had never been tested with that matching function.
412        Oops. What is worse, even when it was passed study data, there was a bug in
413        pcre_dfa_exec() that meant it never actually used it. Double oops. There
414        were also very few tests of studied patterns with pcre_dfa_exec().
415    
416    27. If (?| is used to create subpatterns with duplicate numbers, they are now
417        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
418        on the other side of the coin, they are no longer allowed to have different
419        names, because these cannot be distinguished in PCRE, and this has caused
420        confusion. (This is a difference from Perl.)
421    
422    28. When duplicate subpattern names are present (necessarily with different
423        numbers, as required by 27 above), and a test is made by name in a
424        conditional pattern, either for a subpattern having been matched, or for
425        recursion in such a pattern, all the associated numbered subpatterns are
426        tested, and the overall condition is true if the condition is true for any
427        one of them. This is the way Perl works, and is also more like the way
428        testing by number works.
429    
430    
431    Version 7.9 11-Apr-09
432    ---------------------
433    
434    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
435        (pcretest), all targets were linked against these libraries. This included
436        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
437        libraries. This caused unwanted dependencies to be created. This problem
438        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
439        pcretest is linked with readline.
440    
441    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
442        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
443        moved outside it again, because FALSE and TRUE are already defined in AIX,
444        but BOOL is not.
445    
446    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
447        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
448    
449    4.  The pcregrep documentation said spaces were inserted as well as colons (or
450        hyphens) following file names and line numbers when outputting matching
451        lines. This is not true; no spaces are inserted. I have also clarified the
452        wording for the --colour (or --color) option.
453    
454    5.  In pcregrep, when --colour was used with -o, the list of matching strings
455        was not coloured; this is different to GNU grep, so I have changed it to be
456        the same.
457    
458    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
459        each matching line was coloured. Now it goes on to look for further matches
460        of any of the test patterns, which is the same behaviour as GNU grep.
461    
462    7.  A pattern that could match an empty string could cause pcregrep to loop; it
463        doesn't make sense to accept an empty string match in pcregrep, so I have
464        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
465        seems to be how GNU grep behaves.
466    
467    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
468        start or after a newline", because the conditional assertion was not being
469        correctly handled. The rule now is that both the assertion and what follows
470        in the first alternative must satisfy the test.
471    
472    9.  If auto-callout was enabled in a pattern with a conditional group whose
473        condition was an assertion, PCRE could crash during matching, both with
474        pcre_exec() and pcre_dfa_exec().
475    
476    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
477        used for matching.
478    
479    11. Unicode property support in character classes was not working for
480        characters (bytes) greater than 127 when not in UTF-8 mode.
481    
482    12. Added the -M command line option to pcretest.
483    
484    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
485    
486    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
487    
488    16. Added comments and documentation about mis-use of no_arg in the C++
489        wrapper.
490    
491    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
492        from Martin Jerabek that uses macro names for all relevant character and
493        string constants.
494    
495    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
496        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
497        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
498        these, but not everybody uses configure.
499    
500    19. A conditional group that had only one branch was not being correctly
501        recognized as an item that could match an empty string. This meant that an
502        enclosing group might also not be so recognized, causing infinite looping
503        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
504        with the subject "ab", where knowledge that the repeated group can match
505        nothing is needed in order to break the loop.
506    
507    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
508        exec(), but without supplying a callout function, matching went wrong.
509    
510    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
511        leak if the size of the offset vector was greater than 30. When the vector
512        is smaller, the saved offsets during recursion go onto a local stack
513        vector, but for larger vectors malloc() is used. It was failing to free
514        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
515        error, in fact).
516    
517    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
518        heapframe that is used only when UTF-8 support is enabled. This caused no
519        problem, but was untidy.
520    
521    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
522        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
523        included within another project.
524    
525    24. Steven Van Ingelgem's patches to add more options to the CMake support,
526        slightly modified by me:
527    
528          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
529              not building pcregrep.
530    
531          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
532              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
533    
534    25. Forward references, both numeric and by name, in patterns that made use of
535        duplicate group numbers, could behave incorrectly or give incorrect errors,
536        because when scanning forward to find the reference group, PCRE was not
537        taking into account the duplicate group numbers. A pattern such as
538        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
539    
540    26. Changed a few more instances of "const unsigned char *" to USPTR, making
541        the feature of a custom pointer more persuasive (as requested by a user).
542    
543    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
544        pcretest.c, inside #ifndefs, because it seems they are sometimes already
545        pre-defined.
546    
547    28. Added support for (*UTF8) at the start of a pattern.
548    
549    29. Arrange for flags added by the "release type" setting in CMake to be shown
550        in the configuration summary.
551    
552    
553    Version 7.8 05-Sep-08
554    ---------------------
555    
556    1.  Replaced UCP searching code with optimized version as implemented for Ad
557        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
558        stage table and inline lookup instead of a function, giving speed ups of 2
559        to 5 times on some simple patterns that I tested. Permission was given to
560        distribute the MultiStage2.py script that generates the tables (it's not in
561        the tarball, but is in the Subversion repository).
562    
563    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
564        scripts.
565    
566    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
567        a group with a zero qualifier. The result of the study could be incorrect,
568        or the function might crash, depending on the pattern.
569    
570    4.  Caseless matching was not working for non-ASCII characters in back
571        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
572        It now works when Unicode Property Support is available.
573    
574    5.  In pcretest, an escape such as \x{de} in the data was always generating
575        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
576        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
577        truncation.
578    
579    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
580    
581    7.  Added two (int) casts to pcregrep when printing the difference of two
582        pointers, in case they are 64-bit values.
583    
584    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
585        test 2 if it fails.
586    
587    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
588        and a #define of that name to empty if it is not externally set. This is to
589        allow users of MSVC to set it if necessary.
590    
591    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
592        the convenience functions in the pcre_get.c source file.
593    
594    11. An option change at the start of a pattern that had top-level alternatives
595        could cause overwriting and/or a crash. This command provoked a crash in
596        some environments:
597    
598          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
599    
600        This potential security problem was recorded as CVE-2008-2371.
601    
602    12. For a pattern where the match had to start at the beginning or immediately
603        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
604        pcre_dfa_exec() could read past the end of the passed subject if there was
605        no match. To help with detecting such bugs (e.g. with valgrind), I modified
606        pcretest so that it places the subject at the end of its malloc-ed buffer.
607    
608    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
609        exec() might read past the end of the data buffer in UTF-8 mode.
610    
611    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
612        the data contained the byte 0x85 as part of a UTF-8 character within its
613        first line. This applied both to normal and DFA matching.
614    
615    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
616        /^[^d]*?$/8 failed to match "abc".
617    
618    16. Added a missing copyright notice to pcrecpp_internal.h.
619    
620    17. Make it more clear in the documentation that values returned from
621        pcre_exec() in ovector are byte offsets, not character counts.
622    
623    18. Tidied a few places to stop certain compilers from issuing warnings.
624    
625    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
626        supplied by Stefan Weber. I made a further small update for 7.8 because
627        there is a change of source arrangements: the pcre_searchfuncs.c module is
628        replaced by pcre_ucd.c.
629    
630    
631    Version 7.7 07-May-08
632    ---------------------
633    
634    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
635        a string to a long long, pretend we don't even have a long long." This is
636        done by checking for the strtoq, strtoll, and _strtoi64 functions.
637    
638    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
639        pre-7.6 versions, which defined a global no_arg variable instead of putting
640        it in the RE class. (See also #8 below.)
641    
642    3.  Remove a line of dead code, identified by coverity and reported by Nuno
643        Lopes.
644    
645    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
646    
647        (1) The include/exclude patterns were being applied to the whole pathnames
648            of files, instead of just to the final components.
649    
650        (2) If there was more than one level of directory, the subdirectories were
651            skipped unless they satisfied the include/exclude conditions. This is
652            inconsistent with GNU grep (and could even be seen as contrary to the
653            pcregrep specification - which I improved to make it absolutely clear).
654            The action now is always to scan all levels of directory, and just
655            apply the include/exclude patterns to regular files.
656    
657    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
658        --exclude_dir in the tests to avoid scanning .svn directories.
659    
660    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
661        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
662        doesn't support NULs in patterns.
663    
664    7.  Added some missing "const"s to declarations of static tables in
665        pcre_compile.c and pcre_dfa_exec.c.
666    
667    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
668        caused by fix #2  above. (Subsequently also a second patch to fix the
669        first patch. And a third patch - this was a messy problem.)
670    
671    9.  Applied Craig's patch to remove the use of push_back().
672    
673    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
674        matching function regexec().
675    
676    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
677        which, however, unlike Perl's \g{...}, are subroutine calls, not back
678        references. PCRE supports relative numbers with this syntax (I don't think
679        Oniguruma does).
680    
681    12. Previously, a group with a zero repeat such as (...){0} was completely
682        omitted from the compiled regex. However, this means that if the group
683        was called as a subroutine from elsewhere in the pattern, things went wrong
684        (an internal error was given). Such groups are now left in the compiled
685        pattern, with a new opcode that causes them to be skipped at execution
686        time.
687    
688    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
689        to the way PCRE behaves:
690    
691        (a) A lone ] character is dis-allowed (Perl treats it as data).
692    
693        (b) A back reference to an unmatched subpattern matches an empty string
694            (Perl fails the current match path).
695    
696        (c) A data ] in a character class must be notated as \] because if the
697            first data character in a class is ], it defines an empty class. (In
698            Perl it is not possible to have an empty class.) The empty class []
699            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
700            The negative empty class [^] matches any one character, independently
701            of the DOTALL setting.
702    
703    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
704        non-existent subpattern following a character class starting with ']' and
705        containing () gave an internal compiling error instead of "reference to
706        non-existent subpattern". Fortunately, when the pattern did exist, the
707        compiled code was correct. (When scanning forwards to check for the
708        existencd of the subpattern, it was treating the data ']' as terminating
709        the class, so got the count wrong. When actually compiling, the reference
710        was subsequently set up correctly.)
711    
712    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
713        it was being rejected as not supported by pcre_dfa_exec(), even though
714        other assertions are supported. I have made pcre_dfa_exec() support
715        (*FAIL).
716    
717    16. The implementation of 13c above involved the invention of a new opcode,
718        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
719        cannot be changed at match time, I realized I could make a small
720        improvement to matching performance by compiling OP_ALLANY instead of
721        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
722        on the OP_ANY path.
723    
724    17. Compiling pcretest on Windows with readline support failed without the
725        following two fixes: (1) Make the unistd.h include conditional on
726        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
727    
728    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
729        ncurses library to be included for pcretest when ReadLine support is
730        requested, but also to allow for it to be overridden. This patch came from
731        Daniel Bergström.
732    
733    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
734        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
735        any errors with the current Unicode tables. Thanks to Peter Kankowski for
736        spotting this.
737    
738    
739    Version 7.6 28-Jan-08
740    ---------------------
741    
742    1.  A character class containing a very large number of characters with
743        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
744        overflow.
745    
746    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
747        HAVE_LONG_LONG is not defined.
748    
749    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
750        bring them up to date and include new features. This patch includes:
751    
752        - Fixed PH's badly added libz and libbz2 support.
753        - Fixed a problem with static linking.
754        - Added pcredemo. [But later removed - see 7 below.]
755        - Fixed dftables problem and added an option.
756        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
757            HAVE_LONG_LONG.
758        - Added readline support for pcretest.
759        - Added an listing of the option settings after cmake has run.
760    
761    4.  A user submitted a patch to Makefile that makes it easy to create
762        "pcre.dll" under mingw when using Configure/Make. I added stuff to
763        Makefile.am that cause it to include this special target, without
764        affecting anything else. Note that the same mingw target plus all
765        the other distribution libraries and programs are now supported
766        when configuring with CMake (see 6 below) instead of with
767        Configure/Make.
768    
769    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
770        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
771        exported in the Windows port". It has not yet been confirmed that the patch
772        solves the problem, but it does no harm.
773    
774    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
775        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
776        with CMake, and also correct the comment about stack recursion.
777    
778    7.  Remove the automatic building of pcredemo from the ./configure system and
779        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
780        of a program that users should build themselves after PCRE is installed, so
781        building it automatically is not really right. What is more, it gave
782        trouble in some build environments.
783    
784    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
785    
786    
787    Version 7.5 10-Jan-08
788    ---------------------
789    
790    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
791        values in parens when parsing an RE using the C++ wrapper."
792    
793    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
794        Characters greater than 255 were excluded from the class instead of being
795        included.
796    
797    3.  The same bug as (2) above applied to negated POSIX classes such as
798        [:^space:].
799    
800    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
801        defined or documented. It seems to have been a typo for PCRE_STATIC, so
802        I have changed it.
803    
804    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
805        first named subpattern) and a construct such as (?&a) would reference the
806        first named subpattern whose name started with "a" (in other words, the
807        length check was missing). Both these problems are fixed. "Subpattern name
808        expected" is now given for (?&) (a zero-length name), and this patch also
809        makes it give the same error for \k'' (previously it complained that that
810        was a reference to a non-existent subpattern).
811    
812    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
813        this is right because (?- can be followed by option settings as well as by
814        digits. I have, however, made the messages clearer.
815    
816    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
817        than the number used in the conditional) now cause a compile-time error.
818        This is actually not compatible with Perl, which accepts such patterns, but
819        treats the conditional as always being FALSE (as PCRE used to), but it
820        seems to me that giving a diagnostic is better.
821    
822    8.  Change "alphameric" to the more common word "alphanumeric" in comments
823        and messages.
824    
825    9.  Fix two occurrences of "backslash" in comments that should have been
826        "backspace".
827    
828    10. Remove two redundant lines of code that can never be obeyed (their function
829        was moved elsewhere).
830    
831    11. The program that makes PCRE's Unicode character property table had a bug
832        which caused it to generate incorrect table entries for sequences of
833        characters that have the same character type, but are in different scripts.
834        It amalgamated them into a single range, with the script of the first of
835        them. In other words, some characters were in the wrong script. There were
836        thirteen such cases, affecting characters in the following ranges:
837    
838          U+002b0 - U+002c1
839          U+0060c - U+0060d
840          U+0061e - U+00612
841          U+0064b - U+0065e
842          U+0074d - U+0076d
843          U+01800 - U+01805
844          U+01d00 - U+01d77
845          U+01d9b - U+01dbf
846          U+0200b - U+0200f
847          U+030fc - U+030fe
848          U+03260 - U+0327f
849          U+0fb46 - U+0fbb1
850          U+10450 - U+1049d
851    
852    12. The -o option (show only the matching part of a line) for pcregrep was not
853        compatible with GNU grep in that, if there was more than one match in a
854        line, it showed only the first of them. It now behaves in the same way as
855        GNU grep.
856    
857    13. If the -o and -v options were combined for pcregrep, it printed a blank
858        line for every non-matching line. GNU grep prints nothing, and pcregrep now
859        does the same. The return code can be used to tell if there were any
860        non-matching lines.
861    
862    14. Added --file-offsets and --line-offsets to pcregrep.
863    
864    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
865        infinitely looping recursion. The bug was that positive lookaheads were not
866        being skipped when checking for a possible empty match (negative lookaheads
867        and both kinds of lookbehind were skipped).
868    
869    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
870        inclusion of <windows.h> to before rather than after the definition of
871        INVALID_FILE_ATTRIBUTES (patch from David Byron).
872    
873    17. Specifying a possessive quantifier with a specific limit for a Unicode
874        character property caused pcre_compile() to compile bad code, which led at
875        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
876        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
877        caused the error; without that there was no problem.
878    
879    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
880    
881    19. Added --enable-pcretest-libreadline.
882    
883    20. In pcrecpp.cc, the variable 'count' was incremented twice in
884        RE::GlobalReplace(). As a result, the number of replacements returned was
885        double what it should be. I removed one of the increments, but Craig sent a
886        later patch that removed the other one (the right fix) and added unit tests
887        that check the return values (which was not done before).
888    
889    21. Several CMake things:
890    
891        (1) Arranged that, when cmake is used on Unix, the libraries end up with
892            the names libpcre and libpcreposix, not just pcre and pcreposix.
893    
894        (2) The above change means that pcretest and pcregrep are now correctly
895            linked with the newly-built libraries, not previously installed ones.
896    
897        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
898    
899    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
900        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
901        UTF-8 newline character). The key issue is that the pattern starts .*;
902        this means that the match must be either at the beginning, or after a
903        newline. The bug was in the code for advancing after a failed match and
904        checking that the new position followed a newline. It was not taking
905        account of UTF-8 characters correctly.
906    
907    23. PCRE was behaving differently from Perl in the way it recognized POSIX
908        character classes. PCRE was not treating the sequence [:...:] as a
909        character class unless the ... were all letters. Perl, however, seems to
910        allow any characters between [: and :], though of course it rejects as
911        unknown any "names" that contain non-letters, because all the known class
912        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
913        for example, whereas PCRE did not - it did not recognize a POSIX character
914        class. This seemed a bit dangerous, so the code has been changed to be
915        closer to Perl. The behaviour is not identical to Perl, because PCRE will
916        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
917        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
918        Perl does, and where it didn't before.
919    
920    24. Rewrite so as to remove the single use of %n from pcregrep because in some
921        Windows environments %n is disabled by default.
922    
923    
924    Version 7.4 21-Sep-07
925    ---------------------
926    
927    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
928        means that a class such as [\s] counted as "explicit reference to CR or
929        LF". That isn't really right - the whole point of the change was to try to
930        help when there was an actual mention of one of the two characters. So now
931        the change happens only if \r or \n (or a literal CR or LF) character is
932        encountered.
933    
934    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
935        of both had grown to the point where there were only 3 bits left.
936        Fortunately, there was spare space in the data structure, and so I have
937        moved the internal flags into a new 16-bit field to free up more option
938        bits.
939    
940    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
941        but did not set the internal JCHANGED flag - either of these is enough to
942        control the way the "get" function works - but the PCRE_INFO_JCHANGED
943        facility is supposed to tell if (?J) was ever used, so now (?J) at the
944        start sets both bits.
945    
946    4.  Added options (at build time, compile time, exec time) to change \R from
947        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
948    
949    5.  doc/pcresyntax.html was missing from the distribution.
950    
951    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
952        compatibility, even though it is no longer used.
953    
954    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
955        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
956        windows.h file is present (where different names are used). [This was
957        reversed later after testing - see 16 below.]
958    
959    8.  Changed all #include <config.h> to #include "config.h". There were also
960        some further <pcre.h> cases that I changed to "pcre.h".
961    
962    9.  When pcregrep was used with the --colour option, it missed the line ending
963        sequence off the lines that it output.
964    
965    10. It was pointed out to me that arrays of string pointers cause lots of
966        relocations when a shared library is dynamically loaded. A technique of
967        using a single long string with a table of offsets can drastically reduce
968        these. I have refactored PCRE in four places to do this. The result is
969        dramatic:
970    
971          Originally:                          290
972          After changing UCP table:            187
973          After changing error message table:   43
974          After changing table of "verbs"       36
975          After changing table of Posix names   22
976    
977        Thanks to the folks working on Gregex for glib for this insight.
978    
979    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
980        unicode-properties was also set.
981    
982    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
983    
984    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
985        checked only for CRLF.
986    
987    14. Added casts to pcretest.c to avoid compiler warnings.
988    
989    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
990    
991    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
992        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
993        entirely. This removes changes made in 7 above.
994    
995    17. The CMake files have been updated, and there is now more information about
996        building with CMake in the NON-UNIX-USE document.
997    
998    
999    Version 7.3 28-Aug-07
1000    ---------------------
1001    
1002     1. In the rejigging of the build system that eventually resulted in 7.1, the
1003        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1004        brackets there is not right, since it causes compilers to look for an
1005        installed pcre.h, not the version that is in the source that is being
1006        compiled (which of course may be different). I have changed it back to:
1007    
1008          #include "pcre.h"
1009    
1010        I have a vague recollection that the change was concerned with compiling in
1011        different directories, but in the new build system, that is taken care of
1012        by the VPATH setting the Makefile.
1013    
1014     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1015        when the subject happened to end in the byte 0x85 (e.g. if the last
1016        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1017        characters but of course it shouldn't be taken as a newline when it is part
1018        of another character. The bug was that, for an unlimited repeat of . in
1019        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1020        characters when looking for a newline.
1021    
1022     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1023    
1024     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1025        in debug output.
1026    
1027     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1028        long printing in the pcrecpp unittest when running under MinGW.
1029    
1030     6. ESC_K was left out of the EBCDIC table.
1031    
1032     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1033        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1034        limit also applies to "virtual nesting" when a pattern is recursive, and in
1035        this case 1000 isn't so big. I have been able to remove this limit at the
1036        expense of backing off one optimization in certain circumstances. Normally,
1037        when pcre_exec() would call its internal match() function recursively and
1038        immediately return the result unconditionally, it uses a "tail recursion"
1039        feature to save stack. However, when a subpattern that can match an empty
1040        string has an unlimited repetition quantifier, it no longer makes this
1041        optimization. That gives it a stack frame in which to save the data for
1042        checking that an empty string has been matched. Previously this was taken
1043        from the 1000-entry workspace that had been reserved. So now there is no
1044        explicit limit, but more stack is used.
1045    
1046     8. Applied Daniel's patches to solve problems with the import/export magic
1047        syntax that is required for Windows, and which was going wrong for the
1048        pcreposix and pcrecpp parts of the library. These were overlooked when this
1049        problem was solved for the main library.
1050    
1051     9. There were some crude static tests to avoid integer overflow when computing
1052        the size of patterns that contain repeated groups with explicit upper
1053        limits. As the maximum quantifier is 65535, the maximum group length was
1054        set at 30,000 so that the product of these two numbers did not overflow a
1055        32-bit integer. However, it turns out that people want to use groups that
1056        are longer than 30,000 bytes (though not repeat them that many times).
1057        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1058        made it possible to implement the integer overflow checks in a much more
1059        dynamic way, which I have now done. The artificial limitation on group
1060        length has been removed - we now have only the limit on the total length of
1061        the compiled pattern, which depends on the LINK_SIZE setting.
1062    
1063    10. Fixed a bug in the documentation for get/copy named substring when
1064        duplicate names are permitted. If none of the named substrings are set, the
1065        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1066        empty string.
1067    
1068    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1069        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1070        because the ] is interpreted as the first data character and the
1071        terminating ] is not found. PCRE has been made compatible with Perl in this
1072        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1073        cause memory overwriting.
1074    
1075    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1076        string has been matched (to stop an infinite loop). It was not recognizing
1077        a conditional subpattern that could match an empty string if that
1078        subpattern was within another subpattern. For example, it looped when
1079        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1080        condition was not nested. This bug has been fixed.
1081    
1082    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1083        past the start of the subject in the presence of bytes with the top bit
1084        set, for example "\x8aBCD".
1085    
1086    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1087        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1088    
1089    14. Optimized (?!) to (*FAIL).
1090    
1091    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1092        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1093        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1094        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1095        does: it's just the validity check that is more restrictive.
1096    
1097    16. Inserted checks for integer overflows during escape sequence (backslash)
1098        processing, and also fixed erroneous offset values for syntax errors during
1099        backslash processing.
1100    
1101    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1102        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1103    
1104    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1105        caused an overrun.
1106    
1107    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1108        something other than just ASCII characters) inside a group that had an
1109        unlimited repeat caused a loop at compile time (while checking to see
1110        whether the group could match an empty string).
1111    
1112    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1113        [\P{Any}] did so. (Error in the code for printing property names.)
1114    
1115    21. An orphan \E inside a character class could cause a crash.
1116    
1117    22. A repeated capturing bracket such as (A)? could cause a wild memory
1118        reference during compilation.
1119    
1120    23. There are several functions in pcre_compile() that scan along a compiled
1121        expression for various reasons (e.g. to see if it's fixed length for look
1122        behind). There were bugs in these functions when a repeated \p or \P was
1123        present in the pattern. These operators have additional parameters compared
1124        with \d, etc, and these were not being taken into account when moving along
1125        the compiled data. Specifically:
1126    
1127        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1128            length.
1129    
1130        (b) An item such as \pL+ within a repeated group could cause crashes or
1131            loops.
1132    
1133        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1134            "reference to non-existent subpattern" error.
1135    
1136        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1137    
1138    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1139        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1140    
1141    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1142    
1143    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1144        character were causing crashes (broken optimization).
1145    
1146    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1147        \p or \P) caused a compile-time loop.
1148    
1149    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1150        break. For example, the unstudied pattern [\r\n]A does not match the string
1151        "\r\nA" because change 7.0/46 below moves the current point on by two
1152        characters after failing to match at the start. However, the pattern \nA
1153        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1154        the same is true. There doesn't seem any very clean way out of this, but
1155        what I have chosen to do makes the common cases work: PCRE now takes note
1156        of whether there can be an explicit match for \r or \n anywhere in the
1157        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1158        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1159        pattern has explicit CR or LF references.
1160    
1161    29. Added (*CR) etc for changing newline setting at start of pattern.
1162    
1163    
1164    Version 7.2 19-Jun-07
1165    ---------------------
1166    
1167     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1168        which is apparently normally available under Windows.
1169    
1170     2. Re-jig the pcregrep tests with different newline settings in an attempt
1171        to make them independent of the local environment's newline setting.
1172    
1173     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1174    
1175     4. Some of the "internals" tests were previously cut out when the link size
1176        was not 2, because the output contained actual offsets. The recent new
1177        "Z" feature of pcretest means that these can be cut out, making the tests
1178        usable with all link sizes.
1179    
1180     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1181        stack recursion. This gives a massive performance boost under BSD, but just
1182        a small improvement under Linux. However, it saves one field in the frame
1183        in all cases.
1184    
1185     6. Added more features from the forthcoming Perl 5.10:
1186    
1187        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1188            recursion call. It refers to the nth most recently opened parentheses.
1189    
1190        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1191            to be opened parentheses.
1192    
1193        (c) Conditions that refer to capturing parentheses can be specified
1194            relatively, for example, (?(-2)... or (?(+3)...
1195    
1196        (d) \K resets the start of the current match so that everything before
1197            is not part of it.
1198    
1199        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1200    
1201        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1202            reference syntax.
1203    
1204        (g) (?| introduces a group in which the numbering of parentheses in each
1205            alternative starts with the same number.
1206    
1207        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1208    
1209     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1210        PCRE_INFO_JCHANGED.
1211    
1212     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1213        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1214        for detecting groups that can match an empty string.
1215    
1216     9. A pattern with a very large number of alternatives (more than several
1217        hundred) was running out of internal workspace during the pre-compile
1218        phase, where pcre_compile() figures out how much memory will be needed. A
1219        bit of new cunning has reduced the workspace needed for groups with
1220        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1221        workspace instead of running out of the 4096 that are available.
1222    
1223    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1224    
1225    11. Applied patch from Google to remove an optimization that didn't quite work.
1226        The report of the bug said:
1227    
1228          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1229          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1230          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1231    
1232    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1233        it matched the wrong number of bytes.
1234    
1235    
1236    Version 7.1 24-Apr-07
1237    ---------------------
1238    
1239     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1240        that is more "standard", making use of automake and other Autotools. There
1241        is some re-arrangement of the files and adjustment of comments consequent
1242        on this.
1243    
1244     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1245        for recursive directory scanning broke on some systems because the files
1246        are not scanned in any specific order and on different systems the order
1247        was different. A call to "sort" has been inserted into RunGrepTest for the
1248        approprate test as a short-term fix. In the longer term there may be an
1249        alternative.
1250    
1251     3. I had an email from Eric Raymond about problems translating some of PCRE's
1252        man pages to HTML (despite the fact that I distribute HTML pages, some
1253        people do their own conversions for various reasons). The problems
1254        concerned the use of low-level troff macros .br and .in. I have therefore
1255        removed all such uses from the man pages (some were redundant, some could
1256        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1257        HTML has been updated to handle .nf/.fi and to complain if it encounters
1258        .br or .in.
1259    
1260     4. Updated comments in configure.ac that get placed in config.h.in and also
1261        arranged for config.h to be included in the distribution, with the name
1262        config.h.generic, for the benefit of those who have to compile without
1263        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1264    
1265     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1266        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1267        makevp.bat for the current PCRE, using the additional files
1268        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1269    
1270     6. A Windows user reported a minor discrepancy with test 2, which turned out
1271        to be caused by a trailing space on an input line that had got lost in his
1272        copy. The trailing space was an accident, so I've just removed it.
1273    
1274     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1275        that is needed.
1276    
1277     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1278        as "const" (a) because they are and (b) because it helps the PHP
1279        maintainers who have recently made a script to detect big data structures
1280        in the php code that should be moved to the .rodata section. I remembered
1281        to update Builducptable as well, so it won't revert if ucptable.h is ever
1282        re-created.
1283    
1284     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1285        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1286        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1287        support is not required. This saves 1.5-2K of code, which is important in
1288        some applications.
1289    
1290        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1291        so as not to refer to the tables, even though these functions will never be
1292        called when UTF-8 support is disabled. Otherwise there are problems with a
1293        shared library.
1294    
1295    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1296    
1297        (a) It was defining its arguments as char * instead of void *.
1298    
1299        (b) It was assuming that all moves were upwards in memory; this was true
1300            a long time ago when I wrote it, but is no longer the case.
1301    
1302        The emulated memove() is provided for those environments that have neither
1303        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1304        is clearly not the case, as these two bugs were recently reported.
1305    
1306    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1307        and Detrail to create the HTML documentation, the .txt form of the man
1308        pages, and it removes trailing spaces from listed files. It also creates
1309        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1310        case, it wraps all the #defines with #ifndefs. This script should be run
1311        before "make dist".
1312    
1313    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1314        with Unicode property support.
1315    
1316        (a) For a maximizing quantifier, if the two different cases of the
1317            character were of different lengths in their UTF-8 codings (there are
1318            some cases like this - I found 11), and the matching function had to
1319            back up over a mixture of the two cases, it incorrectly assumed they
1320            were both the same length.
1321    
1322        (b) When PCRE was configured to use the heap rather than the stack for
1323            recursion during matching, it was not correctly preserving the data for
1324            the other case of a UTF-8 character when checking ahead for a match
1325            while processing a minimizing repeat. If the check also involved
1326            matching a wide character, but failed, corruption could cause an
1327            erroneous result when trying to check for a repeat of the original
1328            character.
1329    
1330    13. Some tidying changes to the testing mechanism:
1331    
1332        (a) The RunTest script now detects the internal link size and whether there
1333            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1334            values substituted by "configure". (The RunGrepTest script already did
1335            this for UTF-8.) The configure.ac script no longer substitutes the
1336            relevant variables.
1337    
1338        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1339            with length and offset values. This means that the output is different
1340            for different internal link sizes. Test 2 is skipped for link sizes
1341            other than 2 because of this, bypassing the problem. Unfortunately,
1342            there was also a test in test 3 (the locale tests) that used /B and
1343            failed for link sizes other than 2. Rather than cut the whole test out,
1344            I have added a new /Z option to pcretest that replaces the length and
1345            offset values with spaces. This is now used to make test 3 independent
1346            of link size. (Test 2 will be tidied up later.)
1347    
1348    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1349        segmentation fault instead of returning the appropriate error message.
1350    
1351    15. In multiline mode when the newline sequence was set to "any", the pattern
1352        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1353        This doesn't seem right; it now treats the CRLF combination as the line
1354        ending, and so does not match in that case. It's only a pattern such as ^$
1355        that would hit this one: something like ^ABC$ would have failed after \r
1356        and then tried again after \r\n.
1357    
1358    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1359        in an attempt to make files that differ only in their line terminators
1360        compare equal. This works on Linux.
1361    
1362    17. Under certain error circumstances pcregrep might try to free random memory
1363        as it exited. This is now fixed, thanks to valgrind.
1364    
1365    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1366        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1367        was because its rules for how to advance for /g after matching an empty
1368        string at the end of a line did not allow for this case. They now check for
1369        it specially.
1370    
1371    20. pcretest is supposed to handle patterns and data of any length, by
1372        extending its buffers when necessary. It was getting this wrong when the
1373        buffer for a data line had to be extended.
1374    
1375    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1376        CRLF as a newline sequence.
1377    
1378    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1379        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1380        I have nevertheless tidied it up.
1381    
1382    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1383    
1384    24. Added a man page for pcre-config.
1385    
1386    
1387    Version 7.0 19-Dec-06
1388    ---------------------
1389    
1390     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1391        moving to gcc 4.1.1.
1392    
1393     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1394        sys/time.h, which is documented as needed for this function. It doesn't
1395        seem to matter on Linux, but it showed up on some releases of OS X.
1396    
1397     3. It seems that there are systems where bytes whose values are greater than
1398        127 match isprint() in the "C" locale. The "C" locale should be the
1399        default when a C program starts up. In most systems, only ASCII printing
1400        characters match isprint(). This difference caused the output from pcretest
1401        to vary, making some of the tests fail. I have changed pcretest so that:
1402    
1403        (a) When it is outputting text in the compiled version of a pattern, bytes
1404            other than 32-126 are always shown as hex escapes.
1405    
1406        (b) When it is outputting text that is a matched part of a subject string,
1407            it does the same, unless a different locale has been set for the match
1408            (using the /L modifier). In this case, it uses isprint() to decide.
1409    
1410     4. Fixed a major bug that caused incorrect computation of the amount of memory
1411        required for a compiled pattern when options that changed within the
1412        pattern affected the logic of the preliminary scan that determines the
1413        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1414        that the computed length was too small. The symptoms of this bug were
1415        either the PCRE error "internal error: code overflow" from pcre_compile(),
1416        or a glibc crash with a message such as "pcretest: free(): invalid next
1417        size (fast)". Examples of patterns that provoked this bug (shown in
1418        pcretest format) are:
1419    
1420          /(?-x: )/x
1421          /(?x)(?-x: \s*#\s*)/
1422          /((?i)[\x{c0}])/8
1423          /(?i:[\x{c0}])/8
1424    
1425        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1426        is now done differently.
1427    
1428     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1429        wrapper classes; (b) implement a new function in the C++ scanner that is
1430        more efficient than the old way of doing things because it avoids levels of
1431        recursion in the regex matching; (c) add a paragraph to the documentation
1432        for the FullMatch() function.
1433    
1434     6. The escape sequence \n was being treated as whatever was defined as
1435        "newline". Not only was this contrary to the documentation, which states
1436        that \n is character 10 (hex 0A), but it also went horribly wrong when
1437        "newline" was defined as CRLF. This has been fixed.
1438    
1439     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1440        was being set to -1 for the "end of line" case (supposedly a value that no
1441        character can have). Though this value is never used (the check for end of
1442        line is "zero bytes in current character"), it caused compiler complaints.
1443        I've changed it to 0xffffffff.
1444    
1445     8. In pcre_version.c, the version string was being built by a sequence of
1446        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1447        string (as it is for production releases) called a macro with an empty
1448        argument. The C standard says the result of this is undefined. The gcc
1449        compiler treats it as an empty string (which was what was wanted) but it is
1450        reported that Visual C gives an error. The source has been hacked around to
1451        avoid this problem.
1452    
1453     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1454        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1455        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1456        of them did).
1457    
1458    10. Originally, pcretest opened its input and output without "b"; then I was
1459        told that "b" was needed in some environments, so it was added for release
1460        5.0 to both the input and output. (It makes no difference on Unix-like
1461        systems.) Later I was told that it is wrong for the input on Windows. I've
1462        now abstracted the modes into two macros, to make it easier to fiddle with
1463        them, and removed "b" from the input mode under Windows.
1464    
1465    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1466    
1467    12. Added -help and --help to pcretest as an official way of being reminded
1468        of the options.
1469    
1470    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1471        and pcrecpp.cc because they annoy compilers at high warning levels.
1472    
1473    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1474    
1475    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1476        scripts are not C programs :-) and which was not noticed because it works
1477        on Linux.
1478    
1479    16. pcretest is supposed to handle any length of pattern and data line (as one
1480        line or as a continued sequence of lines) by extending its input buffer if
1481        necessary. This feature was broken for very long pattern lines, leading to
1482        a string of junk being passed to pcre_compile() if the pattern was longer
1483        than about 50K.
1484    
1485    17. I have done a major re-factoring of the way pcre_compile() computes the
1486        amount of memory needed for a compiled pattern. Previously, there was code
1487        that made a preliminary scan of the pattern in order to do this. That was
1488        OK when PCRE was new, but as the facilities have expanded, it has become
1489        harder and harder to keep it in step with the real compile phase, and there
1490        have been a number of bugs (see for example, 4 above). I have now found a
1491        cunning way of running the real compile function in a "fake" mode that
1492        enables it to compute how much memory it would need, while actually only
1493        ever using a few hundred bytes of working memory and without too many
1494        tests of the mode. This should make future maintenance and development
1495        easier. A side effect of this work is that the limit of 200 on the nesting
1496        depth of parentheses has been removed (though this was never a serious
1497        limitation, I suspect). However, there is a downside: pcre_compile() now
1498        runs more slowly than before (30% or more, depending on the pattern). I
1499        hope this isn't a big issue. There is no effect on runtime performance.
1500    
1501    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1502        newline (only possible for the last line of a file) and it was a
1503        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1504    
1505    19. Added additional timing features to pcretest. (1) The -tm option now times
1506        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1507        separate command line item, by a number that specifies the number of
1508        repeats to use when timing. The default is 50000; this gives better
1509        precision, but takes uncomfortably long for very large patterns.
1510    
1511    20. Extended pcre_study() to be more clever in cases where a branch of a
1512        subpattern has no definite first character. For example, (a*|b*)[cd] would
1513        previously give no result from pcre_study(). Now it recognizes that the
1514        first character must be a, b, c, or d.
1515    
1516    21. There was an incorrect error "recursive call could loop indefinitely" if
1517        a subpattern (or the entire pattern) that was being tested for matching an
1518        empty string contained only one non-empty item after a nested subpattern.
1519        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1520        incorrectly, because the \d was being skipped in the check.
1521    
1522    22. The pcretest program now has a new pattern option /B and a command line
1523        option -b, which is equivalent to adding /B to every pattern. This causes
1524        it to show the compiled bytecode, without the additional information that
1525        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1526        is the same as /B/I).
1527    
1528    23. A new optimization is now able automatically to treat some sequences such
1529        as a*b as a*+b. More specifically, if something simple (such as a character
1530        or a simple class like \d) has an unlimited quantifier, and is followed by
1531        something that cannot possibly match the quantified thing, the quantifier
1532        is automatically "possessified".
1533    
1534    24. A recursive reference to a subpattern whose number was greater than 39
1535        went wrong under certain circumstances in UTF-8 mode. This bug could also
1536        have affected the operation of pcre_study().
1537    
1538    25. Realized that a little bit of performance could be had by replacing
1539        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1540    
1541    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1542    
1543    27. Possessive quantifiers such as a++ were previously implemented by turning
1544        them into atomic groups such as ($>a+). Now they have their own opcodes,
1545        which improves performance. This includes the automatically created ones
1546        from 23 above.
1547    
1548    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1549        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1550        the first matched character to be a colon. This applied both to named and
1551        numbered groups.
1552    
1553    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1554    
1555    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1556        building PCRE on Windows easier, so I have included it in the distribution.
1557    
1558    31. There is now a check in pcretest against a ridiculously large number being
1559        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1560        loop, the loop is abandoned.
1561    
1562    32. Forward references to subpatterns in conditions such as (?(2)...) where
1563        subpattern 2 is defined later cause pcre_compile() to search forwards in
1564        the pattern for the relevant set of parentheses. This search went wrong
1565        when there were unescaped parentheses in a character class, parentheses
1566        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1567    
1568    33. "Subroutine" calls and backreferences were previously restricted to
1569        referencing subpatterns earlier in the regex. This restriction has now
1570        been removed.
1571    
1572    34. Added a number of extra features that are going to be in Perl 5.10. On the
1573        whole, these are just syntactic alternatives for features that PCRE had
1574        previously implemented using the Python syntax or my own invention. The
1575        other formats are all retained for compatibility.
1576    
1577        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1578            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1579            also .NET compatible.
1580    
1581        (b) A recursion or subroutine call to a named group can now be defined as
1582            (?&name) as well as (?P>name).
1583    
1584        (c) A backreference to a named group can now be defined as \k<name> or
1585            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1586            5.10, are also .NET compatible.
1587    
1588        (d) A conditional reference to a named group can now use the syntax
1589            (?(<name>) or (?('name') as well as (?(name).
1590    
1591        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1592            groups (named and numbered) that are never evaluated inline, but can be
1593            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1594            is always false. There may be only one alternative in such a group.
1595    
1596        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1597            as the simple (?(R). The condition is true only if the most recent
1598            recursion is that of the given number or name. It does not search out
1599            through the entire recursion stack.
1600    
1601        (g) The escape \gN or \g{N} has been added, where N is a positive or
1602            negative number, specifying an absolute or relative reference.
1603    
1604    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1605        some "unreachable code" warnings.
1606    
1607    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1608        things, this adds five new scripts.
1609    
1610    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1611        There were also incompatibilities regarding the handling of \Q..\E inside
1612        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1613        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1614    
1615    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1616        matches an empty string, and forcibly breaks the loop. There were bugs in
1617        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1618        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1619        separate and independent bugs (that affected different cases) have been
1620        fixed.
1621    
1622    39. Refactored the code to abolish the use of different opcodes for small
1623        capturing bracket numbers. This is a tidy that I avoided doing when I
1624        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1625        The new approach is not only tidier, it makes it possible to reduce the
1626        memory needed to fix the previous bug (38).
1627    
1628    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1629        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1630        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1631        mode.
1632    
1633    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1634        report.
1635    
1636    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1637        copy construction and assignment in the C++ wrapper.
1638    
1639    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1640        couple of bugs that could have given wrong results in the "--newline=crlf"
1641        case.
1642    
1643    44. Added a number of casts and did some reorganization of signed/unsigned int
1644        variables following suggestions from Dair Grant. Also renamed the variable
1645        "this" as "item" because it is a C++ keyword.
1646    
1647    45. Arranged for dftables to add
1648    
1649          #include "pcre_internal.h"
1650    
1651        to pcre_chartables.c because without it, gcc 4.x may remove the array
1652        definition from the final binary if PCRE is built into a static library and
1653        dead code stripping is activated.
1654    
1655    46. For an unanchored pattern, if a match attempt fails at the start of a
1656        newline sequence, and the newline setting is CRLF or ANY, and the next two
1657        characters are CRLF, advance by two characters instead of one.
1658    
1659    
1660    Version 6.7 04-Jul-06
1661    ---------------------
1662    
1663     1. In order to handle tests when input lines are enormously long, pcretest has
1664        been re-factored so that it automatically extends its buffers when
1665        necessary. The code is crude, but this _is_ just a test program. The
1666        default size has been increased from 32K to 50K.
1667    
1668     2. The code in pcre_study() was using the value of the re argument before
1669        testing it for NULL. (Of course, in any sensible call of the function, it
1670        won't be NULL.)
1671    
1672     3. The memmove() emulation function in pcre_internal.h, which is used on
1673        systems that lack both memmove() and bcopy() - that is, hardly ever -
1674        was missing a "static" storage class specifier.
1675    
1676     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1677        containing an extended class (one that cannot be represented by a bitmap
1678        because it contains high-valued characters or Unicode property items, e.g.
1679        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1680        pattern, but PCRE should not loop if you do not (it no longer does).
1681        [Detail: two cases were found: (a) a repeated subpattern containing an
1682        extended class; (b) a recursive reference to a subpattern that followed a
1683        previous extended class. It wasn't skipping over the extended class
1684        correctly when UTF-8 mode was not set.]
1685    
1686     5. A negated single-character class was not being recognized as fixed-length
1687        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1688        compile error "lookbehind assertion is not fixed length".
1689    
1690     6. The RunPerlTest auxiliary script was showing an unexpected difference
1691        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1692        write a Perl script that can interpret lines of an input file either as
1693        byte characters or as UTF-8, which is what "perltest" was being required to
1694        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1695        can't do is switch easily at run time between having the "use utf8;" pragma
1696        or not. In the end, I fudged it by using the RunPerlTest script to insert
1697        "use utf8;" explicitly for the UTF-8 tests.
1698    
1699     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1700        the end of the subject string, contrary to the documentation and to what
1701        Perl does. This was true of both matching functions. Now it matches only at
1702        the start of the subject and immediately after *internal* newlines.
1703    
1704     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1705        a pointer to an int instead of a pointer to an unsigned long int. This
1706        caused problems on 64-bit systems.
1707    
1708     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1709        instance of the 'standard' template library not being so standard".
1710    
1711    10. There was no check on the number of named subpatterns nor the maximum
1712        length of a subpattern name. The product of these values is used to compute
1713        the size of the memory block for a compiled pattern. By supplying a very
1714        long subpattern name and a large number of named subpatterns, the size
1715        computation could be caused to overflow. This is now prevented by limiting
1716        the length of names to 32 characters, and the number of named subpatterns
1717        to 10,000.
1718    
1719    11. Subpatterns that are repeated with specific counts have to be replicated in
1720        the compiled pattern. The size of memory for this was computed from the
1721        length of the subpattern and the repeat count. The latter is limited to
1722        65535, but there was no limit on the former, meaning that integer overflow
1723        could in principle occur. The compiled length of a repeated subpattern is
1724        now limited to 30,000 bytes in order to prevent this.
1725    
1726    12. Added the optional facility to have named substrings with the same name.
1727    
1728    13. Added the ability to use a named substring as a condition, using the
1729        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1730        are numbers (not recommended). Forward references are permitted.
1731    
1732    14. Added forward references in named backreferences (if you see what I mean).
1733    
1734    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1735        pattern could run off the end of the subject. For example, the pattern
1736        "(?s)(.{1,5})"8 did this with the subject "ab".
1737    
1738    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1739        PCRE_CASELESS was set when matching characters that were quantified with ?
1740        or *.
1741    
1742    17. A character class other than a single negated character that had a minimum
1743        but no maximum quantifier - for example [ab]{6,} - was not handled
1744        correctly by pce_dfa_exec(). It would match only one character.
1745    
1746    18. A valid (though odd) pattern that looked like a POSIX character
1747        class but used an invalid character after [ (for example [[,abc,]]) caused
1748        pcre_compile() to give the error "Failed: internal error: code overflow" or
1749        in some cases to crash with a glibc free() error. This could even happen if
1750        the pattern terminated after [[ but there just happened to be a sequence of
1751        letters, a binary zero, and a closing ] in the memory that followed.
1752    
1753    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1754        over the years. Originally (before any Unicode support), just the bottom 8
1755        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1756        output from "man perlunicode" includes this:
1757    
1758          The regular expression compiler produces polymorphic opcodes.  That
1759          is, the pattern adapts to the data and automatically switches to
1760          the Unicode character scheme when presented with Unicode data--or
1761          instead uses a traditional byte scheme when presented with byte
1762          data.
1763    
1764        Sadly, a wide octal escape does not cause a switch, and in a string with
1765        no other multibyte characters, these octal escapes are treated as before.
1766        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1767        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1768        Unicode string.
1769    
1770        I have not perpetrated such confusion in PCRE. Up till now, it took just
1771        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1772        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1773        translate to the appropriate multibyte character.
1774    
1775    29. Applied some refactoring to reduce the number of warnings from Microsoft
1776        and Borland compilers. This has included removing the fudge introduced
1777        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1778        a warning about an unused variable.
1779    
1780    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1781        characters since release 4.0, because Perl (from release 5.004) does not.
1782        [Or at least, is documented not to: some releases seem to be in conflict
1783        with the documentation.] However, when a pattern was studied with
1784        pcre_study() and all its branches started with \s, PCRE still included VT
1785        as a possible starting character. Of course, this did no harm; it just
1786        caused an unnecessary match attempt.
1787    
1788    22. Removed a now-redundant internal flag bit that recorded the fact that case
1789        dependency changed within the pattern. This was once needed for "required
1790        byte" processing, but is no longer used. This recovers a now-scarce options
1791        bit. Also moved the least significant internal flag bit to the most-
1792        significant bit of the word, which was not previously used (hangover from
1793        the days when it was an int rather than a uint) to free up another bit for
1794        the future.
1795    
1796    23. Added support for CRLF line endings as well as CR and LF. As well as the
1797        default being selectable at build time, it can now be changed at runtime
1798        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1799        specify that it is scanning data with non-default line endings.
1800    
1801    24. Changed the definition of CXXLINK to make it agree with the definition of
1802        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1803    
1804    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1805        recursions. This makes a big different to stack usage for some patterns.
1806    
1807    26. If a subpattern containing a named recursion or subroutine reference such
1808        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1809        the space required for the compiled pattern went wrong and gave too small a
1810        value. Depending on the environment, this could lead to "Failed: internal
1811        error: code overflow at offset 49" or "glibc detected double free or
1812        corruption" errors.
1813    
1814    27. Applied patches from Google (a) to support the new newline modes and (b) to
1815        advance over multibyte UTF-8 characters in GlobalReplace.
1816    
1817    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1818        difference for some implementation of PCRE in some Windows version.
1819    
1820    29. Added some extra testing facilities to pcretest:
1821    
1822        \q<number>   in a data line sets the "match limit" value
1823        \Q<number>   in a data line sets the "match recursion limt" value
1824        -S <number>  sets the stack size, where <number> is in megabytes
1825    
1826        The -S option isn't available for Windows.
1827    
1828    
1829    Version 6.6 06-Feb-06
1830    ---------------------
1831    
1832     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1833        in pcreposix.h. I have copied the definition from pcre.h.
1834    
1835     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1836        because pcre.h is no longer a built file.
1837    
1838     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1839        not normally included in the compiled code.
1840    
1841    
1842    Version 6.5 01-Feb-06
1843    ---------------------
1844    
1845     1. When using the partial match feature with pcre_dfa_exec(), it was not
1846        anchoring the second and subsequent partial matches at the new starting
1847        point. This could lead to incorrect results. For example, with the pattern
1848        /1234/, partially matching against "123" and then "a4" gave a match.
1849    
1850     2. Changes to pcregrep:
1851    
1852        (a) All non-match returns from pcre_exec() were being treated as failures
1853            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1854            error message is output. Some extra information is given for the
1855            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1856            probably the only errors that are likely to be caused by users (by
1857            specifying a regex that has nested indefinite repeats, for instance).
1858            If there are more than 20 of these errors, pcregrep is abandoned.
1859    
1860        (b) A binary zero was treated as data while matching, but terminated the
1861            output line if it was written out. This has been fixed: binary zeroes
1862            are now no different to any other data bytes.
1863    
1864        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1865            used to set a locale for matching. The --locale=xxxx long option has
1866            been added (no short equivalent) to specify a locale explicitly on the
1867            pcregrep command, overriding the environment variables.
1868    
1869        (d) When -B was used with -n, some line numbers in the output were one less
1870            than they should have been.
1871    
1872        (e) Added the -o (--only-matching) option.
1873    
1874        (f) If -A or -C was used with -c (count only), some lines of context were
1875            accidentally printed for the final match.
1876    
1877        (g) Added the -H (--with-filename) option.
1878    
1879        (h) The combination of options -rh failed to suppress file names for files
1880            that were found from directory arguments.
1881    
1882        (i) Added the -D (--devices) and -d (--directories) options.
1883    
1884        (j) Added the -F (--fixed-strings) option.
1885    
1886        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1887    
1888        (l) Added the --colo(u)r option.
1889    
1890        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1891            is not present by default.
1892    
1893     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1894        items such as (?R) or (?1), when the recursion could match a number of
1895        alternatives. If it matched one of the alternatives, but subsequently,
1896        outside the recursion, there was a failure, the code tried to back up into
1897        the recursion. However, because of the way PCRE is implemented, this is not
1898        possible, and the result was an incorrect result from the match.
1899    
1900        In order to prevent this happening, the specification of recursion has
1901        been changed so that all such subpatterns are automatically treated as
1902        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1903    
1904     4. I had overlooked the fact that, in some locales, there are characters for
1905        which isalpha() is true but neither isupper() nor islower() are true. In
1906        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1907        and ordfeminine) are like this. This affected the treatment of \w and \W
1908        when they appeared in character classes, but not when they appeared outside
1909        a character class. The bit map for "word" characters is now created
1910        separately from the results of isalnum() instead of just taking it from the
1911        upper, lower, and digit maps. (Plus the underscore character, of course.)
1912    
1913     5. The above bug also affected the handling of POSIX character classes such as
1914        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1915        permanent tables. Instead, the bit maps for such a class were previously
1916        created as the appropriate unions of the upper, lower, and digit bitmaps.
1917        Now they are created by subtraction from the [[:word:]] class, which has
1918        its own bitmap.
1919    
1920     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1921        It is created by subtracting the vertical space characters (\x09, \x0a,
1922        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1923        subtraction was done in the overall bitmap for a character class, meaning
1924        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1925        be recognized. This bug has been fixed.
1926    
1927     7. Patches from the folks at Google:
1928    
1929          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1930          real life, but is still worth protecting against".
1931    
1932          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1933          regular expressions".
1934    
1935          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1936          have it.
1937    
1938          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1939          "configure" and the latter not, in order to fix a problem somebody had
1940          with compiling the Arg class on HP-UX.
1941    
1942          (e) Improve the error-handling of the C++ wrapper a little bit.
1943    
1944          (f) New tests for checking recursion limiting.
1945    
1946     8. The pcre_memmove() function, which is used only if the environment does not
1947        have a standard memmove() function (and is therefore rarely compiled),
1948        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1949        returning a result (though PCRE never actually uses the result).
1950    
1951     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1952        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1953        returned instead of calling malloc() with an overflowing number that would
1954        most likely cause subsequent chaos.
1955    
1956    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1957    
1958    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1959        with this option is matched, the nmatch and pmatch options of regexec() are
1960        ignored.
1961    
1962    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1963        provided in case anyone wants to the the POSIX interface with UTF-8
1964        strings.
1965    
1966    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1967        C++ linking (needed for some HP-UX environments).
1968    
1969    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1970        (unused parameter) and in the pcre_printint() function (omitted "default"
1971        switch label when the default is to do nothing).
1972    
1973    15. Added some code to make it possible, when PCRE is compiled as a C++
1974        library, to replace subject pointers for pcre_exec() with a smart pointer
1975        class, thus making it possible to process discontinuous strings.
1976    
1977    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1978        much the same function. They were added by different people who were trying
1979        to make PCRE easy to compile on non-Unix systems. It has been suggested
1980        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1981        for compiling on Windows systems. I have therefore replaced it with
1982        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1983        defaults to "extern" for C or "extern C" for C++, which works fine on
1984        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1985        SCOPE with something explicit in config.h. In addition:
1986    
1987        (a) pcreposix.h still had just "extern" instead of either of these macros;
1988            I have replaced it with PCRE_DATA_SCOPE.
1989    
1990        (b) Functions such as _pcre_xclass(), which are internal to the library,
1991            but external in the C sense, all had PCRE_EXPORT in their definitions.
1992            This is apparently wrong for the Windows case, so I have removed it.
1993            (It makes no difference on Unix-like systems.)
1994    
1995    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1996        of recursive calls to match(). This is different to MATCH_LIMIT because
1997        that limits the total number of calls to match(), not all of which increase
1998        the depth of recursion. Limiting the recursion depth limits the amount of
1999        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2000        when PCRE is compiled, and changed at run time. A patch from Google adds
2001        this functionality to the C++ interface.
2002    
2003    18. Changes to the handling of Unicode character properties:
2004    
2005        (a) Updated the table to Unicode 4.1.0.
2006    
2007        (b) Recognize characters that are not in the table as "Cn" (undefined).
2008    
2009        (c) I revised the way the table is implemented to a much improved format
2010            which includes recognition of ranges. It now supports the ranges that
2011            are defined in UnicodeData.txt, and it also amalgamates other
2012            characters into ranges. This has reduced the number of entries in the
2013            table from around 16,000 to around 3,000, thus reducing its size
2014            considerably. I realized I did not need to use a tree structure after
2015            all - a binary chop search is just as efficient. Having reduced the
2016            number of entries, I extended their size from 6 bytes to 8 bytes to
2017            allow for more data.
2018    
2019        (d) Added support for Unicode script names via properties such as \p{Han}.
2020    
2021    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2022        matching that character.
2023    
2024    20. When matching a repeated Unicode property with a minimum greater than zero,
2025        (for example \pL{2,}), PCRE could look past the end of the subject if it
2026        reached it while seeking the minimum number of characters. This could
2027        happen only if some of the characters were more than one byte long, because
2028        there is a check for at least the minimum number of bytes.
2029    
2030    21. Refactored the implementation of \p and \P so as to be more general, to
2031        allow for more different types of property in future. This has changed the
2032        compiled form incompatibly. Anybody with saved compiled patterns that use
2033        \p or \P will have to recompile them.
2034    
2035    22. Added "Any" and "L&" to the supported property types.
2036    
2037    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2038        but give a compile time error if the value is greater than 0xff.
2039    
2040    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2041        accidentally not being installed or uninstalled.
2042    
2043    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2044        made were to insert the current release number. This seemed silly, because
2045        it made things harder for people building PCRE on systems that don't run
2046        "configure". I have turned pcre.h into a distributed file, no longer built
2047        by "configure", with the version identification directly included. There is
2048        no longer a pcre.h.in file.
2049    
2050        However, this change necessitated a change to the pcre-config script as
2051        well. It is built from pcre-config.in, and one of the substitutions was the
2052        release number. I have updated configure.ac so that ./configure now finds
2053        the release number by grepping pcre.h.
2054    
2055    26. Added the ability to run the tests under valgrind.
2056    
2057    
2058    Version 6.4 05-Sep-05
2059    ---------------------
2060    
2061     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2062        "--" to be printed when multiple files were scanned, even when none of the
2063        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2064        consider it to be a bug, and have restored the previous behaviour.
2065    
2066     2. A couple of code tidies to get rid of compiler warnings.
2067    
2068     3. The pcretest program used to cheat by referring to symbols in the library
2069        whose names begin with _pcre_. These are internal symbols that are not
2070        really supposed to be visible externally, and in some environments it is
2071        possible to suppress them. The cheating is now confined to including
2072        certain files from the library's source, which is a bit cleaner.
2073    
2074     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2075        file's purpose clearer.
2076    
2077     5. Reorganized pcre_ucp_findchar().
2078    
2079    
2080    Version 6.3 15-Aug-05
2081    ---------------------
2082    
2083     1. The file libpcre.pc.in did not have general read permission in the tarball.
2084    
2085     2. There were some problems when building without C++ support:
2086    
2087        (a) If C++ support was not built, "make install" and "make test" still
2088            tried to test it.
2089    
2090        (b) There were problems when the value of CXX was explicitly set. Some
2091            changes have been made to try to fix these, and ...
2092    
2093        (c) --disable-cpp can now be used to explicitly disable C++ support.
2094    
2095        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2096            backslash in a target when C++ was disabled. This confuses some
2097            versions of "make", apparently. Using an intermediate variable solves
2098            this. (Same for CPP_LOBJ.)
2099    
2100     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2101        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2102        necessary on certain architectures.
2103    
2104     4. Added a setting of -export-symbols-regex to the link command to remove
2105        those symbols that are exported in the C sense, but actually are local
2106        within the library, and not documented. Their names all begin with
2107        "_pcre_". This is not a perfect job, because (a) we have to except some
2108        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2109        available (and never for static libraries). I have made a note to try to
2110        find a way round (a) in the future.
2111    
2112    
2113    Version 6.2 01-Aug-05
2114    ---------------------
2115    
2116     1. There was no test for integer overflow of quantifier values. A construction
2117        such as {1111111111111111} would give undefined results. What is worse, if
2118        a minimum quantifier for a parenthesized subpattern overflowed and became
2119        negative, the calculation of the memory size went wrong. This could have
2120        led to memory overwriting.
2121    
2122     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2123    
2124     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2125        operating environments where this matters.
2126    
2127     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2128        PCRE options from within the C++ wrapper.
2129    
2130     5. Named capturing subpatterns were not being correctly counted when a pattern
2131        was compiled. This caused two problems: (a) If there were more than 100
2132        such subpatterns, the calculation of the memory needed for the whole
2133        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2134        back references of the form \12, where the number was greater than 9, were
2135        not recognized as back references, even though there were sufficient
2136        previous subpatterns.
2137    
2138     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2139        versions of gcc, e.g. 2.95.4.
2140    
2141    
2142    Version 6.1 21-Jun-05
2143    ---------------------
2144    
2145     1. There was one reference to the variable "posix" in pcretest.c that was not
2146        surrounded by "#if !defined NOPOSIX".
2147    
2148     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2149        the cross-check on the old pcre_info() function, for the benefit of the
2150        cut-down version of PCRE that is currently imported into Exim.
2151    
2152     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2153        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2154        patterns that start (?i) but I don't think that matters. The use of (?i) is
2155        just an example; this all applies to the other options as well.
2156    
2157     4. Since libtool seems to echo the compile commands it is issuing, the output
2158        from "make" can be reduced a bit by putting "@" in front of each libtool
2159        compile command.
2160    
2161     5. Patch from the folks at Google for configure.in to be a bit more thorough
2162        in checking for a suitable C++ installation before trying to compile the
2163        C++ stuff. This should fix a reported problem when a compiler was present,
2164        but no suitable headers.
2165    
2166     6. The man pages all had just "PCRE" as their title. I have changed them to
2167        be the relevant file name. I have also arranged that these names are
2168        retained in the file doc/pcre.txt, which is a concatenation in text format
2169        of all the man pages except the little individual ones for each function.
2170    
2171     7. The NON-UNIX-USE file had not been updated for the different set of source
2172        files that come with release 6. I also added a few comments about the C++
2173        wrapper.
2174    
2175    
2176    Version 6.0 07-Jun-05
2177    ---------------------
2178    
2179     1. Some minor internal re-organization to help with my DFA experiments.
2180    
2181     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2182        didn't matter for the library itself when fully configured, but did matter
2183        when compiling without UCP support, or within Exim, where the ucp files are
2184        not imported.
2185    
2186     3. Refactoring of the library code to split up the various functions into
2187        different source modules. The addition of the new DFA matching code (see
2188        below) to a single monolithic source would have made it really too
2189        unwieldy, quite apart from causing all the code to be include in a
2190        statically linked application, when only some functions are used. This is
2191        relevant even without the DFA addition now that patterns can be compiled in
2192        one application and matched in another.
2193    
2194        The downside of splitting up is that there have to be some external
2195        functions and data tables that are used internally in different modules of
2196        the library but which are not part of the API. These have all had their
2197        names changed to start with "_pcre_" so that they are unlikely to clash
2198        with other external names.
2199    
2200     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2201        a different (DFA) algorithm. Although it is slower than the original
2202        function, it does have some advantages for certain types of matching
2203        problem.
2204    
2205     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2206        including restarting after a partial match.
2207    
2208     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2209        defined when compiling for Windows was sent to me. I have put it into the
2210        code, though I have no means of testing or verifying it.
2211    
2212     7. Added the pcre_refcount() auxiliary function.
2213    
2214     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2215        match before or at the first newline in the subject string. In pcretest,
2216        the /f option on a pattern can be used to set this.
2217    
2218     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2219        would behave wrongly. This has been present in PCRE since release 4.0.
2220    
2221    10. A number of changes to the pcregrep command:
2222    
2223        (a) Refactored how -x works; insert ^(...)$ instead of setting
2224            PCRE_ANCHORED and checking the length, in preparation for adding
2225            something similar for -w.
2226    
2227        (b) Added the -w (match as a word) option.
2228    
2229        (c) Refactored the way lines are read and buffered so as to have more
2230            than one at a time available.
2231    
2232        (d) Implemented a pcregrep test script.
2233    
2234        (e) Added the -M (multiline match) option. This allows patterns to match
2235            over several lines of the subject. The buffering ensures that at least
2236            8K, or the rest of the document (whichever is the shorter) is available
2237            for matching (and similarly the previous 8K for lookbehind assertions).
2238    
2239        (f) Changed the --help output so that it now says
2240    
2241              -w, --word-regex(p)
2242    
2243            instead of two lines, one with "regex" and the other with "regexp"
2244            because that confused at least one person since the short forms are the
2245            same. (This required a bit of code, as the output is generated
2246            automatically from a table. It wasn't just a text change.)
2247    
2248        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2249            option but starts with a hyphen. Could be a pattern or a path name
2250            starting with a hyphen, for instance.
2251    
2252        (h) "-" can be given as a file name to represent stdin.
2253    
2254        (i) When file names are being printed, "(standard input)" is used for
2255            the standard input, for compatibility with GNU grep. Previously
2256            "<stdin>" was used.
2257    
2258        (j) The option --label=xxx can be used to supply a name to be used for
2259            stdin when file names are being printed. There is no short form.
2260    
2261        (k) Re-factored the options decoding logic because we are going to add
2262            two more options that take data. Such options can now be given in four
2263            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2264    
2265        (l) Added the -A, -B, and -C options for requesting that lines of context
2266            around matches be printed.
2267    
2268        (m) Added the -L option to print the names of files that do not contain
2269            any matching lines, that is, the complement of -l.
2270    
2271        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2272            continue to scan other files.
2273    
2274        (o) The -s option was incorrectly implemented. For compatibility with other
2275            greps, it now suppresses the error message for a non-existent or non-
2276            accessible file (but not the return code). There is a new option called
2277            -q that suppresses the output of matching lines, which was what -s was
2278            previously doing.
2279    
2280        (p) Added --include and --exclude options to specify files for inclusion
2281            and exclusion when recursing.
2282    
2283    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2284        Hopefully, it now does.
2285    
2286    12. Missing cast in pcre_study().
2287    
2288    13. Added an "uninstall" target to the makefile.
2289    
2290    14. Replaced "extern" in the function prototypes in Makefile.in with
2291        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2292        world, but is set differently for Windows.
2293    
2294    15. Added a second compiling function called pcre_compile2(). The only
2295        difference is that it has an extra argument, which is a pointer to an
2296        integer error code. When there is a compile-time failure, this is set
2297        non-zero, in addition to the error test pointer being set to point to an
2298        error message. The new argument may be NULL if no error number is required
2299        (but then you may as well call pcre_compile(), which is now just a
2300        wrapper). This facility is provided because some applications need a
2301        numeric error indication, but it has also enabled me to tidy up the way
2302        compile-time errors are handled in the POSIX wrapper.
2303    
2304    16. Added VPATH=.libs to the makefile; this should help when building with one
2305        prefix path and installing with another. (Or so I'm told by someone who
2306        knows more about this stuff than I do.)
2307    
2308    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2309        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2310        match everything, including newlines. This is not POSIX-compatible, but
2311        somebody wanted the feature. From pcretest it can be activated by using
2312        both the P and the s flags.
2313    
2314    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2315    
2316    19. libpcre.pc was being incorrectly installed as executable.
2317    
2318    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2319        it now also looks for '\r' so that it will work unmodified on Windows.
2320    
2321    21. Added Google's contributed C++ wrapper to the distribution.
2322    
2323    22. Added some untidy missing memory free() calls in pcretest, to keep
2324        Electric Fence happy when testing.
2325    
2326    
2327    
2328    Version 5.0 13-Sep-04
2329    ---------------------
2330    
2331     1. Internal change: literal characters are no longer packed up into items
2332        containing multiple characters in a single byte-string. Each character
2333        is now matched using a separate opcode. However, there may be more than one
2334        byte in the character in UTF-8 mode.
2335    
2336     2. The pcre_callout_block structure has two new fields: pattern_position and
2337        next_item_length. These contain the offset in the pattern to the next match
2338        item, and its length, respectively.
2339    
2340     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2341        insertion of callouts before each pattern item. Added the /C option to
2342        pcretest to make use of this.
2343    
2344     4. On the advice of a Windows user, the lines
2345    
2346          #if defined(_WIN32) || defined(WIN32)
2347          _setmode( _fileno( stdout ), 0x8000 );
2348          #endif  /* defined(_WIN32) || defined(WIN32) */
2349    
2350        have been added to the source of pcretest. This apparently does useful
2351        magic in relation to line terminators.
2352    
2353     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2354        for the benefit of those environments where the "b" makes a difference.
2355    
2356     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2357        to know about it. I have put a hack into configure.in that adds in code
2358        to set GCC=yes if CC=icc. This seems to end up at a point in the
2359        generated configure script that is early enough to affect the setting of
2360        compiler options, which is what is needed, but I have no means of testing
2361        whether it really works. (The user who reported this had patched the
2362        generated configure script, which of course I cannot do.)
2363    
2364        LATER: After change 22 below (new libtool files), the configure script
2365        seems to know about icc (and also ecc). Therefore, I have commented out
2366        this hack in configure.in.
2367    
2368     7. Added support for pkg-config (2 patches were sent in).
2369    
2370     8. Negated POSIX character classes that used a combination of internal tables
2371        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2372        [[:^ascii]]. Typically, they would match almost any characters. The other
2373        POSIX classes were not broken in this way.
2374    
2375     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2376        to find the match, as PCRE was deluded into thinking that the match had to
2377        start at the start point or following a newline. The same bug applied to
2378        patterns with negative forward assertions or any backward assertions
2379        preceding ".*" at the start, unless the pattern required a fixed first
2380        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2381    
2382    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2383        starting at the last subject character, bytes beyond the end of the subject
2384        string were read.
2385    
2386    11. Renamed the variable "class" as "classbits" to make life easier for C++
2387        users. (Previously there was a macro definition, but it apparently wasn't
2388        enough.)
2389    
2390    12. Added the new field "tables" to the extra data so that tables can be passed
2391        in at exec time, or the internal tables can be re-selected. This allows
2392        a compiled regex to be saved and re-used at a later time by a different
2393        program that might have everything at different addresses.
2394    
2395    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2396        -R library as well as a -L library.
2397    
2398    14. The debugging options of pcretest (-d on the command line or D on a
2399        pattern) showed incorrect output for anything following an extended class
2400        that contained multibyte characters and which was followed by a quantifier.
2401    
2402    15. Added optional support for general category Unicode character properties
2403        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2404        support. It adds about 90K to the size of the library. The meanings of the
2405        inbuilt class escapes such as \d and \s have NOT been changed.
2406    
2407    16. Updated pcredemo.c to include calls to free() to release the memory for the
2408        compiled pattern.
2409    
2410    17. The generated file chartables.c was being created in the source directory
2411        instead of in the building directory. This caused the build to fail if the
2412        source directory was different from the building directory, and was
2413        read-only.
2414    
2415    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2416        file. No doubt somebody will tell me if they don't make sense... Also added
2417        Dan Mooney's comments about building on OpenVMS.
2418    
2419    19. Added support for partial matching via the PCRE_PARTIAL option for
2420        pcre_exec() and the \P data escape in pcretest.
2421    
2422    20. Extended pcretest with 3 new pattern features:
2423    
2424        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2425              write the compiled pattern to the file whose name is "rest-of-line".
2426              This is a straight binary dump of the data, with the saved pointer to
2427              the character tables forced to be NULL. The study data, if any, is
2428              written too. After writing, pcretest reads a new pattern.
2429    
2430        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2431              compiled pattern from the given file. There must not be any
2432              occurrences of "<" in the file name (pretty unlikely); if there are,
2433              pcretest will instead treat the initial "<" as a pattern delimiter.
2434              After reading in the pattern, pcretest goes on to read data lines as
2435              usual.
2436    
2437        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2438              and 16-bit fields in a compiled pattern, to simulate a pattern that
2439              was compiled on a host of opposite endianness.
2440    
2441    21. The pcre-exec() function can now cope with patterns that were compiled on
2442        hosts of opposite endianness, with this restriction:
2443    
2444          As for any compiled expression that is saved and used later, the tables
2445          pointer field cannot be preserved; the extra_data field in the arguments
2446          to pcre_exec() should be used to pass in a tables address if a value
2447          other than the default internal tables were used at compile time.
2448    
2449    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2450        now diagnosed as an error. Previously, most of the time, a negative number
2451        would have been treated as zero, but if in addition "ovector" was passed as
2452        NULL, a crash could occur.
2453    
2454    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2455        new versions from the libtool 1.5 distribution (the last one is a copy of
2456        a file called libtool.m4). This seems to have fixed the need to patch
2457        "configure" to support Darwin 1.3 (which I used to do). However, I still
2458        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2459        workstation).
2460    
2461    24. Changed the PCRE licence to be the more standard "BSD" licence.
2462    
2463    
2464    Version 4.5 01-Dec-03
2465    ---------------------
2466    
2467     1. There has been some re-arrangement of the code for the match() function so
2468        that it can be compiled in a version that does not call itself recursively.
2469        Instead, it keeps those local variables that need separate instances for
2470        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2471        needs to "recurse". Keeping track of where control must go is done by means
2472        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2473        hide most of the details from the main code, and operates only if
2474        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2475        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2476        operating.
2477    
2478        To make it easier for callers to provide specially tailored get/free
2479        functions for this usage, two new functions, pcre_stack_malloc, and
2480        pcre_stack_free, are used. They are always called in strict stacking order,
2481        and the size of block requested is always the same.
2482    
2483        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2484        PCRE has been compiled to use the stack or the heap for recursion. The
2485        -C option of pcretest uses this to show which version is compiled.
2486    
2487        A new data escape \S, is added to pcretest; it causes the amounts of store
2488        obtained and freed by both kinds of malloc/free at match time to be added
2489        to the output.
2490    
2491     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2492        what's available on my current Linux desktop machine.
2493    
2494     3. When matching a UTF-8 string, the test for a valid string at the start has
2495        been extended. If start_offset is not zero, PCRE now checks that it points
2496        to a byte that is the start of a UTF-8 character. If not, it returns
2497        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2498        this is necessary because there may be backward assertions in the pattern.
2499        When matching the same subject several times, it may save resources to use
2500        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2501    
2502     4. The code for checking the validity of UTF-8 strings has been tightened so
2503        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2504        containing "overlong sequences".
2505    
2506     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2507        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2508        should have been "&", but it just so happened that all the cases this let
2509        through by mistake were picked up later in the function.
2510    
2511     6. I had used a variable called "isblank" - this is a C99 function, causing
2512        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2513    
2514     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2515        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2516        so the version is shown; (c) stop "make test" echoing "./RunTest".
2517    
2518     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2519    
2520     9. The prototype for memmove() for systems that don't have it was using
2521        size_t, but the inclusion of the header that defines size_t was later. I've
2522        moved the #includes for the C headers earlier to avoid this.
2523    
2524    10. Added some adjustments to the code to make it easier to compiler on certain
2525        special systems:
2526    
2527          (a) Some "const" qualifiers were missing.
2528          (b) Added the macro EXPORT before all exported functions; by default this
2529              is defined to be empty.
2530          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2531              that it reads its output file name as an argument instead of writing
2532              to the standard output and assuming this can be redirected.
2533    
2534    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2535        class containing characters with values greater than 255, PCRE compilation
2536        went into a loop.
2537    
2538    12. A recursive reference to a subpattern that was within another subpattern
2539        that had a minimum quantifier of zero caused PCRE to crash. For example,
2540        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2541        recursion. If the recursively-called subpattern itself had a zero repeat,
2542        that was OK.
2543    
2544    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2545        buffer into which it was copied (for escape processing) was still set at
2546        1024, so long lines caused crashes.
2547    
2548    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2549        "internal error: code overflow...". This applied to any character class
2550        that was followed by a possessive quantifier.
2551    
2552    15. Modified the Makefile to add libpcre.la as a prerequisite for
2553        libpcreposix.la because I was told this is needed for a parallel build to
2554        work.
2555    
2556    16. If a pattern that contained .* following optional items at the start was
2557        studied, the wrong optimizing data was generated, leading to matching
2558        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2559        matching string must start with a or b or c. The correct conclusion for
2560        this pattern is that a match can start with any character.
2561    
2562    
2563    Version 4.4 13-Aug-03
2564    ---------------------
2565    
2566     1. In UTF-8 mode, a character class containing characters with values between
2567        127 and 255 was not handled correctly if the compiled pattern was studied.
2568        In fixing this, I have also improved the studying algorithm for such
2569        classes (slightly).
2570    
2571     2. Three internal functions had redundant arguments passed to them. Removal
2572        might give a very teeny performance improvement.
2573    
2574     3. Documentation bug: the value of the capture_top field in a callout is *one
2575        more than* the number of the hightest numbered captured substring.
2576    
2577     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2578        in incorrectly linking with a previously installed version. They now link
2579        explicitly with libpcre.la.
2580    
2581     5. configure.in no longer needs to recognize Cygwin specially.
2582    
2583     6. A problem in pcre.in for Windows platforms is fixed.
2584    
2585     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2586        pcretest, it used to include the size of the study block as part of its
2587        output. Unfortunately, the structure contains a field that has a different
2588        size on different hardware architectures. This meant that the tests that
2589        showed this size failed. As the block is currently always of a fixed size,
2590        this information isn't actually particularly useful in pcretest output, so
2591        I have just removed it.
2592    
2593     8. Three pre-processor statements accidentally did not start in column 1.
2594        Sadly, there are *still* compilers around that complain, even though
2595        standard C has not required this for well over a decade. Sigh.
2596    
2597     9. In pcretest, the code for checking callouts passed small integers in the
2598        callout_data field, which is a void * field. However, some picky compilers
2599        complained about the casts involved for this on 64-bit systems. Now
2600        pcretest passes the address of the small integer instead, which should get
2601        rid of the warnings.
2602    
2603    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2604        both compile and run time, and gives an error if an invalid UTF-8 sequence
2605        is found. There is a option for disabling this check in cases where the
2606        string is known to be correct and/or the maximum performance is wanted.
2607    
2608    11. In response to a bug report, I changed one line in Makefile.in from
2609    
2610            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2611        to
2612            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2613    
2614        to look similar to other lines, but I have no way of telling whether this
2615        is the right thing to do, as I do not use Windows. No doubt I'll get told
2616        if it's wrong...
2617    
2618    
2619    Version 4.3 21-May-03
2620    ---------------------
2621    
2622    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2623       Makefile.
2624    
2625    2. Some refactoring to improve the quality of the code:
2626    
2627       (i)   The utf8_table... variables are now declared "const".
2628    
2629       (ii)  The code for \cx, which used the "case flipping" table to upper case
2630             lower case letters, now just substracts 32. This is ASCII-specific,
2631             but the whole concept of \cx is ASCII-specific, so it seems
2632             reasonable.
2633    
2634       (iii) PCRE was using its character types table to recognize decimal and
2635             hexadecimal digits in the pattern. This is silly, because it handles
2636             only 0-9, a-f, and A-F, but the character types table is locale-
2637             specific, which means strange things might happen. A private
2638             table is now used for this - though it costs 256 bytes, a table is
2639             much faster than multiple explicit tests. Of course, the standard
2640             character types table is still used for matching digits in subject
2641             strings against \d.
2642    
2643       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2644             ending in _t are). So I've renamed it as ESC_tee.
2645    
2646    3. The first argument for regexec() in the POSIX wrapper should have been
2647       defined as "const".
2648    
2649    4. Changed pcretest to use malloc() for its buffers so that they can be
2650       Electric Fenced for debugging.
2651    
2652    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2653       to read one or more bytes before the start of the subject string. Often this
2654       had no effect on PCRE's behaviour, but in some circumstances it could
2655       provoke a segmentation fault.
2656    
2657    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2658       to try to read one or more bytes before the start of the subject string.
2659    
2660    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2661       UTF-8 support could misbehave in various ways if the subject string
2662       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2663       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2664       back over UTF-8 characters.)
2665    
2666    
2667    Version 4.2 14-Apr-03
2668    ---------------------
2669    
2670    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2671    
2672    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2673         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2674         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2675         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2676         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2677           and BUILD_EXEEXT
2678         Note: automatic setting of the BUILD variables is not yet working
2679         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2680           compile-time but not at link-time
2681         [LINK]: use for linking executables only
2682         make different versions for Windows and non-Windows
2683         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2684           libraries
2685         [LINK_FOR_BUILD]: new variable
2686         [OBJEXT]: use throughout
2687         [EXEEXT]: use throughout
2688         <winshared>: new target
2689         <wininstall>: new target
2690         <dftables.o>: use native compiler
2691         <dftables>: use native linker
2692         <install>: handle Windows platform correctly
2693         <clean>: ditto
2694         <check>: ditto
2695         copy DLL to top builddir before testing
2696    
2697       As part of these changes, -no-undefined was removed again. This was reported
2698       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2699       in any case.
2700    
2701    3. Some tidies to get rid of compiler warnings:
2702    
2703       . In the match_data structure, match_limit was an unsigned long int, whereas
2704         match_call_count was an int. I've made them both unsigned long ints.
2705    
2706       . In pcretest the fact that a const uschar * doesn't automatically cast to
2707         a void * provoked a warning.
2708    
2709       . Turning on some more compiler warnings threw up some "shadow" variables
2710         and a few more missing casts.
2711    
2712    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2713       option, a class that contained a single character with a value between 128
2714       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2715    
2716    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2717       option, a class that contained several characters, but with at least one
2718       whose value was between 128 and 255 caused PCRE to crash.
2719    
2720    
2721    Version 4.1 12-Mar-03
2722    ---------------------
2723    
2724    1. Compiling with gcc -pedantic found a couple of places where casts were
2725    needed, and a string in dftables.c that was longer than standard compilers are
2726    required to support.
2727    
2728    2. Compiling with Sun's compiler found a few more places where the code could
2729    be tidied up in order to avoid warnings.
2730    
2731    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2732    first of these names is deprecated in the latest Autoconf in favour of the name
2733    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2734    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2735    analogy I have changed it to CFLAGS_FOR_BUILD.
2736    
2737    4. Added -no-undefined to the linking command in the Makefile, because this is
2738    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2739    linking step for the pcreposix library.
2740    
2741    5. PCRE was failing to diagnose the case of two named groups with the same
2742    name.
2743    
2744    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2745    literal character that is needed in the subject for a match, and scans along to
2746    ensure that it is present before embarking on the full matching process. This
2747    saves time in cases of nested unlimited repeats that are never going to match.
2748    Problem: the scan can take a lot of time if the subject is very long (e.g.
2749    megabytes), thus penalizing straightforward matches. It is now done only if the
2750    amount of subject to be scanned is less than 1000 bytes.
2751    
2752    7. A lesser problem with the same optimization is that it was recording the
2753    first character of an anchored pattern as "needed", thus provoking a search
2754    right along the subject, even when the first match of the pattern was going to
2755    fail. The "needed" character is now not set for anchored patterns, unless it
2756    follows something in the pattern that is of non-fixed length. Thus, it still
2757    fulfils its original purpose of finding quick non-matches in cases of nested
2758    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2759    
2760    
2761    Version 4.0 17-Feb-03
2762    ---------------------
2763    
2764    1. If a comment in an extended regex that started immediately after a meta-item
2765    extended to the end of string, PCRE compiled incorrect data. This could lead to
2766    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2767    
2768    2. Moved to autoconf 2.53 and libtool 1.4.2.
2769    
2770    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2771    the special perltest8 script is no longer needed - all the tests can be run
2772    from a single perltest script.
2773    
2774    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2775    by \s. It has now been removed in PCRE. This means it isn't recognized as
2776    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2777    class [:space:] *does* include VT, thereby creating a mess.
2778    
2779    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2780    space and tab.
2781    
2782    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2783    its new features into the main test script, reducing the number of scripts.
2784    
2785    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2786    were backward compatible, and made the (?i) apply to the whole pattern, as if
2787    /i were given. Now it behaves more logically, and applies the option setting
2788    only to what follows. PCRE has been changed to follow suit. However, if it
2789    finds options settings right at the start of the pattern, it extracts them into
2790    the global options, as before. Thus, they show up in the info data.
2791    
2792    8. Added support for the \Q...\E escape sequence. Characters in between are
2793    treated as literals. This is slightly different from Perl in that $ and @ are
2794    also handled as literals inside the quotes. In Perl, they will cause variable
2795    interpolation. Note the following examples:
2796    
2797        Pattern            PCRE matches      Perl matches
2798    
2799        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2800        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2801        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2802    
2803    For compatibility with Perl, \Q...\E sequences are recognized inside character
2804    classes as well as outside them.
2805    
2806    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2807    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2808    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2809    signed/unsigned warnings.
2810    
2811    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2812    option for pcretest, so I've replaced it by a simple function that does just
2813    that job.
2814    
2815    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2816    "pcregrep -".
2817    
2818    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2819    Java package. This provides some syntactic sugar for simple cases of what my
2820    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2821    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2822    item, you can use this simplified notation. Note that only makes sense with
2823    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2824    greediness, whatever the setting of the PCRE_UNGREEDY option.
2825    
2826    13. A change of greediness default within a pattern was not taking effect at
2827    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2828    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2829    was abstracted outside.
2830    
2831    14. PCRE now supports the \G assertion. It is true when the current matching
2832    position is at the start point of the match. This differs from \A when the
2833    starting offset is non-zero. Used with the /g option of pcretest (or similar
2834    code), it works in the same way as it does for Perl's /g option. If all
2835    alternatives of a regex begin with \G, the expression is anchored to the start
2836    match position, and the "anchored" flag is set in the compiled expression.
2837    
2838    15. Some bugs concerning the handling of certain option changes within patterns
2839    have been fixed. These applied to options other than (?ims). For example,
2840    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2841    been the other way round. Some of this was related to change 7 above.
2842    
2843    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2844    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2845    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2846    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2847    
2848    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2849    mode. Unlike ".", it always matches newline, whatever the setting of
2850    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2851    assertions. Perl allows it, but it doesn't (in general) work because it can't
2852    calculate the length of the lookbehind. At least, that's the case for Perl
2853    5.8.0 - I've been told they are going to document that it doesn't work in
2854    future.
2855    
2856    18. Added an error diagnosis for escapes that PCRE does not support: these are
2857    \L, \l, \N, \P, \p, \U, \u, and \X.
2858    
2859    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2860    reading past the end of the pattern in cases such as /[abcd/.
2861    
2862    20. PCRE was getting more memory than necessary for patterns with classes that
2863    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2864    
2865    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2866    compiling PCRE for use with Virtual Pascal.
2867    
2868    22. Small fix to the Makefile to make it work properly if the build is done
2869    outside the source tree.
2870    
2871    23. Added a new extension: a condition to go with recursion. If a conditional
2872    subpattern starts with (?(R) the "true" branch is used if recursion has
2873    happened, whereas the "false" branch is used only at the top level.
2874    
2875    24. When there was a very long string of literal characters (over 255 bytes
2876    without UTF support, over 250 bytes with UTF support), the computation of how
2877    much memory was required could be incorrect, leading to segfaults or other
2878    strange effects.
2879    
2880    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2881    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2882    there was a subsequent back reference to those brackets. This meant that, for
2883    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2884    possible to check for precisely this case. All we can do is abandon the
2885    optimization if .* occurs inside capturing brackets when there are any back
2886    references whatsoever. (See below for a better fix that came later.)
2887    
2888    26. The handling of the optimization for finding the first character of a
2889    non-anchored pattern, and for finding a character that is required later in the
2890    match were failing in some cases. This didn't break the matching; it just
2891    failed to optimize when it could. The way this is done has been re-implemented.
2892    
2893    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2894    
2895    28. Added a new feature that provides some of the functionality that Perl
2896    provides with (?{...}). The facility is termed a "callout". The way it is done
2897    in PCRE is for the caller to provide an optional function, by setting
2898    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2899    global variable. By default it is unset, which disables all calling out. To get
2900    the function called, the regex must include (?C) at appropriate points. This
2901    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2902    This provides a means of identifying different callout points. When PCRE
2903    reaches such a point in the regex, if pcre_callout has been set, the external
2904    function is called. It is provided with data in a structure called
2905    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2906    matching continues; if it returns a non-zero value, the match at the current
2907    point fails. However, backtracking will occur if possible. [This was changed
2908    later and other features added - see item 49 below.]
2909    
2910    29. pcretest is upgraded to test the callout functionality. It provides a
2911    callout function that displays information. By default, it shows the start of
2912    the match and the current position in the text. There are some new data escapes
2913    to vary what happens:
2914    
2915        \C+         in addition, show current contents of captured substrings
2916        \C-         do not supply a callout function
2917        \C!n        return 1 when callout number n is reached
2918        \C!n!m      return 1 when callout number n is reached for the mth time
2919    
2920    30. If pcregrep was called with the -l option and just a single file name, it
2921    output "<stdin>" if a match was found, instead of the file name.
2922    
2923    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2924    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2925    pcre_exec(). This saves a malloc/free per call. The default value of
2926    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2927    when configuring.
2928    
2929    32. The default maximum size of a compiled pattern is 64K. There have been a
2930    few cases of people hitting this limit. The code now uses macros to handle the
2931    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2932    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2933    configuring. Tests 2 and 5 work only with 2-byte links because they output
2934    debugging information about compiled patterns.
2935    
2936    33. Internal code re-arrangements:
2937    
2938    (a) Moved the debugging function for printing out a compiled regex into
2939        its own source file (printint.c) and used #include to pull it into
2940        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2941        separate copies.
2942    
2943    (b) Defined the list of op-code names for debugging as a macro in
2944        internal.h so that it is next to the definition of the opcodes.
2945    
2946    (c) Defined a table of op-code lengths for simpler skipping along compiled
2947        code. This is again a macro in internal.h so that it is next to the
2948        definition of the opcodes.
2949    
2950    34. Added support for recursive calls to individual subpatterns, along the
2951    lines of Robin Houston's patch (but implemented somewhat differently).
2952    
2953    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2954    allow it to read and process whole directories in Win32. This code was
2955    contributed by Lionel Fourquaux; it has not been tested by me.
2956    
2957    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2958    used to name a group. Names consist of alphanumerics and underscores, and must
2959    be unique. Back references use the syntax (?P=name) and recursive calls use
2960    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2961    numbers. The function pcre_fullinfo() can be used after compilation to extract
2962    a name/number map. There are three relevant calls:
2963    
2964      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2965      PCRE_INFO_NAMECOUNT            yields the number of entries
2966      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2967    
2968    The map is a vector of fixed-size entries. The size of each entry depends on
2969    the length of the longest name used. The first two bytes of each entry are the
2970    group number, most significant byte first. There follows the corresponding
2971    name, zero terminated. The names are in alphabetical order.
2972    
2973    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2974    case instead of 255. Making it the same both with and without UTF-8 support
2975    means that the same test output works with both.
2976    
2977    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2978    calling malloc() with a zero argument.
2979    
2980    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2981    optimization. I've improved things by keeping a bitmap of backreferences with
2982    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2983    fact referenced, the optimization can be applied. It is unlikely that a
2984    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2985    the match to follow \n) will appear inside brackets with a number greater than
2986    31, but if it does, any back reference > 31 suppresses the optimization.
2987    
2988    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2989    of disabling numbered capturing parentheses. Any opening parenthesis that is
2990    not followed by ? behaves as if it were followed by ?: but named parentheses
2991    can still be used for capturing (and they will acquire numbers in the usual
2992    way).
2993    
2994    41. Redesigned the return codes from the match() function into yes/no/error so
2995    that errors can be passed back from deep inside the nested calls. A malloc
2996    failure while inside a recursive subpattern call now causes the
2997    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2998    
2999    42. It is now possible to set a limit on the number of times the match()
3000    function is called in a call to pcre_exec(). This facility makes it possible to
3001    limit the amount of recursion and backtracking, though not in a directly
3002    obvious way, because the match() function is used in a number of different
3003    circumstances. The count starts from zero for each position in the subject
3004    string (for non-anchored patterns). The default limit is, for compatibility, a
3005    large number, namely 10 000 000. You can change this in two ways:
3006    
3007    (a) When configuring PCRE before making, you can use --with-match-limit=n
3008        to set a default value for the compiled library.
3009    
3010    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3011        a different value is set. See 45 below.
3012    
3013    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3014    
3015    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3016    of things that can be changed at compile time. The first argument specifies
3017    what is wanted and the second points to where the information is to be placed.
3018    The current list of available information is:
3019    
3020      PCRE_CONFIG_UTF8
3021    
3022    The output is an integer that is set to one if UTF-8 support is available;
3023    otherwise it is set to zero.
3024    
3025      PCRE_CONFIG_NEWLINE
3026    
3027    The output is an integer that it set to the value of the code that is used for
3028    newline. It is either LF (10) or CR (13).
3029    
3030      PCRE_CONFIG_LINK_SIZE
3031    
3032    The output is an integer that contains the number of bytes used for internal
3033    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3034    
3035      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3036    
3037    The output is an integer that contains the threshold above which the POSIX
3038    interface uses malloc() for output vectors. See item 31 above.
3039    
3040      PCRE_CONFIG_MATCH_LIMIT
3041    
3042    The output is an unsigned integer that contains the default limit of the number
3043    of match() calls in a pcre_exec() execution. See 42 above.
3044    
3045    44. pcretest has been upgraded by the addition of the -C option. This causes it
3046    to extract all the available output from the new pcre_config() function, and to
3047    output it. The program then exits immediately.
3048    
3049    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3050    order to support additional features. One way would have been to define
3051    pcre_exec2() (for example) with extra arguments, but this would not have been
3052    extensible, and would also have required all calls to the original function to
3053    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3054    is used for passing in "extra" data from pcre_study().
3055    
3056    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3057    contains the following fields:
3058    
3059      flags         a bitmap indicating which of the following fields are set
3060      study_data    opaque data from pcre_study()
3061      match_limit   a way of specifying a limit on match() calls for a specific
3062                      call to pcre_exec()
3063      callout_data  data for callouts (see 49 below)
3064    
3065    The flag bits are also defined in pcre.h, and are
3066    
3067      PCRE_EXTRA_STUDY_DATA
3068      PCRE_EXTRA_MATCH_LIMIT
3069      PCRE_EXTRA_CALLOUT_DATA
3070    
3071    The pcre_study() function now returns one of these new pcre_extra blocks, with
3072    the actual study data pointed to by the study_data field, and the
3073    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3074    before. That is, this change is entirely upwards-compatible and requires no
3075    change to existing code.
3076    
3077    If you want to pass in additional data to pcre_exec(), you can either place it
3078    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3079    block.
3080    
3081    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3082    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3083    times with different match limits, until it finds the minimum value needed for
3084    pcre_exec() to complete. The value is then output. This can be instructive; for
3085    most simple matches the number is quite small, but for pathological cases it
3086    gets very large very quickly.
3087    
3088    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3089    returns the size of the data block pointed to by the study_data field in a
3090    pcre_extra block, that is, the value that was passed as the argument to
3091    pcre_malloc() when PCRE was getting memory in which to place the information
3092    created by pcre_study(). The fourth argument should point to a size_t variable.
3093    pcretest has been extended so that this information is shown after a successful
3094    pcre_study() call when information about the compiled regex is being displayed.
3095    
3096    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3097    because what follows is always an absolute path. (Later: it turns out that this
3098    is more than cosmetic for MinGW, because it doesn't like empty path
3099    components.)
3100    
3101    49. Some changes have been made to the callout feature (see 28 above):
3102    
3103    (i)  A callout function now has three choices for what it returns:
3104    
3105           0  =>  success, carry on matching
3106         > 0  =>  failure at this point, but backtrack if possible
3107         < 0  =>  serious error, return this value from pcre_exec()
3108    
3109         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3110         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3111         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3112         use by callout functions. It will never be used by PCRE itself.
3113    
3114    (ii) The pcre_extra structure (see 45 above) has a void * field called
3115         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3116         pcre_callout_block structure has a field of the same name. The contents of
3117         the field passed in the pcre_extra structure are passed to the callout
3118         function in the corresponding field in the callout block. This makes it
3119         easier to use the same callout-containing regex from multiple threads. For
3120         testing, the pcretest program has a new data escape
3121    
3122           \C*n        pass the number n (may be negative) as callout_data
3123    
3124         If the callout function in pcretest receives a non-zero value as
3125         callout_data, it returns that value.
3126    
3127    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3128    there were some redundant $(CFLAGS) in commands that are now specified as
3129    $(LINK), which already includes $(CFLAGS).
3130    
3131    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3132    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3133    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3134    one-byte characters throughout. Note that case-insensitive matching applies
3135    only to characters whose values are less than 256. PCRE doesn't support the
3136    notion of cases for higher-valued characters.
3137    
3138    (i)   A character class whose characters are all within 0-255 is handled as
3139          a bit map, and the map is inverted for negative classes. Previously, a
3140          character > 255 always failed to match such a class; however it should
3141          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3142    
3143    (ii)  A negated character class with a single character < 255 is coded as
3144          "not this character" (OP_NOT). This wasn't working properly when the test
3145          character was multibyte, either singly or repeated.
3146    
3147    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3148          mode, for example: \x{100}{2,3}.
3149    
3150    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3151          singly or repeated) now correctly test multibyte characters. However,
3152          PCRE doesn't recognize any characters with values greater than 255 as
3153          digits, spaces, or word characters. Such characters always match \D, \S,
3154          and \W, and never match \d, \s, or \w.
3155    
3156    (v)   Classes may now contain characters and character ranges with values
3157          greater than 255. For example: [ab\x{100}-\x{400}].
3158    
3159    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3160          PCRE in UTF-8 mode.
3161    
3162    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3163    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3164    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3165    value.)
3166    
3167    53. The single man page has become too large. I have therefore split it up into
3168    a number of separate man pages. These also give rise to individual HTML pages;
3169    these are now put in a separate directory, and there is an index.html page that
3170    lists them all. Some hyperlinking between the pages has been installed.
3171    
3172    54. Added convenience functions for handling named capturing parentheses.
3173    
3174    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3175    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3176    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3177    are faulted.
3178    
3179    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3180    calling configure. These values are used when compiling the dftables.c program
3181    which is run to generate the source of the default character tables. They
3182    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3183    you will need to set these values.
3184    
3185    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3186    
3187    
3188    Version 3.9 02-Jan-02
3189    ---------------------
3190    
3191    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3192    
3193    2. If --disable-static was given, the building process failed when trying to
3194    build pcretest and pcregrep. (For some reason it was using libtool to compile
3195    them, which is not right, as they aren't part of the library.)
3196    
3197    
3198    Version 3.8 18-Dec-01
3199    ---------------------
3200    
3201    1. The experimental UTF-8 code was completely screwed up. It was packing the
3202    bytes in the wrong order. How dumb can you get?
3203    
3204    
3205    Version 3.7 29-Oct-01
3206    ---------------------
3207    
3208    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3209    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3210    this didn't happen under Solaris 8, where I normally test things.
3211    
3212    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3213    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3214    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3215    this fix an hour or so after the initial 3.7 release.)
3216    
3217    
3218    Version 3.6 23-Oct-01
3219    ---------------------
3220    
3221    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3222    offsets passed as NULL with zero offset count.
3223    
3224    2. The config.guess and config.sub files had not been updated when I moved to
3225    the latest autoconf.
3226    
3227    
3228    Version 3.5 15-Aug-01
3229    ---------------------
3230    
3231    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3232    had been forgotten.
3233    
3234    2. By using declared but undefined structures, we can avoid using "void"
3235    definitions in pcre.h while keeping the internal definitions of the structures
3236    private.
3237    
3238    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3239    user point of view, this means that both static and shared libraries are built
3240    by default, but this can be individually controlled. More of the work of
3241    handling this static/shared cases is now inside libtool instead of PCRE's make
3242    file.
3243    
3244    4. The pcretest utility is now installed along with pcregrep because it is
3245    useful for users (to test regexs) and by doing this, it automatically gets
3246    relinked by libtool. The documentation has been turned into a man page, so
3247    there are now .1, .txt, and .html versions in /doc.
3248    
3249    5. Upgrades to pcregrep:
3250       (i)   Added long-form option names like gnu grep.
3251       (ii)  Added --help to list all options with an explanatory phrase.
3252       (iii) Added -r, --recursive to recurse into sub-directories.
3253       (iv)  Added -f, --file to read patterns from a file.
3254    
3255    6. pcre_exec() was referring to its "code" argument before testing that
3256    argument for NULL (and giving an error if it was NULL).
3257    
3258    7. Upgraded Makefile.in to allow for compiling in a different directory from
3259    the source directory.
3260    
3261    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3262    options bits, the pointer it was passed was to an int instead of to an unsigned
3263    long int. This mattered only on 64-bit systems.
3264    
3265    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3266    generated) instead of pcre.in, which it its source. Also made the same change
3267    in several of the .c files.
3268    
3269    10. A new release of gcc defines printf() as a macro, which broke pcretest
3270    because it had an ifdef in the middle of a string argument for printf(). Fixed
3271    by using separate calls to printf().
3272    
3273    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3274    script, to force use of CR or LF instead of \n in the source. On non-Unix
3275    systems, the value can be set in config.h.
3276    
3277    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3278    absolute limit. Changed the text of the error message to make this clear, and
3279    likewise updated the man page.
3280    
3281    13. The limit of 99 on the number of capturing subpatterns has been removed.
3282    The new limit is 65535, which I hope will not be a "real" limit.
3283    
3284    
3285    Version 3.4 22-Aug-00
3286    ---------------------
3287    
3288    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3289    
3290    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3291    
3292    
3293    Version 3.3 01-Aug-00
3294    ---------------------
3295    
3296    1. If an octal character was given, but the value was greater than \377, it
3297    was not getting masked to the least significant bits, as documented. This could
3298    lead to crashes in some systems.
3299    
3300    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3301    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3302    
3303    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3304    These just pass their arguments on to (pcre_free)(), but they are provided
3305    because some uses of PCRE bind it to non-C systems that can call its functions,
3306    but cannot call free() or pcre_free() directly.
3307    
3308    4. Add "make test" as a synonym for "make check". Corrected some comments in
3309    the Makefile.
3310    
3311    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3312    Makefile.
3313    
3314    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3315    command called pgrep for grepping around the active processes.
3316    
3317    7. Added the beginnings of support for UTF-8 character strings.
3318    
3319    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3320    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3321    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3322    out for the ar command.)
3323    
3324    
3325    Version 3.2 12-May-00
3326    ---------------------
3327    
3328    This is purely a bug fixing release.
3329    
3330    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3331    of ZA. This was just one example of several cases that could provoke this bug,
3332    which was introduced by change 9 of version 2.00. The code for breaking
3333    infinite loops after an iteration that matches an empty string was't working
3334    correctly.
3335    
3336    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3337    when matched against abbab (for example). After matching an empty string, it
3338    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3339    caused it to match further down the string than it should.
3340    
3341    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3342    was there because it doesn't seem to be needed, and it causes trouble on some
3343    systems, as it is not a Standard C header. It has been removed.
3344    
3345    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3346    were reported on the Macintosh. The changes were from
3347    
3348      while ((c = *(++ptr)) != 0 && c != '\n');
3349    to
3350      while ((c = *(++ptr)) != 0 && c != '\n') ;
3351    
3352    Totally extraordinary, but if that's what it takes...
3353    
3354    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3355    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3356    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3357    assumes the way PCRE uses memmove() (always moving upwards).
3358    
3359    6. PCRE is being used in one environment where strchr() is not available. There
3360    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3361    faster code anyway.
3362    
3363    
3364    Version 3.1 09-Feb-00
3365    ---------------------
3366    
3367    The only change in this release is the fixing of some bugs in Makefile.in for
3368    the "install" target:
3369    
3370    (1) It was failing to install pcreposix.h.
3371    
3372    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3373    
3374    
3375    Version 3.0 01-Feb-00
3376    ---------------------
3377    
3378    1. Add support for the /+ modifier to perltest (to output $` like it does in
3379    pcretest).
3380    
3381    2. Add support for the /g modifier to perltest.
3382    
3383    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3384    matches null strings.
3385    
3386    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3387    pattern. Perl treats empty patterns specially - it reuses the most recent
3388    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3389    effect.
3390    
3391    5. The POSIX interface was broken in that it was just handing over the POSIX
3392    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3393    required a bigger vector, with some working space on the end. This means that
3394    the POSIX wrapper now has to get and free some memory, and copy the results.
3395    
3396    6. Added some simple autoconf support, placing the test data and the
3397    documentation in separate directories, re-organizing some of the
3398    information files, and making it build pcre-config (a GNU standard). Also added
3399    libtool support for building PCRE as a shared library, which is now the
3400    default.
3401    
3402    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3403    09 are not valid octal constants. Single digits will be used for minor values
3404    less than 10.
3405    
3406    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3407    existing programs that set these in the POSIX interface can use PCRE without
3408    modification.
3409    
3410    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3411    return all that pcre_info() returns, plus additional data. The pcre_info()
3412    function is retained for compatibility, but is considered to be obsolete.
3413    
3414    10. Added experimental recursion feature (?R) to handle one common case that
3415    Perl 5.6 will be able to do with (?p{...}).
3416    
3417    11. Added support for POSIX character classes like [:alpha:], which Perl is
3418    adopting.
3419    
3420    
3421    Version 2.08 31-Aug-99
3422    ----------------------
3423    
3424    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3425    trying to match at the startoffset position, but instead was moving forward to
3426    the next newline as if a previous match had failed.
3427    
3428    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3429    and could get into a loop if a null string was matched other than at the start
3430    of the subject.
3431    
3432    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3433    be distinguished at compile time, and for completeness also added PCRE_DATE.
3434    
3435    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3436    in GnuWin32 environments.
3437    
3438    
3439    Version 2.07 29-Jul-99
3440    ----------------------
3441    
3442    1. The documentation is now supplied in plain text form and HTML as well as in
3443    the form of man page sources.
3444    
3445    2. C++ compilers don't like assigning (void *) values to other pointer types.
3446    In particular this affects malloc(). Although there is no problem in Standard
3447    C, I've put in casts to keep C++ compilers happy.
3448    
3449    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3450    should be (const char *).
3451    
3452    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3453    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3454    However, I haven't made this a standard facility. The documentation doesn't
3455    mention it, and the Makefile doesn't support it.
3456    
3457    5. The Makefile now contains an "install" target, with editable destinations at
3458    the top of the file. The pcretest program is not installed.
3459    
3460    6. pgrep -V now gives the PCRE version number and date.
3461    
3462    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3463    causing the entire string to be ignored, instead of just the last character.
3464    
3465    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3466    non-matching string, it can take a very, very long time, even for strings of
3467    quite modest length, because of the nested recursion. PCRE now does better in
3468    some of these cases. It does this by remembering the last required literal
3469    character in the pattern, and pre-searching the subject to ensure it is present
3470    before running the real match. In other words, it applies a heuristic to detect
3471    some types of certain failure quickly, and in the above example, if presented
3472    with a string that has no trailing " it gives "no match" very quickly.
3473    
3474    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3475    other alternatives are tried instead.
3476    
3477    
3478    Version 2.06 09-Jun-99
3479    ----------------------
3480    
3481    1. Change pcretest's output for amount of store used to show just the code
3482    space, because the remainder (the data block) varies in size between 32-bit and
3483    64-bit systems.
3484    
3485    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3486    start matching at. This allows lookbehinds to work when searching for multiple
3487    occurrences in a string.
3488    
3489    3. Added additional options to pcretest for testing multiple occurrences:
3490    
3491       /+   outputs the rest of the string that follows a match
3492       /g   loops for multiple occurrences, using the new startoffset argument
3493       /G   loops for multiple occurrences by passing an incremented pointer
3494    
3495    4. PCRE wasn't doing the "first character" optimization for patterns starting
3496    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3497    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3498    the letter 'x'. On long subject strings, this gives a significant speed-up.
3499    
3500    
3501    Version 2.05 21-Apr-99
3502    ----------------------
3503    
3504    1. Changed the type of magic_number from int to long int so that it works
3505    properly on 16-bit systems.
3506    
3507    2. Fixed a bug which caused patterns starting with .* not to work correctly
3508    when the subject string contained newline characters. PCRE was assuming
3509    anchoring for such patterns in all cases, which is not correct because .* will
3510    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3511    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3512    must be retried after every newline in the subject.
3513    
3514    
3515    Version 2.04 18-Feb-99
3516    ----------------------
3517    
3518    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3519    computation of the store needed to hold the pattern was incorrect (too large).
3520    If such patterns were nested a few deep, this could multiply and become a real
3521    problem.
3522    
3523    2. Added /M option to pcretest to show the memory requirement of a specific
3524    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3525    
3526    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3527