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

Diff of /code/trunk/ChangeLog

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

revision 53 by nigel, Sat Feb 24 21:39:42 2007 UTC revision 564 by ph10, Sun Oct 31 16:07:24 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.11 10-Oct-2010
5    ------------------------
6    
7    1.  (*THEN) was not working properly if there were untried alternatives prior
8        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
9        backtracked to try for "b" instead of moving to the next alternative branch
10        at the same level (in this case, to look for "c"). The Perl documentation
11        is clear that when (*THEN) is backtracked onto, it goes to the "next
12        alternative in the innermost enclosing group".
13    
14    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
15        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
16        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
17        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
18        (*THEN).
19    
20    3.  If \s appeared in a character class, it removed the VT character from
21        the class, even if it had been included by some previous item, for example
22        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
23        of \s, but is part of the POSIX "space" class.)
24    
25    4.  A partial match never returns an empty string (because you can always
26        match an empty string at the end of the subject); however the checking for
27        an empty string was starting at the "start of match" point. This has been
28        changed to the "earliest inspected character" point, because the returned
29        data for a partial match starts at this character. This means that, for
30        example, /(?<=abc)def/ gives a partial match for the subject "abc"
31        (previously it gave "no match").
32    
33    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
34        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
35        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
36        has an implication that the given string is incomplete (because a partial
37        match is preferred over a full match). For this reason, these items now
38        give a partial match in this situation. [Aside: previously, the one case
39        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
40        match rather than a full match, which was wrong by the old rules, but is
41        now correct.]
42    
43    6.  There was a bug in the handling of #-introduced comments, recognized when
44        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
45        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
46        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
47        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
48        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
49        places in pcre_compile().
50    
51    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
52        comments when looking ahead for named forward references to subpatterns,
53        the only newline sequence it recognized was NL. It now handles newlines
54        according to the set newline convention.
55    
56    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
57        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
58        cater for a lack of strerror(). These oversights have been fixed.
59    
60    9.  Added --match-limit and --recursion-limit to pcregrep.
61    
62    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
63    
64    11. When the -o option was used, pcregrep was setting a return code of 1, even
65        when matches were found, and --line-buffered was not being honoured.
66    
67    
68    Version 8.10 25-Jun-2010
69    ------------------------
70    
71    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
72        THEN.
73    
74    2.  (*ACCEPT) was not working when inside an atomic group.
75    
76    3.  Inside a character class, \B is treated as a literal by default, but
77        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
78        causes the error). The code is unchanged, but I tidied the documentation.
79    
80    4.  Inside a character class, PCRE always treated \R and \X as literals,
81        whereas Perl faults them if its -w option is set. I have changed PCRE so
82        that it faults them when PCRE_EXTRA is set.
83    
84    5.  Added support for \N, which always matches any character other than
85        newline. (It is the same as "." when PCRE_DOTALL is not set.)
86    
87    6.  When compiling pcregrep with newer versions of gcc which may have
88        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
89        declared with attribute warn_unused_result" were given. Just casting the
90        result to (void) does not stop the warnings; a more elaborate fudge is
91        needed. I've used a macro to implement this.
92    
93    7.  Minor change to pcretest.c to avoid a compiler warning.
94    
95    8.  Added four artifical Unicode properties to help with an option to make
96        \s etc use properties (see next item). The new properties are: Xan
97        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
98    
99    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
100        use Unicode properties. (*UCP) at the start of a pattern can be used to set
101        this option. Modified pcretest to add /W to test this facility. Added
102        REG_UCP to make it available via the POSIX interface.
103    
104    10. Added --line-buffered to pcregrep.
105    
106    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
107        studied, and the match started with a letter with a code point greater than
108        127 whose first byte was different to the first byte of the other case of
109        the letter, the other case of this starting letter was not recognized
110        (#976).
111    
112    12. If a pattern that was studied started with a repeated Unicode property
113        test, for example, \p{Nd}+, there was the theoretical possibility of
114        setting up an incorrect bitmap of starting bytes, but fortunately it could
115        not have actually happened in practice until change 8 above was made (it
116        added property types that matched character-matching opcodes).
117    
118    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
119        possible starting bytes for non-anchored patterns.
120    
121    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
122        \R, and also a number of cases that involve Unicode properties, both
123        explicit and implicit when PCRE_UCP is set.
124    
125    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
126        input, it could crash or give wrong results if characters with values
127        greater than 0xc0 were present in the subject string. (Detail: it assumed
128        UTF-8 input when processing these items.)
129    
130    16. Added a lot of (int) casts to avoid compiler warnings in systems where
131        size_t is 64-bit (#991).
132    
133    17. Added a check for running out of memory when PCRE is compiled with
134        --disable-stack-for-recursion (#990).
135    
136    18. If the last data line in a file for pcretest does not have a newline on
137        the end, a newline was missing in the output.
138    
139    19. The default pcre_chartables.c file recognizes only ASCII characters (values
140        less than 128) in its various bitmaps. However, there is a facility for
141        generating tables according to the current locale when PCRE is compiled. It
142        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
143        characters, are recognized by isspace() and therefore were getting set in
144        these tables, and indeed these tables seem to approximate to ISO 8859. This
145        caused a problem in UTF-8 mode when pcre_study() was used to create a list
146        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
147        which of course cannot start UTF-8 characters. I have changed the code so
148        that only real ASCII characters (less than 128) and the correct starting
149        bytes for UTF-8 encodings are set for characters greater than 127 when in
150        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
151        altogether.)
152    
153    20. Added the /T option to pcretest so as to be able to run tests with non-
154        standard character tables, thus making it possible to include the tests
155        used for 19 above in the standard set of tests.
156    
157    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
158        reference to a subpattern the other side of a comment that contains an
159        opening parenthesis caused either an internal compiling error, or a
160        reference to the wrong subpattern.
161    
162    
163    Version 8.02 19-Mar-2010
164    ------------------------
165    
166    1.  The Unicode data tables have been updated to Unicode 5.2.0.
167    
168    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
169        configured.
170    
171    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
172        original author of that file, following a query about its status.
173    
174    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
175        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
176    
177    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
178        quantifier applied to a forward-referencing subroutine call, could compile
179        incorrect code or give the error "internal error: previously-checked
180        referenced subpattern not found".
181    
182    6.  Both MS Visual Studio and Symbian OS have problems with initializing
183        variables to point to external functions. For these systems, therefore,
184        pcre_malloc etc. are now initialized to local functions that call the
185        relevant global functions.
186    
187    7.  There were two entries missing in the vectors called coptable and poptable
188        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
189        I've fixed the data, and added a kludgy way of testing at compile time that
190        the lengths are correct (equal to the number of opcodes).
191    
192    8.  Following on from 7, I added a similar kludge to check the length of the
193        eint vector in pcreposix.c.
194    
195    9.  Error texts for pcre_compile() are held as one long string to avoid too
196        much relocation at load time. To find a text, the string is searched,
197        counting zeros. There was no check for running off the end of the string,
198        which could happen if a new error number was added without updating the
199        string.
200    
201    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
202    
203    11. \K was not working if it appeared in an atomic group or in a group that
204        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
205        \K is "not well defined" if used in an assertion. PCRE now accepts it if
206        the assertion is positive, but not if it is negative.
207    
208    12. Change 11 fortuitously reduced the size of the stack frame used in the
209        "match()" function of pcre_exec.c by one pointer. Forthcoming
210        implementation of support for (*MARK) will need an extra pointer on the
211        stack; I have reserved it now, so that the stack frame size does not
212        decrease.
213    
214    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
215        item in branch that calls a recursion is a subroutine call - as in the
216        second branch in the above example - was incorrectly given the compile-
217        time error "recursive call could loop indefinitely" because pcre_compile()
218        was not correctly checking the subroutine for matching a non-empty string.
219    
220    14. The checks for overrunning compiling workspace could trigger after an
221        overrun had occurred. This is a "should never occur" error, but it can be
222        triggered by pathological patterns such as hundreds of nested parentheses.
223        The checks now trigger 100 bytes before the end of the workspace.
224    
225    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
226    
227    
228    Version 8.01 19-Jan-2010
229    ------------------------
230    
231    1.  If a pattern contained a conditional subpattern with only one branch (in
232        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
233        computed the wrong minimum data length (which is of course zero for such
234        subpatterns). This could cause incorrect "no match" results.
235    
236    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
237        the pattern is reset in the first branch, pcre_compile() failed with
238        "internal error: code overflow at offset...". This happened only when
239        the reset was to the original external option setting. (An optimization
240        abstracts leading options settings into an external setting, which was the
241        cause of this.)
242    
243    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
244        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
245        assertion pattern did not match (meaning that the assertion was true), it
246        was incorrectly treated as false if the SKIP had been reached during the
247        matching. This also applied to assertions used as conditions.
248    
249    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
250        assertion subpattern, including such a pattern used as a condition,
251        unpredictable results occurred, instead of the error return
252        PCRE_ERROR_DFA_UITEM.
253    
254    5.  The C++ GlobalReplace function was not working like Perl for the special
255        situation when an empty string is matched. It now does the fancy magic
256        stuff that is necessary.
257    
258    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
259        removed. (These were left over from very, very early versions of PCRE.)
260    
261    7.  Some cosmetic changes to the code to make life easier when compiling it
262        as part of something else:
263    
264        (a) Change DEBUG to PCRE_DEBUG.
265    
266        (b) In pcre_compile(), rename the member of the "branch_chain" structure
267            called "current" as "current_branch", to prevent a collision with the
268            Linux macro when compiled as a kernel module.
269    
270        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
271            prevent a collision with the Linux macro when compiled as a kernel
272            module.
273    
274    8.  In pcre_compile() there are some checks for integer overflows that used to
275        cast potentially large values to (double). This has been changed to that
276        when building, a check for int64_t is made, and if it is found, it is used
277        instead, thus avoiding the use of floating point arithmetic. (There is no
278        other use of FP in PCRE.) If int64_t is not found, the fallback is to
279        double.
280    
281    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
282        2005 (difference between two addresses compared to an unsigned value).
283    
284    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
285        custom one, because of the following reported problem in Windows:
286    
287          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
288              under Win32.
289          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
290              therefore missing the function definition.
291          - The compiler thus generates a "C" signature for the test function.
292          - The linker fails to find the "C" function.
293          - PCRE fails to configure if asked to do so against libbz2.
294    
295    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
296        messages were output:
297    
298          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
299          rerunning libtoolize, to keep the correct libtool macros in-tree.
300          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
301    
302        I have done both of these things.
303    
304    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
305        most of the time, it *can* run out if it is given a pattern that contains a
306        runaway infinite recursion. I updated the discussion in the pcrestack man
307        page.
308    
309    13. Now that we have gone to the x.xx style of version numbers, the minor
310        version may start with zero. Using 08 or 09 is a bad idea because users
311        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
312        interpreted as invalid octal numbers. I've updated the previous comment in
313        configure.ac, and also added a check that gives an error if 08 or 09 are
314        used.
315    
316    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
317        causing partial matching to fail when the end of the subject matched \W
318        in a UTF-8 pattern where \W was quantified with a minimum of 3.
319    
320    15. There were some discrepancies between the declarations in pcre_internal.h
321        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
322        their definitions. The declarations used "const uschar *" and the
323        definitions used USPTR. Even though USPTR is normally defined as "const
324        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
325        reported that: "This difference in casting confuses some C++ compilers, for
326        example, SunCC recognizes above declarations as different functions and
327        generates broken code for hbpcre." I have changed the declarations to use
328        USPTR.
329    
330    16. GNU libtool is named differently on some systems. The autogen.sh script now
331        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
332        (FreeBSD).
333    
334    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
335        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
336        comment: "Figure out how to create a longlong from a string: strtoll and
337        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
338        instance, but it only takes 2 args instead of 3!"
339    
340    18. A subtle bug concerned with back references has been fixed by a change of
341        specification, with a corresponding code fix. A pattern such as
342        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
343        refers, was giving matches when it shouldn't. For example, xa=xaaa would
344        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
345        same bug. Such groups have to be quantified to be useful, or contained
346        inside another quantified group. (If there's no repetition, the reference
347        can never match.) The problem arises because, having left the group and
348        moved on to the rest of the pattern, a later failure that backtracks into
349        the group uses the captured value from the final iteration of the group
350        rather than the correct earlier one. I have fixed this in PCRE by forcing
351        any group that contains a reference to itself to be an atomic group; that
352        is, there cannot be any backtracking into it once it has completed. This is
353        similar to recursive and subroutine calls.
354    
355    
356    Version 8.00 19-Oct-09
357    ----------------------
358    
359    1.  The table for translating pcre_compile() error codes into POSIX error codes
360        was out-of-date, and there was no check on the pcre_compile() error code
361        being within the table. This could lead to an OK return being given in
362        error.
363    
364    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
365        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
366        in a Windows environment.
367    
368    3.  The pcregrep --count option prints the count for each file even when it is
369        zero, as does GNU grep. However, pcregrep was also printing all files when
370        --files-with-matches was added. Now, when both options are given, it prints
371        counts only for those files that have at least one match. (GNU grep just
372        prints the file name in this circumstance, but including the count seems
373        more useful - otherwise, why use --count?) Also ensured that the
374        combination -clh just lists non-zero counts, with no names.
375    
376    4.  The long form of the pcregrep -F option was incorrectly implemented as
377        --fixed_strings instead of --fixed-strings. This is an incompatible change,
378        but it seems right to fix it, and I didn't think it was worth preserving
379        the old behaviour.
380    
381    5.  The command line items --regex=pattern and --regexp=pattern were not
382        recognized by pcregrep, which required --regex pattern or --regexp pattern
383        (with a space rather than an '='). The man page documented the '=' forms,
384        which are compatible with GNU grep; these now work.
385    
386    6.  No libpcreposix.pc file was created for pkg-config; there was just
387        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
388    
389    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
390        when UCP support is not needed, by modifying the Python script that
391        generates it from Unicode data files. This should not matter if the module
392        is correctly used as a library, but I received one complaint about 50K of
393        unwanted data. My guess is that the person linked everything into his
394        program rather than using a library. Anyway, it does no harm.
395    
396    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
397        was a minimum greater than 1 for a wide character in a possessive
398        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
399        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
400        character. Chaos in the form of incorrect output or a compiling loop could
401        result.
402    
403    9.  The restrictions on what a pattern can contain when partial matching is
404        requested for pcre_exec() have been removed. All patterns can now be
405        partially matched by this function. In addition, if there are at least two
406        slots in the offset vector, the offset of the earliest inspected character
407        for the match and the offset of the end of the subject are set in them when
408        PCRE_ERROR_PARTIAL is returned.
409    
410    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
411        synonymous with PCRE_PARTIAL, for backwards compatibility, and
412        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
413        and may be more useful for multi-segment matching.
414    
415    11. Partial matching with pcre_exec() is now more intuitive. A partial match
416        used to be given if ever the end of the subject was reached; now it is
417        given only if matching could not proceed because another character was
418        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
419        string "Z", which now yields "no match" instead of "partial match". In the
420        case of pcre_dfa_exec(), "no match" is given if every matching path for the
421        final character ended with (*FAIL).
422    
423    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
424        if the pattern had a "must contain" character that was already found in the
425        earlier partial match, unless partial matching was again requested. For
426        example, with the pattern /dog.(body)?/, the "must contain" character is
427        "g". If the first part-match was for the string "dog", restarting with
428        "sbody" failed. This bug has been fixed.
429    
430    13. The string returned by pcre_dfa_exec() after a partial match has been
431        changed so that it starts at the first inspected character rather than the
432        first character of the match. This makes a difference only if the pattern
433        starts with a lookbehind assertion or \b or \B (\K is not supported by
434        pcre_dfa_exec()). It's an incompatible change, but it makes the two
435        matching functions compatible, and I think it's the right thing to do.
436    
437    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
438        so that the demonstration program is easily available in environments where
439        PCRE has not been installed from source.
440    
441    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
442        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
443        library.
444    
445    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
446        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
447        is not the first non-POSIX option to be added. Clearly some people find
448        these options useful.
449    
450    17. If a caller to the POSIX matching function regexec() passes a non-zero
451        value for nmatch with a NULL value for pmatch, the value of
452        nmatch is forced to zero.
453    
454    18. RunGrepTest did not have a test for the availability of the -u option of
455        the diff command, as RunTest does. It now checks in the same way as
456        RunTest, and also checks for the -b option.
457    
458    19. If an odd number of negated classes containing just a single character
459        interposed, within parentheses, between a forward reference to a named
460        subpattern and the definition of the subpattern, compilation crashed with
461        an internal error, complaining that it could not find the referenced
462        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
463        [The bug was that it was starting one character too far in when skipping
464        over the character class, thus treating the ] as data rather than
465        terminating the class. This meant it could skip too much.]
466    
467    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
468        /g option in pcretest when the pattern contains \K, which makes it possible
469        to have an empty string match not at the start, even when the pattern is
470        anchored. Updated pcretest and pcredemo to use this option.
471    
472    21. If the maximum number of capturing subpatterns in a recursion was greater
473        than the maximum at the outer level, the higher number was returned, but
474        with unset values at the outer level. The correct (outer level) value is
475        now given.
476    
477    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
478        PCRE did not set those parentheses (unlike Perl). I have now found a way to
479        make it do so. The string so far is captured, making this feature
480        compatible with Perl.
481    
482    23. The tests have been re-organized, adding tests 11 and 12, to make it
483        possible to check the Perl 5.10 features against Perl 5.10.
484    
485    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
486        pattern matches a fixed length string. PCRE did not allow this; now it
487        does. Neither allows recursion.
488    
489    25. I finally figured out how to implement a request to provide the minimum
490        length of subject string that was needed in order to match a given pattern.
491        (It was back references and recursion that I had previously got hung up
492        on.) This code has now been added to pcre_study(); it finds a lower bound
493        to the length of subject needed. It is not necessarily the greatest lower
494        bound, but using it to avoid searching strings that are too short does give
495        some useful speed-ups. The value is available to calling programs via
496        pcre_fullinfo().
497    
498    26. While implementing 25, I discovered to my embarrassment that pcretest had
499        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
500        study optimizations had never been tested with that matching function.
501        Oops. What is worse, even when it was passed study data, there was a bug in
502        pcre_dfa_exec() that meant it never actually used it. Double oops. There
503        were also very few tests of studied patterns with pcre_dfa_exec().
504    
505    27. If (?| is used to create subpatterns with duplicate numbers, they are now
506        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
507        on the other side of the coin, they are no longer allowed to have different
508        names, because these cannot be distinguished in PCRE, and this has caused
509        confusion. (This is a difference from Perl.)
510    
511    28. When duplicate subpattern names are present (necessarily with different
512        numbers, as required by 27 above), and a test is made by name in a
513        conditional pattern, either for a subpattern having been matched, or for
514        recursion in such a pattern, all the associated numbered subpatterns are
515        tested, and the overall condition is true if the condition is true for any
516        one of them. This is the way Perl works, and is also more like the way
517        testing by number works.
518    
519    
520    Version 7.9 11-Apr-09
521    ---------------------
522    
523    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
524        (pcretest), all targets were linked against these libraries. This included
525        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
526        libraries. This caused unwanted dependencies to be created. This problem
527        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
528        pcretest is linked with readline.
529    
530    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
531        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
532        moved outside it again, because FALSE and TRUE are already defined in AIX,
533        but BOOL is not.
534    
535    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
536        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
537    
538    4.  The pcregrep documentation said spaces were inserted as well as colons (or
539        hyphens) following file names and line numbers when outputting matching
540        lines. This is not true; no spaces are inserted. I have also clarified the
541        wording for the --colour (or --color) option.
542    
543    5.  In pcregrep, when --colour was used with -o, the list of matching strings
544        was not coloured; this is different to GNU grep, so I have changed it to be
545        the same.
546    
547    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
548        each matching line was coloured. Now it goes on to look for further matches
549        of any of the test patterns, which is the same behaviour as GNU grep.
550    
551    7.  A pattern that could match an empty string could cause pcregrep to loop; it
552        doesn't make sense to accept an empty string match in pcregrep, so I have
553        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
554        seems to be how GNU grep behaves.
555    
556    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
557        start or after a newline", because the conditional assertion was not being
558        correctly handled. The rule now is that both the assertion and what follows
559        in the first alternative must satisfy the test.
560    
561    9.  If auto-callout was enabled in a pattern with a conditional group whose
562        condition was an assertion, PCRE could crash during matching, both with
563        pcre_exec() and pcre_dfa_exec().
564    
565    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
566        used for matching.
567    
568    11. Unicode property support in character classes was not working for
569        characters (bytes) greater than 127 when not in UTF-8 mode.
570    
571    12. Added the -M command line option to pcretest.
572    
573    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
574    
575    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
576    
577    16. Added comments and documentation about mis-use of no_arg in the C++
578        wrapper.
579    
580    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
581        from Martin Jerabek that uses macro names for all relevant character and
582        string constants.
583    
584    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
585        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
586        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
587        these, but not everybody uses configure.
588    
589    19. A conditional group that had only one branch was not being correctly
590        recognized as an item that could match an empty string. This meant that an
591        enclosing group might also not be so recognized, causing infinite looping
592        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
593        with the subject "ab", where knowledge that the repeated group can match
594        nothing is needed in order to break the loop.
595    
596    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
597        exec(), but without supplying a callout function, matching went wrong.
598    
599    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
600        leak if the size of the offset vector was greater than 30. When the vector
601        is smaller, the saved offsets during recursion go onto a local stack
602        vector, but for larger vectors malloc() is used. It was failing to free
603        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
604        error, in fact).
605    
606    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
607        heapframe that is used only when UTF-8 support is enabled. This caused no
608        problem, but was untidy.
609    
610    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
611        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
612        included within another project.
613    
614    24. Steven Van Ingelgem's patches to add more options to the CMake support,
615        slightly modified by me:
616    
617          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
618              not building pcregrep.
619    
620          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
621              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
622    
623    25. Forward references, both numeric and by name, in patterns that made use of
624        duplicate group numbers, could behave incorrectly or give incorrect errors,
625        because when scanning forward to find the reference group, PCRE was not
626        taking into account the duplicate group numbers. A pattern such as
627        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
628    
629    26. Changed a few more instances of "const unsigned char *" to USPTR, making
630        the feature of a custom pointer more persuasive (as requested by a user).
631    
632    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
633        pcretest.c, inside #ifndefs, because it seems they are sometimes already
634        pre-defined.
635    
636    28. Added support for (*UTF8) at the start of a pattern.
637    
638    29. Arrange for flags added by the "release type" setting in CMake to be shown
639        in the configuration summary.
640    
641    
642    Version 7.8 05-Sep-08
643    ---------------------
644    
645    1.  Replaced UCP searching code with optimized version as implemented for Ad
646        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
647        stage table and inline lookup instead of a function, giving speed ups of 2
648        to 5 times on some simple patterns that I tested. Permission was given to
649        distribute the MultiStage2.py script that generates the tables (it's not in
650        the tarball, but is in the Subversion repository).
651    
652    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
653        scripts.
654    
655    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
656        a group with a zero qualifier. The result of the study could be incorrect,
657        or the function might crash, depending on the pattern.
658    
659    4.  Caseless matching was not working for non-ASCII characters in back
660        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
661        It now works when Unicode Property Support is available.
662    
663    5.  In pcretest, an escape such as \x{de} in the data was always generating
664        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
665        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
666        truncation.
667    
668    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
669    
670    7.  Added two (int) casts to pcregrep when printing the difference of two
671        pointers, in case they are 64-bit values.
672    
673    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
674        test 2 if it fails.
675    
676    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
677        and a #define of that name to empty if it is not externally set. This is to
678        allow users of MSVC to set it if necessary.
679    
680    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
681        the convenience functions in the pcre_get.c source file.
682    
683    11. An option change at the start of a pattern that had top-level alternatives
684        could cause overwriting and/or a crash. This command provoked a crash in
685        some environments:
686    
687          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
688    
689        This potential security problem was recorded as CVE-2008-2371.
690    
691    12. For a pattern where the match had to start at the beginning or immediately
692        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
693        pcre_dfa_exec() could read past the end of the passed subject if there was
694        no match. To help with detecting such bugs (e.g. with valgrind), I modified
695        pcretest so that it places the subject at the end of its malloc-ed buffer.
696    
697    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
698        exec() might read past the end of the data buffer in UTF-8 mode.
699    
700    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
701        the data contained the byte 0x85 as part of a UTF-8 character within its
702        first line. This applied both to normal and DFA matching.
703    
704    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
705        /^[^d]*?$/8 failed to match "abc".
706    
707    16. Added a missing copyright notice to pcrecpp_internal.h.
708    
709    17. Make it more clear in the documentation that values returned from
710        pcre_exec() in ovector are byte offsets, not character counts.
711    
712    18. Tidied a few places to stop certain compilers from issuing warnings.
713    
714    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
715        supplied by Stefan Weber. I made a further small update for 7.8 because
716        there is a change of source arrangements: the pcre_searchfuncs.c module is
717        replaced by pcre_ucd.c.
718    
719    
720    Version 7.7 07-May-08
721    ---------------------
722    
723    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
724        a string to a long long, pretend we don't even have a long long." This is
725        done by checking for the strtoq, strtoll, and _strtoi64 functions.
726    
727    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
728        pre-7.6 versions, which defined a global no_arg variable instead of putting
729        it in the RE class. (See also #8 below.)
730    
731    3.  Remove a line of dead code, identified by coverity and reported by Nuno
732        Lopes.
733    
734    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
735    
736        (1) The include/exclude patterns were being applied to the whole pathnames
737            of files, instead of just to the final components.
738    
739        (2) If there was more than one level of directory, the subdirectories were
740            skipped unless they satisfied the include/exclude conditions. This is
741            inconsistent with GNU grep (and could even be seen as contrary to the
742            pcregrep specification - which I improved to make it absolutely clear).
743            The action now is always to scan all levels of directory, and just
744            apply the include/exclude patterns to regular files.
745    
746    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
747        --exclude_dir in the tests to avoid scanning .svn directories.
748    
749    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
750        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
751        doesn't support NULs in patterns.
752    
753    7.  Added some missing "const"s to declarations of static tables in
754        pcre_compile.c and pcre_dfa_exec.c.
755    
756    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
757        caused by fix #2  above. (Subsequently also a second patch to fix the
758        first patch. And a third patch - this was a messy problem.)
759    
760    9.  Applied Craig's patch to remove the use of push_back().
761    
762    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
763        matching function regexec().
764    
765    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
766        which, however, unlike Perl's \g{...}, are subroutine calls, not back
767        references. PCRE supports relative numbers with this syntax (I don't think
768        Oniguruma does).
769    
770    12. Previously, a group with a zero repeat such as (...){0} was completely
771        omitted from the compiled regex. However, this means that if the group
772        was called as a subroutine from elsewhere in the pattern, things went wrong
773        (an internal error was given). Such groups are now left in the compiled
774        pattern, with a new opcode that causes them to be skipped at execution
775        time.
776    
777    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
778        to the way PCRE behaves:
779    
780        (a) A lone ] character is dis-allowed (Perl treats it as data).
781    
782        (b) A back reference to an unmatched subpattern matches an empty string
783            (Perl fails the current match path).
784    
785        (c) A data ] in a character class must be notated as \] because if the
786            first data character in a class is ], it defines an empty class. (In
787            Perl it is not possible to have an empty class.) The empty class []
788            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
789            The negative empty class [^] matches any one character, independently
790            of the DOTALL setting.
791    
792    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
793        non-existent subpattern following a character class starting with ']' and
794        containing () gave an internal compiling error instead of "reference to
795        non-existent subpattern". Fortunately, when the pattern did exist, the
796        compiled code was correct. (When scanning forwards to check for the
797        existencd of the subpattern, it was treating the data ']' as terminating
798        the class, so got the count wrong. When actually compiling, the reference
799        was subsequently set up correctly.)
800    
801    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
802        it was being rejected as not supported by pcre_dfa_exec(), even though
803        other assertions are supported. I have made pcre_dfa_exec() support
804        (*FAIL).
805    
806    16. The implementation of 13c above involved the invention of a new opcode,
807        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
808        cannot be changed at match time, I realized I could make a small
809        improvement to matching performance by compiling OP_ALLANY instead of
810        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
811        on the OP_ANY path.
812    
813    17. Compiling pcretest on Windows with readline support failed without the
814        following two fixes: (1) Make the unistd.h include conditional on
815        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
816    
817    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
818        ncurses library to be included for pcretest when ReadLine support is
819        requested, but also to allow for it to be overridden. This patch came from
820        Daniel Bergström.
821    
822    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
823        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
824        any errors with the current Unicode tables. Thanks to Peter Kankowski for
825        spotting this.
826    
827    
828    Version 7.6 28-Jan-08
829    ---------------------
830    
831    1.  A character class containing a very large number of characters with
832        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
833        overflow.
834    
835    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
836        HAVE_LONG_LONG is not defined.
837    
838    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
839        bring them up to date and include new features. This patch includes:
840    
841        - Fixed PH's badly added libz and libbz2 support.
842        - Fixed a problem with static linking.
843        - Added pcredemo. [But later removed - see 7 below.]
844        - Fixed dftables problem and added an option.
845        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
846            HAVE_LONG_LONG.
847        - Added readline support for pcretest.
848        - Added an listing of the option settings after cmake has run.
849    
850    4.  A user submitted a patch to Makefile that makes it easy to create
851        "pcre.dll" under mingw when using Configure/Make. I added stuff to
852        Makefile.am that cause it to include this special target, without
853        affecting anything else. Note that the same mingw target plus all
854        the other distribution libraries and programs are now supported
855        when configuring with CMake (see 6 below) instead of with
856        Configure/Make.
857    
858    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
859        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
860        exported in the Windows port". It has not yet been confirmed that the patch
861        solves the problem, but it does no harm.
862    
863    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
864        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
865        with CMake, and also correct the comment about stack recursion.
866    
867    7.  Remove the automatic building of pcredemo from the ./configure system and
868        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
869        of a program that users should build themselves after PCRE is installed, so
870        building it automatically is not really right. What is more, it gave
871        trouble in some build environments.
872    
873    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
874    
875    
876    Version 7.5 10-Jan-08
877    ---------------------
878    
879    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
880        values in parens when parsing an RE using the C++ wrapper."
881    
882    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
883        Characters greater than 255 were excluded from the class instead of being
884        included.
885    
886    3.  The same bug as (2) above applied to negated POSIX classes such as
887        [:^space:].
888    
889    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
890        defined or documented. It seems to have been a typo for PCRE_STATIC, so
891        I have changed it.
892    
893    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
894        first named subpattern) and a construct such as (?&a) would reference the
895        first named subpattern whose name started with "a" (in other words, the
896        length check was missing). Both these problems are fixed. "Subpattern name
897        expected" is now given for (?&) (a zero-length name), and this patch also
898        makes it give the same error for \k'' (previously it complained that that
899        was a reference to a non-existent subpattern).
900    
901    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
902        this is right because (?- can be followed by option settings as well as by
903        digits. I have, however, made the messages clearer.
904    
905    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
906        than the number used in the conditional) now cause a compile-time error.
907        This is actually not compatible with Perl, which accepts such patterns, but
908        treats the conditional as always being FALSE (as PCRE used to), but it
909        seems to me that giving a diagnostic is better.
910    
911    8.  Change "alphameric" to the more common word "alphanumeric" in comments
912        and messages.
913    
914    9.  Fix two occurrences of "backslash" in comments that should have been
915        "backspace".
916    
917    10. Remove two redundant lines of code that can never be obeyed (their function
918        was moved elsewhere).
919    
920    11. The program that makes PCRE's Unicode character property table had a bug
921        which caused it to generate incorrect table entries for sequences of
922        characters that have the same character type, but are in different scripts.
923        It amalgamated them into a single range, with the script of the first of
924        them. In other words, some characters were in the wrong script. There were
925        thirteen such cases, affecting characters in the following ranges:
926    
927          U+002b0 - U+002c1
928          U+0060c - U+0060d
929          U+0061e - U+00612
930          U+0064b - U+0065e
931          U+0074d - U+0076d
932          U+01800 - U+01805
933          U+01d00 - U+01d77
934          U+01d9b - U+01dbf
935          U+0200b - U+0200f
936          U+030fc - U+030fe
937          U+03260 - U+0327f
938          U+0fb46 - U+0fbb1
939          U+10450 - U+1049d
940    
941    12. The -o option (show only the matching part of a line) for pcregrep was not
942        compatible with GNU grep in that, if there was more than one match in a
943        line, it showed only the first of them. It now behaves in the same way as
944        GNU grep.
945    
946    13. If the -o and -v options were combined for pcregrep, it printed a blank
947        line for every non-matching line. GNU grep prints nothing, and pcregrep now
948        does the same. The return code can be used to tell if there were any
949        non-matching lines.
950    
951    14. Added --file-offsets and --line-offsets to pcregrep.
952    
953    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
954        infinitely looping recursion. The bug was that positive lookaheads were not
955        being skipped when checking for a possible empty match (negative lookaheads
956        and both kinds of lookbehind were skipped).
957    
958    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
959        inclusion of <windows.h> to before rather than after the definition of
960        INVALID_FILE_ATTRIBUTES (patch from David Byron).
961    
962    17. Specifying a possessive quantifier with a specific limit for a Unicode
963        character property caused pcre_compile() to compile bad code, which led at
964        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
965        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
966        caused the error; without that there was no problem.
967    
968    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
969    
970    19. Added --enable-pcretest-libreadline.
971    
972    20. In pcrecpp.cc, the variable 'count' was incremented twice in
973        RE::GlobalReplace(). As a result, the number of replacements returned was
974        double what it should be. I removed one of the increments, but Craig sent a
975        later patch that removed the other one (the right fix) and added unit tests
976        that check the return values (which was not done before).
977    
978    21. Several CMake things:
979    
980        (1) Arranged that, when cmake is used on Unix, the libraries end up with
981            the names libpcre and libpcreposix, not just pcre and pcreposix.
982    
983        (2) The above change means that pcretest and pcregrep are now correctly
984            linked with the newly-built libraries, not previously installed ones.
985    
986        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
987    
988    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
989        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
990        UTF-8 newline character). The key issue is that the pattern starts .*;
991        this means that the match must be either at the beginning, or after a
992        newline. The bug was in the code for advancing after a failed match and
993        checking that the new position followed a newline. It was not taking
994        account of UTF-8 characters correctly.
995    
996    23. PCRE was behaving differently from Perl in the way it recognized POSIX
997        character classes. PCRE was not treating the sequence [:...:] as a
998        character class unless the ... were all letters. Perl, however, seems to
999        allow any characters between [: and :], though of course it rejects as
1000        unknown any "names" that contain non-letters, because all the known class
1001        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1002        for example, whereas PCRE did not - it did not recognize a POSIX character
1003        class. This seemed a bit dangerous, so the code has been changed to be
1004        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1005        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1006        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1007        Perl does, and where it didn't before.
1008    
1009    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1010        Windows environments %n is disabled by default.
1011    
1012    
1013    Version 7.4 21-Sep-07
1014    ---------------------
1015    
1016    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1017        means that a class such as [\s] counted as "explicit reference to CR or
1018        LF". That isn't really right - the whole point of the change was to try to
1019        help when there was an actual mention of one of the two characters. So now
1020        the change happens only if \r or \n (or a literal CR or LF) character is
1021        encountered.
1022    
1023    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1024        of both had grown to the point where there were only 3 bits left.
1025        Fortunately, there was spare space in the data structure, and so I have
1026        moved the internal flags into a new 16-bit field to free up more option
1027        bits.
1028    
1029    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1030        but did not set the internal JCHANGED flag - either of these is enough to
1031        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1032        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1033        start sets both bits.
1034    
1035    4.  Added options (at build time, compile time, exec time) to change \R from
1036        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1037    
1038    5.  doc/pcresyntax.html was missing from the distribution.
1039    
1040    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1041        compatibility, even though it is no longer used.
1042    
1043    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1044        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1045        windows.h file is present (where different names are used). [This was
1046        reversed later after testing - see 16 below.]
1047    
1048    8.  Changed all #include <config.h> to #include "config.h". There were also
1049        some further <pcre.h> cases that I changed to "pcre.h".
1050    
1051    9.  When pcregrep was used with the --colour option, it missed the line ending
1052        sequence off the lines that it output.
1053    
1054    10. It was pointed out to me that arrays of string pointers cause lots of
1055        relocations when a shared library is dynamically loaded. A technique of
1056        using a single long string with a table of offsets can drastically reduce
1057        these. I have refactored PCRE in four places to do this. The result is
1058        dramatic:
1059    
1060          Originally:                          290
1061          After changing UCP table:            187
1062          After changing error message table:   43
1063          After changing table of "verbs"       36
1064          After changing table of Posix names   22
1065    
1066        Thanks to the folks working on Gregex for glib for this insight.
1067    
1068    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1069        unicode-properties was also set.
1070    
1071    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1072    
1073    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1074        checked only for CRLF.
1075    
1076    14. Added casts to pcretest.c to avoid compiler warnings.
1077    
1078    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1079    
1080    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1081        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1082        entirely. This removes changes made in 7 above.
1083    
1084    17. The CMake files have been updated, and there is now more information about
1085        building with CMake in the NON-UNIX-USE document.
1086    
1087    
1088    Version 7.3 28-Aug-07
1089    ---------------------
1090    
1091     1. In the rejigging of the build system that eventually resulted in 7.1, the
1092        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1093        brackets there is not right, since it causes compilers to look for an
1094        installed pcre.h, not the version that is in the source that is being
1095        compiled (which of course may be different). I have changed it back to:
1096    
1097          #include "pcre.h"
1098    
1099        I have a vague recollection that the change was concerned with compiling in
1100        different directories, but in the new build system, that is taken care of
1101        by the VPATH setting the Makefile.
1102    
1103     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1104        when the subject happened to end in the byte 0x85 (e.g. if the last
1105        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1106        characters but of course it shouldn't be taken as a newline when it is part
1107        of another character. The bug was that, for an unlimited repeat of . in
1108        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1109        characters when looking for a newline.
1110    
1111     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1112    
1113     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1114        in debug output.
1115    
1116     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1117        long printing in the pcrecpp unittest when running under MinGW.
1118    
1119     6. ESC_K was left out of the EBCDIC table.
1120    
1121     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1122        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1123        limit also applies to "virtual nesting" when a pattern is recursive, and in
1124        this case 1000 isn't so big. I have been able to remove this limit at the
1125        expense of backing off one optimization in certain circumstances. Normally,
1126        when pcre_exec() would call its internal match() function recursively and
1127        immediately return the result unconditionally, it uses a "tail recursion"
1128        feature to save stack. However, when a subpattern that can match an empty
1129        string has an unlimited repetition quantifier, it no longer makes this
1130        optimization. That gives it a stack frame in which to save the data for
1131        checking that an empty string has been matched. Previously this was taken
1132        from the 1000-entry workspace that had been reserved. So now there is no
1133        explicit limit, but more stack is used.
1134    
1135     8. Applied Daniel's patches to solve problems with the import/export magic
1136        syntax that is required for Windows, and which was going wrong for the
1137        pcreposix and pcrecpp parts of the library. These were overlooked when this
1138        problem was solved for the main library.
1139    
1140     9. There were some crude static tests to avoid integer overflow when computing
1141        the size of patterns that contain repeated groups with explicit upper
1142        limits. As the maximum quantifier is 65535, the maximum group length was
1143        set at 30,000 so that the product of these two numbers did not overflow a
1144        32-bit integer. However, it turns out that people want to use groups that
1145        are longer than 30,000 bytes (though not repeat them that many times).
1146        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1147        made it possible to implement the integer overflow checks in a much more
1148        dynamic way, which I have now done. The artificial limitation on group
1149        length has been removed - we now have only the limit on the total length of
1150        the compiled pattern, which depends on the LINK_SIZE setting.
1151    
1152    10. Fixed a bug in the documentation for get/copy named substring when
1153        duplicate names are permitted. If none of the named substrings are set, the
1154        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1155        empty string.
1156    
1157    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1158        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1159        because the ] is interpreted as the first data character and the
1160        terminating ] is not found. PCRE has been made compatible with Perl in this
1161        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1162        cause memory overwriting.
1163    
1164    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1165        string has been matched (to stop an infinite loop). It was not recognizing
1166        a conditional subpattern that could match an empty string if that
1167        subpattern was within another subpattern. For example, it looped when
1168        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1169        condition was not nested. This bug has been fixed.
1170    
1171    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1172        past the start of the subject in the presence of bytes with the top bit
1173        set, for example "\x8aBCD".
1174    
1175    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1176        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1177    
1178    14. Optimized (?!) to (*FAIL).
1179    
1180    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1181        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1182        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1183        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1184        does: it's just the validity check that is more restrictive.
1185    
1186    16. Inserted checks for integer overflows during escape sequence (backslash)
1187        processing, and also fixed erroneous offset values for syntax errors during
1188        backslash processing.
1189    
1190    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1191        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1192    
1193    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1194        caused an overrun.
1195    
1196    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1197        something other than just ASCII characters) inside a group that had an
1198        unlimited repeat caused a loop at compile time (while checking to see
1199        whether the group could match an empty string).
1200    
1201    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1202        [\P{Any}] did so. (Error in the code for printing property names.)
1203    
1204    21. An orphan \E inside a character class could cause a crash.
1205    
1206    22. A repeated capturing bracket such as (A)? could cause a wild memory
1207        reference during compilation.
1208    
1209    23. There are several functions in pcre_compile() that scan along a compiled
1210        expression for various reasons (e.g. to see if it's fixed length for look
1211        behind). There were bugs in these functions when a repeated \p or \P was
1212        present in the pattern. These operators have additional parameters compared
1213        with \d, etc, and these were not being taken into account when moving along
1214        the compiled data. Specifically:
1215    
1216        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1217            length.
1218    
1219        (b) An item such as \pL+ within a repeated group could cause crashes or
1220            loops.
1221    
1222        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1223            "reference to non-existent subpattern" error.
1224    
1225        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1226    
1227    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1228        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1229    
1230    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1231    
1232    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1233        character were causing crashes (broken optimization).
1234    
1235    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1236        \p or \P) caused a compile-time loop.
1237    
1238    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1239        break. For example, the unstudied pattern [\r\n]A does not match the string
1240        "\r\nA" because change 7.0/46 below moves the current point on by two
1241        characters after failing to match at the start. However, the pattern \nA
1242        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1243        the same is true. There doesn't seem any very clean way out of this, but
1244        what I have chosen to do makes the common cases work: PCRE now takes note
1245        of whether there can be an explicit match for \r or \n anywhere in the
1246        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1247        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1248        pattern has explicit CR or LF references.
1249    
1250    29. Added (*CR) etc for changing newline setting at start of pattern.
1251    
1252    
1253    Version 7.2 19-Jun-07
1254    ---------------------
1255    
1256     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1257        which is apparently normally available under Windows.
1258    
1259     2. Re-jig the pcregrep tests with different newline settings in an attempt
1260        to make them independent of the local environment's newline setting.
1261    
1262     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1263    
1264     4. Some of the "internals" tests were previously cut out when the link size
1265        was not 2, because the output contained actual offsets. The recent new
1266        "Z" feature of pcretest means that these can be cut out, making the tests
1267        usable with all link sizes.
1268    
1269     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1270        stack recursion. This gives a massive performance boost under BSD, but just
1271        a small improvement under Linux. However, it saves one field in the frame
1272        in all cases.
1273    
1274     6. Added more features from the forthcoming Perl 5.10:
1275    
1276        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1277            recursion call. It refers to the nth most recently opened parentheses.
1278    
1279        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1280            to be opened parentheses.
1281    
1282        (c) Conditions that refer to capturing parentheses can be specified
1283            relatively, for example, (?(-2)... or (?(+3)...
1284    
1285        (d) \K resets the start of the current match so that everything before
1286            is not part of it.
1287    
1288        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1289    
1290        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1291            reference syntax.
1292    
1293        (g) (?| introduces a group in which the numbering of parentheses in each
1294            alternative starts with the same number.
1295    
1296        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1297    
1298     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1299        PCRE_INFO_JCHANGED.
1300    
1301     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1302        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1303        for detecting groups that can match an empty string.
1304    
1305     9. A pattern with a very large number of alternatives (more than several
1306        hundred) was running out of internal workspace during the pre-compile
1307        phase, where pcre_compile() figures out how much memory will be needed. A
1308        bit of new cunning has reduced the workspace needed for groups with
1309        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1310        workspace instead of running out of the 4096 that are available.
1311    
1312    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1313    
1314    11. Applied patch from Google to remove an optimization that didn't quite work.
1315        The report of the bug said:
1316    
1317          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1318          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1319          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1320    
1321    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1322        it matched the wrong number of bytes.
1323    
1324    
1325    Version 7.1 24-Apr-07
1326    ---------------------
1327    
1328     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1329        that is more "standard", making use of automake and other Autotools. There
1330        is some re-arrangement of the files and adjustment of comments consequent
1331        on this.
1332    
1333     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1334        for recursive directory scanning broke on some systems because the files
1335        are not scanned in any specific order and on different systems the order
1336        was different. A call to "sort" has been inserted into RunGrepTest for the
1337        approprate test as a short-term fix. In the longer term there may be an
1338        alternative.
1339    
1340     3. I had an email from Eric Raymond about problems translating some of PCRE's
1341        man pages to HTML (despite the fact that I distribute HTML pages, some
1342        people do their own conversions for various reasons). The problems
1343        concerned the use of low-level troff macros .br and .in. I have therefore
1344        removed all such uses from the man pages (some were redundant, some could
1345        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1346        HTML has been updated to handle .nf/.fi and to complain if it encounters
1347        .br or .in.
1348    
1349     4. Updated comments in configure.ac that get placed in config.h.in and also
1350        arranged for config.h to be included in the distribution, with the name
1351        config.h.generic, for the benefit of those who have to compile without
1352        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1353    
1354     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1355        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1356        makevp.bat for the current PCRE, using the additional files
1357        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1358    
1359     6. A Windows user reported a minor discrepancy with test 2, which turned out
1360        to be caused by a trailing space on an input line that had got lost in his
1361        copy. The trailing space was an accident, so I've just removed it.
1362    
1363     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1364        that is needed.
1365    
1366     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1367        as "const" (a) because they are and (b) because it helps the PHP
1368        maintainers who have recently made a script to detect big data structures
1369        in the php code that should be moved to the .rodata section. I remembered
1370        to update Builducptable as well, so it won't revert if ucptable.h is ever
1371        re-created.
1372    
1373     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1374        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1375        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1376        support is not required. This saves 1.5-2K of code, which is important in
1377        some applications.
1378    
1379        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1380        so as not to refer to the tables, even though these functions will never be
1381        called when UTF-8 support is disabled. Otherwise there are problems with a
1382        shared library.
1383    
1384    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1385    
1386        (a) It was defining its arguments as char * instead of void *.
1387    
1388        (b) It was assuming that all moves were upwards in memory; this was true
1389            a long time ago when I wrote it, but is no longer the case.
1390    
1391        The emulated memove() is provided for those environments that have neither
1392        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1393        is clearly not the case, as these two bugs were recently reported.
1394    
1395    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1396        and Detrail to create the HTML documentation, the .txt form of the man
1397        pages, and it removes trailing spaces from listed files. It also creates
1398        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1399        case, it wraps all the #defines with #ifndefs. This script should be run
1400        before "make dist".
1401    
1402    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1403        with Unicode property support.
1404    
1405        (a) For a maximizing quantifier, if the two different cases of the
1406            character were of different lengths in their UTF-8 codings (there are
1407            some cases like this - I found 11), and the matching function had to
1408            back up over a mixture of the two cases, it incorrectly assumed they
1409            were both the same length.
1410    
1411        (b) When PCRE was configured to use the heap rather than the stack for
1412            recursion during matching, it was not correctly preserving the data for
1413            the other case of a UTF-8 character when checking ahead for a match
1414            while processing a minimizing repeat. If the check also involved
1415            matching a wide character, but failed, corruption could cause an
1416            erroneous result when trying to check for a repeat of the original
1417            character.
1418    
1419    13. Some tidying changes to the testing mechanism:
1420    
1421        (a) The RunTest script now detects the internal link size and whether there
1422            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1423            values substituted by "configure". (The RunGrepTest script already did
1424            this for UTF-8.) The configure.ac script no longer substitutes the
1425            relevant variables.
1426    
1427        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1428            with length and offset values. This means that the output is different
1429            for different internal link sizes. Test 2 is skipped for link sizes
1430            other than 2 because of this, bypassing the problem. Unfortunately,
1431            there was also a test in test 3 (the locale tests) that used /B and
1432            failed for link sizes other than 2. Rather than cut the whole test out,
1433            I have added a new /Z option to pcretest that replaces the length and
1434            offset values with spaces. This is now used to make test 3 independent
1435            of link size. (Test 2 will be tidied up later.)
1436    
1437    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1438        segmentation fault instead of returning the appropriate error message.
1439    
1440    15. In multiline mode when the newline sequence was set to "any", the pattern
1441        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1442        This doesn't seem right; it now treats the CRLF combination as the line
1443        ending, and so does not match in that case. It's only a pattern such as ^$
1444        that would hit this one: something like ^ABC$ would have failed after \r
1445        and then tried again after \r\n.
1446    
1447    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1448        in an attempt to make files that differ only in their line terminators
1449        compare equal. This works on Linux.
1450    
1451    17. Under certain error circumstances pcregrep might try to free random memory
1452        as it exited. This is now fixed, thanks to valgrind.
1453    
1454    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1455        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1456        was because its rules for how to advance for /g after matching an empty
1457        string at the end of a line did not allow for this case. They now check for
1458        it specially.
1459    
1460    20. pcretest is supposed to handle patterns and data of any length, by
1461        extending its buffers when necessary. It was getting this wrong when the
1462        buffer for a data line had to be extended.
1463    
1464    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1465        CRLF as a newline sequence.
1466    
1467    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1468        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1469        I have nevertheless tidied it up.
1470    
1471    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1472    
1473    24. Added a man page for pcre-config.
1474    
1475    
1476    Version 7.0 19-Dec-06
1477    ---------------------
1478    
1479     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1480        moving to gcc 4.1.1.
1481    
1482     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1483        sys/time.h, which is documented as needed for this function. It doesn't
1484        seem to matter on Linux, but it showed up on some releases of OS X.
1485    
1486     3. It seems that there are systems where bytes whose values are greater than
1487        127 match isprint() in the "C" locale. The "C" locale should be the
1488        default when a C program starts up. In most systems, only ASCII printing
1489        characters match isprint(). This difference caused the output from pcretest
1490        to vary, making some of the tests fail. I have changed pcretest so that:
1491    
1492        (a) When it is outputting text in the compiled version of a pattern, bytes
1493            other than 32-126 are always shown as hex escapes.
1494    
1495        (b) When it is outputting text that is a matched part of a subject string,
1496            it does the same, unless a different locale has been set for the match
1497            (using the /L modifier). In this case, it uses isprint() to decide.
1498    
1499     4. Fixed a major bug that caused incorrect computation of the amount of memory
1500        required for a compiled pattern when options that changed within the
1501        pattern affected the logic of the preliminary scan that determines the
1502        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1503        that the computed length was too small. The symptoms of this bug were
1504        either the PCRE error "internal error: code overflow" from pcre_compile(),
1505        or a glibc crash with a message such as "pcretest: free(): invalid next
1506        size (fast)". Examples of patterns that provoked this bug (shown in
1507        pcretest format) are:
1508    
1509          /(?-x: )/x
1510          /(?x)(?-x: \s*#\s*)/
1511          /((?i)[\x{c0}])/8
1512          /(?i:[\x{c0}])/8
1513    
1514        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1515        is now done differently.
1516    
1517     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1518        wrapper classes; (b) implement a new function in the C++ scanner that is
1519        more efficient than the old way of doing things because it avoids levels of
1520        recursion in the regex matching; (c) add a paragraph to the documentation
1521        for the FullMatch() function.
1522    
1523     6. The escape sequence \n was being treated as whatever was defined as
1524        "newline". Not only was this contrary to the documentation, which states
1525        that \n is character 10 (hex 0A), but it also went horribly wrong when
1526        "newline" was defined as CRLF. This has been fixed.
1527    
1528     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1529        was being set to -1 for the "end of line" case (supposedly a value that no
1530        character can have). Though this value is never used (the check for end of
1531        line is "zero bytes in current character"), it caused compiler complaints.
1532        I've changed it to 0xffffffff.
1533    
1534     8. In pcre_version.c, the version string was being built by a sequence of
1535        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1536        string (as it is for production releases) called a macro with an empty
1537        argument. The C standard says the result of this is undefined. The gcc
1538        compiler treats it as an empty string (which was what was wanted) but it is
1539        reported that Visual C gives an error. The source has been hacked around to
1540        avoid this problem.
1541    
1542     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1543        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1544        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1545        of them did).
1546    
1547    10. Originally, pcretest opened its input and output without "b"; then I was
1548        told that "b" was needed in some environments, so it was added for release
1549        5.0 to both the input and output. (It makes no difference on Unix-like
1550        systems.) Later I was told that it is wrong for the input on Windows. I've
1551        now abstracted the modes into two macros, to make it easier to fiddle with
1552        them, and removed "b" from the input mode under Windows.
1553    
1554    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1555    
1556    12. Added -help and --help to pcretest as an official way of being reminded
1557        of the options.
1558    
1559    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1560        and pcrecpp.cc because they annoy compilers at high warning levels.
1561    
1562    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1563    
1564    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1565        scripts are not C programs :-) and which was not noticed because it works
1566        on Linux.
1567    
1568    16. pcretest is supposed to handle any length of pattern and data line (as one
1569        line or as a continued sequence of lines) by extending its input buffer if
1570        necessary. This feature was broken for very long pattern lines, leading to
1571        a string of junk being passed to pcre_compile() if the pattern was longer
1572        than about 50K.
1573    
1574    17. I have done a major re-factoring of the way pcre_compile() computes the
1575        amount of memory needed for a compiled pattern. Previously, there was code
1576        that made a preliminary scan of the pattern in order to do this. That was
1577        OK when PCRE was new, but as the facilities have expanded, it has become
1578        harder and harder to keep it in step with the real compile phase, and there
1579        have been a number of bugs (see for example, 4 above). I have now found a
1580        cunning way of running the real compile function in a "fake" mode that
1581        enables it to compute how much memory it would need, while actually only
1582        ever using a few hundred bytes of working memory and without too many
1583        tests of the mode. This should make future maintenance and development
1584        easier. A side effect of this work is that the limit of 200 on the nesting
1585        depth of parentheses has been removed (though this was never a serious
1586        limitation, I suspect). However, there is a downside: pcre_compile() now
1587        runs more slowly than before (30% or more, depending on the pattern). I
1588        hope this isn't a big issue. There is no effect on runtime performance.
1589    
1590    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1591        newline (only possible for the last line of a file) and it was a
1592        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1593    
1594    19. Added additional timing features to pcretest. (1) The -tm option now times
1595        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1596        separate command line item, by a number that specifies the number of
1597        repeats to use when timing. The default is 50000; this gives better
1598        precision, but takes uncomfortably long for very large patterns.
1599    
1600    20. Extended pcre_study() to be more clever in cases where a branch of a
1601        subpattern has no definite first character. For example, (a*|b*)[cd] would
1602        previously give no result from pcre_study(). Now it recognizes that the
1603        first character must be a, b, c, or d.
1604    
1605    21. There was an incorrect error "recursive call could loop indefinitely" if
1606        a subpattern (or the entire pattern) that was being tested for matching an
1607        empty string contained only one non-empty item after a nested subpattern.
1608        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1609        incorrectly, because the \d was being skipped in the check.
1610    
1611    22. The pcretest program now has a new pattern option /B and a command line
1612        option -b, which is equivalent to adding /B to every pattern. This causes
1613        it to show the compiled bytecode, without the additional information that
1614        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1615        is the same as /B/I).
1616    
1617    23. A new optimization is now able automatically to treat some sequences such
1618        as a*b as a*+b. More specifically, if something simple (such as a character
1619        or a simple class like \d) has an unlimited quantifier, and is followed by
1620        something that cannot possibly match the quantified thing, the quantifier
1621        is automatically "possessified".
1622    
1623    24. A recursive reference to a subpattern whose number was greater than 39
1624        went wrong under certain circumstances in UTF-8 mode. This bug could also
1625        have affected the operation of pcre_study().
1626    
1627    25. Realized that a little bit of performance could be had by replacing
1628        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1629    
1630    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1631    
1632    27. Possessive quantifiers such as a++ were previously implemented by turning
1633        them into atomic groups such as ($>a+). Now they have their own opcodes,
1634        which improves performance. This includes the automatically created ones
1635        from 23 above.
1636    
1637    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1638        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1639        the first matched character to be a colon. This applied both to named and
1640        numbered groups.
1641    
1642    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1643    
1644    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1645        building PCRE on Windows easier, so I have included it in the distribution.
1646    
1647    31. There is now a check in pcretest against a ridiculously large number being
1648        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1649        loop, the loop is abandoned.
1650    
1651    32. Forward references to subpatterns in conditions such as (?(2)...) where
1652        subpattern 2 is defined later cause pcre_compile() to search forwards in
1653        the pattern for the relevant set of parentheses. This search went wrong
1654        when there were unescaped parentheses in a character class, parentheses
1655        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1656    
1657    33. "Subroutine" calls and backreferences were previously restricted to
1658        referencing subpatterns earlier in the regex. This restriction has now
1659        been removed.
1660    
1661    34. Added a number of extra features that are going to be in Perl 5.10. On the
1662        whole, these are just syntactic alternatives for features that PCRE had
1663        previously implemented using the Python syntax or my own invention. The
1664        other formats are all retained for compatibility.
1665    
1666        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1667            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1668            also .NET compatible.
1669    
1670        (b) A recursion or subroutine call to a named group can now be defined as
1671            (?&name) as well as (?P>name).
1672    
1673        (c) A backreference to a named group can now be defined as \k<name> or
1674            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1675            5.10, are also .NET compatible.
1676    
1677        (d) A conditional reference to a named group can now use the syntax
1678            (?(<name>) or (?('name') as well as (?(name).
1679    
1680        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1681            groups (named and numbered) that are never evaluated inline, but can be
1682            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1683            is always false. There may be only one alternative in such a group.
1684    
1685        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1686            as the simple (?(R). The condition is true only if the most recent
1687            recursion is that of the given number or name. It does not search out
1688            through the entire recursion stack.
1689    
1690        (g) The escape \gN or \g{N} has been added, where N is a positive or
1691            negative number, specifying an absolute or relative reference.
1692    
1693    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1694        some "unreachable code" warnings.
1695    
1696    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1697        things, this adds five new scripts.
1698    
1699    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1700        There were also incompatibilities regarding the handling of \Q..\E inside
1701        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1702        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1703    
1704    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1705        matches an empty string, and forcibly breaks the loop. There were bugs in
1706        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1707        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1708        separate and independent bugs (that affected different cases) have been
1709        fixed.
1710    
1711    39. Refactored the code to abolish the use of different opcodes for small
1712        capturing bracket numbers. This is a tidy that I avoided doing when I
1713        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1714        The new approach is not only tidier, it makes it possible to reduce the
1715        memory needed to fix the previous bug (38).
1716    
1717    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1718        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1719        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1720        mode.
1721    
1722    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1723        report.
1724    
1725    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1726        copy construction and assignment in the C++ wrapper.
1727    
1728    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1729        couple of bugs that could have given wrong results in the "--newline=crlf"
1730        case.
1731    
1732    44. Added a number of casts and did some reorganization of signed/unsigned int
1733        variables following suggestions from Dair Grant. Also renamed the variable
1734        "this" as "item" because it is a C++ keyword.
1735    
1736    45. Arranged for dftables to add
1737    
1738          #include "pcre_internal.h"
1739    
1740        to pcre_chartables.c because without it, gcc 4.x may remove the array
1741        definition from the final binary if PCRE is built into a static library and
1742        dead code stripping is activated.
1743    
1744    46. For an unanchored pattern, if a match attempt fails at the start of a
1745        newline sequence, and the newline setting is CRLF or ANY, and the next two
1746        characters are CRLF, advance by two characters instead of one.
1747    
1748    
1749    Version 6.7 04-Jul-06
1750    ---------------------
1751    
1752     1. In order to handle tests when input lines are enormously long, pcretest has
1753        been re-factored so that it automatically extends its buffers when
1754        necessary. The code is crude, but this _is_ just a test program. The
1755        default size has been increased from 32K to 50K.
1756    
1757     2. The code in pcre_study() was using the value of the re argument before
1758        testing it for NULL. (Of course, in any sensible call of the function, it
1759        won't be NULL.)
1760    
1761     3. The memmove() emulation function in pcre_internal.h, which is used on
1762        systems that lack both memmove() and bcopy() - that is, hardly ever -
1763        was missing a "static" storage class specifier.
1764    
1765     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1766        containing an extended class (one that cannot be represented by a bitmap
1767        because it contains high-valued characters or Unicode property items, e.g.
1768        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1769        pattern, but PCRE should not loop if you do not (it no longer does).
1770        [Detail: two cases were found: (a) a repeated subpattern containing an
1771        extended class; (b) a recursive reference to a subpattern that followed a
1772        previous extended class. It wasn't skipping over the extended class
1773        correctly when UTF-8 mode was not set.]
1774    
1775     5. A negated single-character class was not being recognized as fixed-length
1776        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1777        compile error "lookbehind assertion is not fixed length".
1778    
1779     6. The RunPerlTest auxiliary script was showing an unexpected difference
1780        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1781        write a Perl script that can interpret lines of an input file either as
1782        byte characters or as UTF-8, which is what "perltest" was being required to
1783        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1784        can't do is switch easily at run time between having the "use utf8;" pragma
1785        or not. In the end, I fudged it by using the RunPerlTest script to insert
1786        "use utf8;" explicitly for the UTF-8 tests.
1787    
1788     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1789        the end of the subject string, contrary to the documentation and to what
1790        Perl does. This was true of both matching functions. Now it matches only at
1791        the start of the subject and immediately after *internal* newlines.
1792    
1793     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1794        a pointer to an int instead of a pointer to an unsigned long int. This
1795        caused problems on 64-bit systems.
1796    
1797     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1798        instance of the 'standard' template library not being so standard".
1799    
1800    10. There was no check on the number of named subpatterns nor the maximum
1801        length of a subpattern name. The product of these values is used to compute
1802        the size of the memory block for a compiled pattern. By supplying a very
1803        long subpattern name and a large number of named subpatterns, the size
1804        computation could be caused to overflow. This is now prevented by limiting
1805        the length of names to 32 characters, and the number of named subpatterns
1806        to 10,000.
1807    
1808    11. Subpatterns that are repeated with specific counts have to be replicated in
1809        the compiled pattern. The size of memory for this was computed from the
1810        length of the subpattern and the repeat count. The latter is limited to
1811        65535, but there was no limit on the former, meaning that integer overflow
1812        could in principle occur. The compiled length of a repeated subpattern is
1813        now limited to 30,000 bytes in order to prevent this.
1814    
1815    12. Added the optional facility to have named substrings with the same name.
1816    
1817    13. Added the ability to use a named substring as a condition, using the
1818        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1819        are numbers (not recommended). Forward references are permitted.
1820    
1821    14. Added forward references in named backreferences (if you see what I mean).
1822    
1823    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1824        pattern could run off the end of the subject. For example, the pattern
1825        "(?s)(.{1,5})"8 did this with the subject "ab".
1826    
1827    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1828        PCRE_CASELESS was set when matching characters that were quantified with ?
1829        or *.
1830    
1831    17. A character class other than a single negated character that had a minimum
1832        but no maximum quantifier - for example [ab]{6,} - was not handled
1833        correctly by pce_dfa_exec(). It would match only one character.
1834    
1835    18. A valid (though odd) pattern that looked like a POSIX character
1836        class but used an invalid character after [ (for example [[,abc,]]) caused
1837        pcre_compile() to give the error "Failed: internal error: code overflow" or
1838        in some cases to crash with a glibc free() error. This could even happen if
1839        the pattern terminated after [[ but there just happened to be a sequence of
1840        letters, a binary zero, and a closing ] in the memory that followed.
1841    
1842    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1843        over the years. Originally (before any Unicode support), just the bottom 8
1844        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1845        output from "man perlunicode" includes this:
1846    
1847          The regular expression compiler produces polymorphic opcodes.  That
1848          is, the pattern adapts to the data and automatically switches to
1849          the Unicode character scheme when presented with Unicode data--or
1850          instead uses a traditional byte scheme when presented with byte
1851          data.
1852    
1853        Sadly, a wide octal escape does not cause a switch, and in a string with
1854        no other multibyte characters, these octal escapes are treated as before.
1855        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1856        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1857        Unicode string.
1858    
1859        I have not perpetrated such confusion in PCRE. Up till now, it took just
1860        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1861        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1862        translate to the appropriate multibyte character.
1863    
1864    29. Applied some refactoring to reduce the number of warnings from Microsoft
1865        and Borland compilers. This has included removing the fudge introduced
1866        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1867        a warning about an unused variable.
1868    
1869    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1870        characters since release 4.0, because Perl (from release 5.004) does not.
1871        [Or at least, is documented not to: some releases seem to be in conflict
1872        with the documentation.] However, when a pattern was studied with
1873        pcre_study() and all its branches started with \s, PCRE still included VT
1874        as a possible starting character. Of course, this did no harm; it just
1875        caused an unnecessary match attempt.
1876    
1877    22. Removed a now-redundant internal flag bit that recorded the fact that case
1878        dependency changed within the pattern. This was once needed for "required
1879        byte" processing, but is no longer used. This recovers a now-scarce options
1880        bit. Also moved the least significant internal flag bit to the most-
1881        significant bit of the word, which was not previously used (hangover from
1882        the days when it was an int rather than a uint) to free up another bit for
1883        the future.
1884    
1885    23. Added support for CRLF line endings as well as CR and LF. As well as the
1886        default being selectable at build time, it can now be changed at runtime
1887        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1888        specify that it is scanning data with non-default line endings.
1889    
1890    24. Changed the definition of CXXLINK to make it agree with the definition of
1891        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1892    
1893    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1894        recursions. This makes a big different to stack usage for some patterns.
1895    
1896    26. If a subpattern containing a named recursion or subroutine reference such
1897        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1898        the space required for the compiled pattern went wrong and gave too small a
1899        value. Depending on the environment, this could lead to "Failed: internal
1900        error: code overflow at offset 49" or "glibc detected double free or
1901        corruption" errors.
1902    
1903    27. Applied patches from Google (a) to support the new newline modes and (b) to
1904        advance over multibyte UTF-8 characters in GlobalReplace.
1905    
1906    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1907        difference for some implementation of PCRE in some Windows version.
1908    
1909    29. Added some extra testing facilities to pcretest:
1910    
1911        \q<number>   in a data line sets the "match limit" value
1912        \Q<number>   in a data line sets the "match recursion limt" value
1913        -S <number>  sets the stack size, where <number> is in megabytes
1914    
1915        The -S option isn't available for Windows.
1916    
1917    
1918    Version 6.6 06-Feb-06
1919    ---------------------
1920    
1921     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1922        in pcreposix.h. I have copied the definition from pcre.h.
1923    
1924     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1925        because pcre.h is no longer a built file.
1926    
1927     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1928        not normally included in the compiled code.
1929    
1930    
1931    Version 6.5 01-Feb-06
1932    ---------------------
1933    
1934     1. When using the partial match feature with pcre_dfa_exec(), it was not
1935        anchoring the second and subsequent partial matches at the new starting
1936        point. This could lead to incorrect results. For example, with the pattern
1937        /1234/, partially matching against "123" and then "a4" gave a match.
1938    
1939     2. Changes to pcregrep:
1940    
1941        (a) All non-match returns from pcre_exec() were being treated as failures
1942            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1943            error message is output. Some extra information is given for the
1944            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1945            probably the only errors that are likely to be caused by users (by
1946            specifying a regex that has nested indefinite repeats, for instance).
1947            If there are more than 20 of these errors, pcregrep is abandoned.
1948    
1949        (b) A binary zero was treated as data while matching, but terminated the
1950            output line if it was written out. This has been fixed: binary zeroes
1951            are now no different to any other data bytes.
1952    
1953        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1954            used to set a locale for matching. The --locale=xxxx long option has
1955            been added (no short equivalent) to specify a locale explicitly on the
1956            pcregrep command, overriding the environment variables.
1957    
1958        (d) When -B was used with -n, some line numbers in the output were one less
1959            than they should have been.
1960    
1961        (e) Added the -o (--only-matching) option.
1962    
1963        (f) If -A or -C was used with -c (count only), some lines of context were
1964            accidentally printed for the final match.
1965    
1966        (g) Added the -H (--with-filename) option.
1967    
1968        (h) The combination of options -rh failed to suppress file names for files
1969            that were found from directory arguments.
1970    
1971        (i) Added the -D (--devices) and -d (--directories) options.
1972    
1973        (j) Added the -F (--fixed-strings) option.
1974    
1975        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1976    
1977        (l) Added the --colo(u)r option.
1978    
1979        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1980            is not present by default.
1981    
1982     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1983        items such as (?R) or (?1), when the recursion could match a number of
1984        alternatives. If it matched one of the alternatives, but subsequently,
1985        outside the recursion, there was a failure, the code tried to back up into
1986        the recursion. However, because of the way PCRE is implemented, this is not
1987        possible, and the result was an incorrect result from the match.
1988    
1989        In order to prevent this happening, the specification of recursion has
1990        been changed so that all such subpatterns are automatically treated as
1991        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1992    
1993     4. I had overlooked the fact that, in some locales, there are characters for
1994        which isalpha() is true but neither isupper() nor islower() are true. In
1995        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1996        and ordfeminine) are like this. This affected the treatment of \w and \W
1997        when they appeared in character classes, but not when they appeared outside
1998        a character class. The bit map for "word" characters is now created
1999        separately from the results of isalnum() instead of just taking it from the
2000        upper, lower, and digit maps. (Plus the underscore character, of course.)
2001    
2002     5. The above bug also affected the handling of POSIX character classes such as
2003        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2004        permanent tables. Instead, the bit maps for such a class were previously
2005        created as the appropriate unions of the upper, lower, and digit bitmaps.
2006        Now they are created by subtraction from the [[:word:]] class, which has
2007        its own bitmap.
2008    
2009     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2010        It is created by subtracting the vertical space characters (\x09, \x0a,
2011        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2012        subtraction was done in the overall bitmap for a character class, meaning
2013        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2014        be recognized. This bug has been fixed.
2015    
2016     7. Patches from the folks at Google:
2017    
2018          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2019          real life, but is still worth protecting against".
2020    
2021          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2022          regular expressions".
2023    
2024          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2025          have it.
2026    
2027          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2028          "configure" and the latter not, in order to fix a problem somebody had
2029          with compiling the Arg class on HP-UX.
2030    
2031          (e) Improve the error-handling of the C++ wrapper a little bit.
2032    
2033          (f) New tests for checking recursion limiting.
2034    
2035     8. The pcre_memmove() function, which is used only if the environment does not
2036        have a standard memmove() function (and is therefore rarely compiled),
2037        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2038        returning a result (though PCRE never actually uses the result).
2039    
2040     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2041        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2042        returned instead of calling malloc() with an overflowing number that would
2043        most likely cause subsequent chaos.
2044    
2045    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2046    
2047    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2048        with this option is matched, the nmatch and pmatch options of regexec() are
2049        ignored.
2050    
2051    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2052        provided in case anyone wants to the the POSIX interface with UTF-8
2053        strings.
2054    
2055    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2056        C++ linking (needed for some HP-UX environments).
2057    
2058    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2059        (unused parameter) and in the pcre_printint() function (omitted "default"
2060        switch label when the default is to do nothing).
2061    
2062    15. Added some code to make it possible, when PCRE is compiled as a C++
2063        library, to replace subject pointers for pcre_exec() with a smart pointer
2064        class, thus making it possible to process discontinuous strings.
2065    
2066    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2067        much the same function. They were added by different people who were trying
2068        to make PCRE easy to compile on non-Unix systems. It has been suggested
2069        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2070        for compiling on Windows systems. I have therefore replaced it with
2071        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2072        defaults to "extern" for C or "extern C" for C++, which works fine on
2073        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2074        SCOPE with something explicit in config.h. In addition:
2075    
2076        (a) pcreposix.h still had just "extern" instead of either of these macros;
2077            I have replaced it with PCRE_DATA_SCOPE.
2078    
2079        (b) Functions such as _pcre_xclass(), which are internal to the library,
2080            but external in the C sense, all had PCRE_EXPORT in their definitions.
2081            This is apparently wrong for the Windows case, so I have removed it.
2082            (It makes no difference on Unix-like systems.)
2083    
2084    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2085        of recursive calls to match(). This is different to MATCH_LIMIT because
2086        that limits the total number of calls to match(), not all of which increase
2087        the depth of recursion. Limiting the recursion depth limits the amount of
2088        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2089        when PCRE is compiled, and changed at run time. A patch from Google adds
2090        this functionality to the C++ interface.
2091    
2092    18. Changes to the handling of Unicode character properties:
2093    
2094        (a) Updated the table to Unicode 4.1.0.
2095    
2096        (b) Recognize characters that are not in the table as "Cn" (undefined).
2097    
2098        (c) I revised the way the table is implemented to a much improved format
2099            which includes recognition of ranges. It now supports the ranges that
2100            are defined in UnicodeData.txt, and it also amalgamates other
2101            characters into ranges. This has reduced the number of entries in the
2102            table from around 16,000 to around 3,000, thus reducing its size
2103            considerably. I realized I did not need to use a tree structure after
2104            all - a binary chop search is just as efficient. Having reduced the
2105            number of entries, I extended their size from 6 bytes to 8 bytes to
2106            allow for more data.
2107    
2108        (d) Added support for Unicode script names via properties such as \p{Han}.
2109    
2110    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2111        matching that character.
2112    
2113    20. When matching a repeated Unicode property with a minimum greater than zero,
2114        (for example \pL{2,}), PCRE could look past the end of the subject if it
2115        reached it while seeking the minimum number of characters. This could
2116        happen only if some of the characters were more than one byte long, because
2117        there is a check for at least the minimum number of bytes.
2118    
2119    21. Refactored the implementation of \p and \P so as to be more general, to
2120        allow for more different types of property in future. This has changed the
2121        compiled form incompatibly. Anybody with saved compiled patterns that use
2122        \p or \P will have to recompile them.
2123    
2124    22. Added "Any" and "L&" to the supported property types.
2125    
2126    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2127        but give a compile time error if the value is greater than 0xff.
2128    
2129    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2130        accidentally not being installed or uninstalled.
2131    
2132    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2133        made were to insert the current release number. This seemed silly, because
2134        it made things harder for people building PCRE on systems that don't run
2135        "configure". I have turned pcre.h into a distributed file, no longer built
2136        by "configure", with the version identification directly included. There is
2137        no longer a pcre.h.in file.
2138    
2139        However, this change necessitated a change to the pcre-config script as
2140        well. It is built from pcre-config.in, and one of the substitutions was the
2141        release number. I have updated configure.ac so that ./configure now finds
2142        the release number by grepping pcre.h.
2143    
2144    26. Added the ability to run the tests under valgrind.
2145    
2146    
2147    Version 6.4 05-Sep-05
2148    ---------------------
2149    
2150     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2151        "--" to be printed when multiple files were scanned, even when none of the
2152        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2153        consider it to be a bug, and have restored the previous behaviour.
2154    
2155     2. A couple of code tidies to get rid of compiler warnings.
2156    
2157     3. The pcretest program used to cheat by referring to symbols in the library
2158        whose names begin with _pcre_. These are internal symbols that are not
2159        really supposed to be visible externally, and in some environments it is
2160        possible to suppress them. The cheating is now confined to including
2161        certain files from the library's source, which is a bit cleaner.
2162    
2163     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2164        file's purpose clearer.
2165    
2166     5. Reorganized pcre_ucp_findchar().
2167    
2168    
2169    Version 6.3 15-Aug-05
2170    ---------------------
2171    
2172     1. The file libpcre.pc.in did not have general read permission in the tarball.
2173    
2174     2. There were some problems when building without C++ support:
2175    
2176        (a) If C++ support was not built, "make install" and "make test" still
2177            tried to test it.
2178    
2179        (b) There were problems when the value of CXX was explicitly set. Some
2180            changes have been made to try to fix these, and ...
2181    
2182        (c) --disable-cpp can now be used to explicitly disable C++ support.
2183    
2184        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2185            backslash in a target when C++ was disabled. This confuses some
2186            versions of "make", apparently. Using an intermediate variable solves
2187            this. (Same for CPP_LOBJ.)
2188    
2189     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2190        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2191        necessary on certain architectures.
2192    
2193     4. Added a setting of -export-symbols-regex to the link command to remove
2194        those symbols that are exported in the C sense, but actually are local
2195        within the library, and not documented. Their names all begin with
2196        "_pcre_". This is not a perfect job, because (a) we have to except some
2197        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2198        available (and never for static libraries). I have made a note to try to
2199        find a way round (a) in the future.
2200    
2201    
2202    Version 6.2 01-Aug-05
2203    ---------------------
2204    
2205     1. There was no test for integer overflow of quantifier values. A construction
2206        such as {1111111111111111} would give undefined results. What is worse, if
2207        a minimum quantifier for a parenthesized subpattern overflowed and became
2208        negative, the calculation of the memory size went wrong. This could have
2209        led to memory overwriting.
2210    
2211     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2212    
2213     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2214        operating environments where this matters.
2215    
2216     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2217        PCRE options from within the C++ wrapper.
2218    
2219     5. Named capturing subpatterns were not being correctly counted when a pattern
2220        was compiled. This caused two problems: (a) If there were more than 100
2221        such subpatterns, the calculation of the memory needed for the whole
2222        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2223        back references of the form \12, where the number was greater than 9, were
2224        not recognized as back references, even though there were sufficient
2225        previous subpatterns.
2226    
2227     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2228        versions of gcc, e.g. 2.95.4.
2229    
2230    
2231    Version 6.1 21-Jun-05
2232    ---------------------
2233    
2234     1. There was one reference to the variable "posix" in pcretest.c that was not
2235        surrounded by "#if !defined NOPOSIX".
2236    
2237     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2238        the cross-check on the old pcre_info() function, for the benefit of the
2239        cut-down version of PCRE that is currently imported into Exim.
2240    
2241     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2242        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2243        patterns that start (?i) but I don't think that matters. The use of (?i) is
2244        just an example; this all applies to the other options as well.
2245    
2246     4. Since libtool seems to echo the compile commands it is issuing, the output
2247        from "make" can be reduced a bit by putting "@" in front of each libtool
2248        compile command.
2249    
2250     5. Patch from the folks at Google for configure.in to be a bit more thorough
2251        in checking for a suitable C++ installation before trying to compile the
2252        C++ stuff. This should fix a reported problem when a compiler was present,
2253        but no suitable headers.
2254    
2255     6. The man pages all had just "PCRE" as their title. I have changed them to
2256        be the relevant file name. I have also arranged that these names are
2257        retained in the file doc/pcre.txt, which is a concatenation in text format
2258        of all the man pages except the little individual ones for each function.
2259    
2260     7. The NON-UNIX-USE file had not been updated for the different set of source
2261        files that come with release 6. I also added a few comments about the C++
2262        wrapper.
2263    
2264    
2265    Version 6.0 07-Jun-05
2266    ---------------------
2267    
2268     1. Some minor internal re-organization to help with my DFA experiments.
2269    
2270     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2271        didn't matter for the library itself when fully configured, but did matter
2272        when compiling without UCP support, or within Exim, where the ucp files are
2273        not imported.
2274    
2275     3. Refactoring of the library code to split up the various functions into
2276        different source modules. The addition of the new DFA matching code (see
2277        below) to a single monolithic source would have made it really too
2278        unwieldy, quite apart from causing all the code to be include in a
2279        statically linked application, when only some functions are used. This is
2280        relevant even without the DFA addition now that patterns can be compiled in
2281        one application and matched in another.
2282    
2283        The downside of splitting up is that there have to be some external
2284        functions and data tables that are used internally in different modules of
2285        the library but which are not part of the API. These have all had their
2286        names changed to start with "_pcre_" so that they are unlikely to clash
2287        with other external names.
2288    
2289     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2290        a different (DFA) algorithm. Although it is slower than the original
2291        function, it does have some advantages for certain types of matching
2292        problem.
2293    
2294     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2295        including restarting after a partial match.
2296    
2297     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2298        defined when compiling for Windows was sent to me. I have put it into the
2299        code, though I have no means of testing or verifying it.
2300    
2301     7. Added the pcre_refcount() auxiliary function.
2302    
2303     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2304        match before or at the first newline in the subject string. In pcretest,
2305        the /f option on a pattern can be used to set this.
2306    
2307     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2308        would behave wrongly. This has been present in PCRE since release 4.0.
2309    
2310    10. A number of changes to the pcregrep command:
2311    
2312        (a) Refactored how -x works; insert ^(...)$ instead of setting
2313            PCRE_ANCHORED and checking the length, in preparation for adding
2314            something similar for -w.
2315    
2316        (b) Added the -w (match as a word) option.
2317    
2318        (c) Refactored the way lines are read and buffered so as to have more
2319            than one at a time available.
2320    
2321        (d) Implemented a pcregrep test script.
2322    
2323        (e) Added the -M (multiline match) option. This allows patterns to match
2324            over several lines of the subject. The buffering ensures that at least
2325            8K, or the rest of the document (whichever is the shorter) is available
2326            for matching (and similarly the previous 8K for lookbehind assertions).
2327    
2328        (f) Changed the --help output so that it now says
2329    
2330              -w, --word-regex(p)
2331    
2332            instead of two lines, one with "regex" and the other with "regexp"
2333            because that confused at least one person since the short forms are the
2334            same. (This required a bit of code, as the output is generated
2335            automatically from a table. It wasn't just a text change.)
2336    
2337        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2338            option but starts with a hyphen. Could be a pattern or a path name
2339            starting with a hyphen, for instance.
2340    
2341        (h) "-" can be given as a file name to represent stdin.
2342    
2343        (i) When file names are being printed, "(standard input)" is used for
2344            the standard input, for compatibility with GNU grep. Previously
2345            "<stdin>" was used.
2346    
2347        (j) The option --label=xxx can be used to supply a name to be used for
2348            stdin when file names are being printed. There is no short form.
2349    
2350        (k) Re-factored the options decoding logic because we are going to add
2351            two more options that take data. Such options can now be given in four
2352            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2353    
2354        (l) Added the -A, -B, and -C options for requesting that lines of context
2355            around matches be printed.
2356    
2357        (m) Added the -L option to print the names of files that do not contain
2358            any matching lines, that is, the complement of -l.
2359    
2360        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2361            continue to scan other files.
2362    
2363        (o) The -s option was incorrectly implemented. For compatibility with other
2364            greps, it now suppresses the error message for a non-existent or non-
2365            accessible file (but not the return code). There is a new option called
2366            -q that suppresses the output of matching lines, which was what -s was
2367            previously doing.
2368    
2369        (p) Added --include and --exclude options to specify files for inclusion
2370            and exclusion when recursing.
2371    
2372    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2373        Hopefully, it now does.
2374    
2375    12. Missing cast in pcre_study().
2376    
2377    13. Added an "uninstall" target to the makefile.
2378    
2379    14. Replaced "extern" in the function prototypes in Makefile.in with
2380        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2381        world, but is set differently for Windows.
2382    
2383    15. Added a second compiling function called pcre_compile2(). The only
2384        difference is that it has an extra argument, which is a pointer to an
2385        integer error code. When there is a compile-time failure, this is set
2386        non-zero, in addition to the error test pointer being set to point to an
2387        error message. The new argument may be NULL if no error number is required
2388        (but then you may as well call pcre_compile(), which is now just a
2389        wrapper). This facility is provided because some applications need a
2390        numeric error indication, but it has also enabled me to tidy up the way
2391        compile-time errors are handled in the POSIX wrapper.
2392    
2393    16. Added VPATH=.libs to the makefile; this should help when building with one
2394        prefix path and installing with another. (Or so I'm told by someone who
2395        knows more about this stuff than I do.)
2396    
2397    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2398        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2399        match everything, including newlines. This is not POSIX-compatible, but
2400        somebody wanted the feature. From pcretest it can be activated by using
2401        both the P and the s flags.
2402    
2403    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2404    
2405    19. libpcre.pc was being incorrectly installed as executable.
2406    
2407    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2408        it now also looks for '\r' so that it will work unmodified on Windows.
2409    
2410    21. Added Google's contributed C++ wrapper to the distribution.
2411    
2412    22. Added some untidy missing memory free() calls in pcretest, to keep
2413        Electric Fence happy when testing.
2414    
2415    
2416    
2417    Version 5.0 13-Sep-04
2418    ---------------------
2419    
2420     1. Internal change: literal characters are no longer packed up into items
2421        containing multiple characters in a single byte-string. Each character
2422        is now matched using a separate opcode. However, there may be more than one
2423        byte in the character in UTF-8 mode.
2424    
2425     2. The pcre_callout_block structure has two new fields: pattern_position and
2426        next_item_length. These contain the offset in the pattern to the next match
2427        item, and its length, respectively.
2428    
2429     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2430        insertion of callouts before each pattern item. Added the /C option to
2431        pcretest to make use of this.
2432    
2433     4. On the advice of a Windows user, the lines
2434    
2435          #if defined(_WIN32) || defined(WIN32)
2436          _setmode( _fileno( stdout ), 0x8000 );
2437          #endif  /* defined(_WIN32) || defined(WIN32) */
2438    
2439        have been added to the source of pcretest. This apparently does useful
2440        magic in relation to line terminators.
2441    
2442     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2443        for the benefit of those environments where the "b" makes a difference.
2444    
2445     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2446        to know about it. I have put a hack into configure.in that adds in code
2447        to set GCC=yes if CC=icc. This seems to end up at a point in the
2448        generated configure script that is early enough to affect the setting of
2449        compiler options, which is what is needed, but I have no means of testing
2450        whether it really works. (The user who reported this had patched the
2451        generated configure script, which of course I cannot do.)
2452    
2453        LATER: After change 22 below (new libtool files), the configure script
2454        seems to know about icc (and also ecc). Therefore, I have commented out
2455        this hack in configure.in.
2456    
2457     7. Added support for pkg-config (2 patches were sent in).
2458    
2459     8. Negated POSIX character classes that used a combination of internal tables
2460        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2461        [[:^ascii]]. Typically, they would match almost any characters. The other
2462        POSIX classes were not broken in this way.
2463    
2464     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2465        to find the match, as PCRE was deluded into thinking that the match had to
2466        start at the start point or following a newline. The same bug applied to
2467        patterns with negative forward assertions or any backward assertions
2468        preceding ".*" at the start, unless the pattern required a fixed first
2469        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2470    
2471    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2472        starting at the last subject character, bytes beyond the end of the subject
2473        string were read.
2474    
2475    11. Renamed the variable "class" as "classbits" to make life easier for C++
2476        users. (Previously there was a macro definition, but it apparently wasn't
2477        enough.)
2478    
2479    12. Added the new field "tables" to the extra data so that tables can be passed
2480        in at exec time, or the internal tables can be re-selected. This allows
2481        a compiled regex to be saved and re-used at a later time by a different
2482        program that might have everything at different addresses.
2483    
2484    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2485        -R library as well as a -L library.
2486    
2487    14. The debugging options of pcretest (-d on the command line or D on a
2488        pattern) showed incorrect output for anything following an extended class
2489        that contained multibyte characters and which was followed by a quantifier.
2490    
2491    15. Added optional support for general category Unicode character properties
2492        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2493        support. It adds about 90K to the size of the library. The meanings of the
2494        inbuilt class escapes such as \d and \s have NOT been changed.
2495    
2496    16. Updated pcredemo.c to include calls to free() to release the memory for the
2497        compiled pattern.
2498    
2499    17. The generated file chartables.c was being created in the source directory
2500        instead of in the building directory. This caused the build to fail if the
2501        source directory was different from the building directory, and was
2502        read-only.
2503    
2504    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2505        file. No doubt somebody will tell me if they don't make sense... Also added
2506        Dan Mooney's comments about building on OpenVMS.
2507    
2508    19. Added support for partial matching via the PCRE_PARTIAL option for
2509        pcre_exec() and the \P data escape in pcretest.
2510    
2511    20. Extended pcretest with 3 new pattern features:
2512    
2513        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2514              write the compiled pattern to the file whose name is "rest-of-line".
2515              This is a straight binary dump of the data, with the saved pointer to
2516              the character tables forced to be NULL. The study data, if any, is
2517              written too. After writing, pcretest reads a new pattern.
2518    
2519        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2520              compiled pattern from the given file. There must not be any
2521              occurrences of "<" in the file name (pretty unlikely); if there are,
2522              pcretest will instead treat the initial "<" as a pattern delimiter.
2523              After reading in the pattern, pcretest goes on to read data lines as
2524              usual.
2525    
2526        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2527              and 16-bit fields in a compiled pattern, to simulate a pattern that
2528              was compiled on a host of opposite endianness.
2529    
2530    21. The pcre-exec() function can now cope with patterns that were compiled on
2531        hosts of opposite endianness, with this restriction:
2532    
2533          As for any compiled expression that is saved and used later, the tables
2534          pointer field cannot be preserved; the extra_data field in the arguments
2535          to pcre_exec() should be used to pass in a tables address if a value
2536          other than the default internal tables were used at compile time.
2537    
2538    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2539        now diagnosed as an error. Previously, most of the time, a negative number
2540        would have been treated as zero, but if in addition "ovector" was passed as
2541        NULL, a crash could occur.
2542    
2543    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2544        new versions from the libtool 1.5 distribution (the last one is a copy of
2545        a file called libtool.m4). This seems to have fixed the need to patch
2546        "configure" to support Darwin 1.3 (which I used to do). However, I still
2547        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2548        workstation).
2549    
2550    24. Changed the PCRE licence to be the more standard "BSD" licence.
2551    
2552    
2553    Version 4.5 01-Dec-03
2554    ---------------------
2555    
2556     1. There has been some re-arrangement of the code for the match() function so
2557        that it can be compiled in a version that does not call itself recursively.
2558        Instead, it keeps those local variables that need separate instances for
2559        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2560        needs to "recurse". Keeping track of where control must go is done by means
2561        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2562        hide most of the details from the main code, and operates only if
2563        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2564        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2565        operating.
2566    
2567        To make it easier for callers to provide specially tailored get/free
2568        functions for this usage, two new functions, pcre_stack_malloc, and
2569        pcre_stack_free, are used. They are always called in strict stacking order,
2570        and the size of block requested is always the same.
2571    
2572        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2573        PCRE has been compiled to use the stack or the heap for recursion. The
2574        -C option of pcretest uses this to show which version is compiled.
2575    
2576        A new data escape \S, is added to pcretest; it causes the amounts of store
2577        obtained and freed by both kinds of malloc/free at match time to be added
2578        to the output.
2579    
2580     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2581        what's available on my current Linux desktop machine.
2582    
2583     3. When matching a UTF-8 string, the test for a valid string at the start has
2584        been extended. If start_offset is not zero, PCRE now checks that it points
2585        to a byte that is the start of a UTF-8 character. If not, it returns
2586        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2587        this is necessary because there may be backward assertions in the pattern.
2588        When matching the same subject several times, it may save resources to use
2589        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2590    
2591     4. The code for checking the validity of UTF-8 strings has been tightened so
2592        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2593        containing "overlong sequences".
2594    
2595     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2596        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2597        should have been "&", but it just so happened that all the cases this let
2598        through by mistake were picked up later in the function.
2599    
2600     6. I had used a variable called "isblank" - this is a C99 function, causing
2601        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2602    
2603     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2604        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2605        so the version is shown; (c) stop "make test" echoing "./RunTest".
2606    
2607     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2608    
2609     9. The prototype for memmove() for systems that don't have it was using
2610        size_t, but the inclusion of the header that defines size_t was later. I've
2611        moved the #includes for the C headers earlier to avoid this.
2612    
2613    10. Added some adjustments to the code to make it easier to compiler on certain
2614        special systems:
2615    
2616          (a) Some "const" qualifiers were missing.
2617          (b) Added the macro EXPORT before all exported functions; by default this
2618              is defined to be empty.
2619          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2620              that it reads its output file name as an argument instead of writing
2621              to the standard output and assuming this can be redirected.
2622    
2623    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2624        class containing characters with values greater than 255, PCRE compilation
2625        went into a loop.
2626    
2627    12. A recursive reference to a subpattern that was within another subpattern
2628        that had a minimum quantifier of zero caused PCRE to crash. For example,
2629        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2630        recursion. If the recursively-called subpattern itself had a zero repeat,
2631        that was OK.
2632    
2633    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2634        buffer into which it was copied (for escape processing) was still set at
2635        1024, so long lines caused crashes.
2636    
2637    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2638        "internal error: code overflow...". This applied to any character class
2639        that was followed by a possessive quantifier.
2640    
2641    15. Modified the Makefile to add libpcre.la as a prerequisite for
2642        libpcreposix.la because I was told this is needed for a parallel build to
2643        work.
2644    
2645    16. If a pattern that contained .* following optional items at the start was
2646        studied, the wrong optimizing data was generated, leading to matching
2647        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2648        matching string must start with a or b or c. The correct conclusion for
2649        this pattern is that a match can start with any character.
2650    
2651    
2652    Version 4.4 13-Aug-03
2653    ---------------------
2654    
2655     1. In UTF-8 mode, a character class containing characters with values between
2656        127 and 255 was not handled correctly if the compiled pattern was studied.
2657        In fixing this, I have also improved the studying algorithm for such
2658        classes (slightly).
2659    
2660     2. Three internal functions had redundant arguments passed to them. Removal
2661        might give a very teeny performance improvement.
2662    
2663     3. Documentation bug: the value of the capture_top field in a callout is *one
2664        more than* the number of the hightest numbered captured substring.
2665    
2666     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2667        in incorrectly linking with a previously installed version. They now link
2668        explicitly with libpcre.la.
2669    
2670     5. configure.in no longer needs to recognize Cygwin specially.
2671    
2672     6. A problem in pcre.in for Windows platforms is fixed.
2673    
2674     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2675        pcretest, it used to include the size of the study block as part of its
2676        output. Unfortunately, the structure contains a field that has a different
2677        size on different hardware architectures. This meant that the tests that
2678        showed this size failed. As the block is currently always of a fixed size,
2679        this information isn't actually particularly useful in pcretest output, so
2680        I have just removed it.
2681    
2682     8. Three pre-processor statements accidentally did not start in column 1.
2683        Sadly, there are *still* compilers around that complain, even though
2684        standard C has not required this for well over a decade. Sigh.
2685    
2686     9. In pcretest, the code for checking callouts passed small integers in the
2687        callout_data field, which is a void * field. However, some picky compilers
2688        complained about the casts involved for this on 64-bit systems. Now
2689        pcretest passes the address of the small integer instead, which should get
2690        rid of the warnings.
2691    
2692    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2693        both compile and run time, and gives an error if an invalid UTF-8 sequence
2694        is found. There is a option for disabling this check in cases where the
2695        string is known to be correct and/or the maximum performance is wanted.
2696    
2697    11. In response to a bug report, I changed one line in Makefile.in from
2698    
2699            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2700        to
2701            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2702    
2703        to look similar to other lines, but I have no way of telling whether this
2704        is the right thing to do, as I do not use Windows. No doubt I'll get told
2705        if it's wrong...
2706    
2707    
2708    Version 4.3 21-May-03
2709    ---------------------
2710    
2711    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2712       Makefile.
2713    
2714    2. Some refactoring to improve the quality of the code:
2715    
2716       (i)   The utf8_table... variables are now declared "const".
2717    
2718       (ii)  The code for \cx, which used the "case flipping" table to upper case
2719             lower case letters, now just substracts 32. This is ASCII-specific,
2720             but the whole concept of \cx is ASCII-specific, so it seems
2721             reasonable.
2722    
2723       (iii) PCRE was using its character types table to recognize decimal and
2724             hexadecimal digits in the pattern. This is silly, because it handles
2725             only 0-9, a-f, and A-F, but the character types table is locale-
2726             specific, which means strange things might happen. A private
2727             table is now used for this - though it costs 256 bytes, a table is
2728             much faster than multiple explicit tests. Of course, the standard
2729             character types table is still used for matching digits in subject
2730             strings against \d.
2731    
2732       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2733             ending in _t are). So I've renamed it as ESC_tee.
2734    
2735    3. The first argument for regexec() in the POSIX wrapper should have been
2736       defined as "const".
2737    
2738    4. Changed pcretest to use malloc() for its buffers so that they can be
2739       Electric Fenced for debugging.
2740    
2741    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2742       to read one or more bytes before the start of the subject string. Often this
2743       had no effect on PCRE's behaviour, but in some circumstances it could
2744       provoke a segmentation fault.
2745    
2746    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2747       to try to read one or more bytes before the start of the subject string.
2748    
2749    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2750       UTF-8 support could misbehave in various ways if the subject string
2751       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2752       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2753       back over UTF-8 characters.)
2754    
2755    
2756    Version 4.2 14-Apr-03
2757    ---------------------
2758    
2759    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2760    
2761    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2762         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2763         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2764         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2765         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2766           and BUILD_EXEEXT
2767         Note: automatic setting of the BUILD variables is not yet working
2768         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2769           compile-time but not at link-time
2770         [LINK]: use for linking executables only
2771         make different versions for Windows and non-Windows
2772         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2773           libraries
2774         [LINK_FOR_BUILD]: new variable
2775         [OBJEXT]: use throughout
2776         [EXEEXT]: use throughout
2777         <winshared>: new target
2778         <wininstall>: new target
2779         <dftables.o>: use native compiler
2780         <dftables>: use native linker
2781         <install>: handle Windows platform correctly
2782         <clean>: ditto
2783         <check>: ditto
2784         copy DLL to top builddir before testing
2785    
2786       As part of these changes, -no-undefined was removed again. This was reported
2787       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2788       in any case.
2789    
2790    3. Some tidies to get rid of compiler warnings:
2791    
2792       . In the match_data structure, match_limit was an unsigned long int, whereas
2793         match_call_count was an int. I've made them both unsigned long ints.
2794    
2795       . In pcretest the fact that a const uschar * doesn't automatically cast to
2796         a void * provoked a warning.
2797    
2798       . Turning on some more compiler warnings threw up some "shadow" variables
2799         and a few more missing casts.
2800    
2801    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2802       option, a class that contained a single character with a value between 128
2803       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2804    
2805    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2806       option, a class that contained several characters, but with at least one
2807       whose value was between 128 and 255 caused PCRE to crash.
2808    
2809    
2810    Version 4.1 12-Mar-03
2811    ---------------------
2812    
2813    1. Compiling with gcc -pedantic found a couple of places where casts were
2814    needed, and a string in dftables.c that was longer than standard compilers are
2815    required to support.
2816    
2817    2. Compiling with Sun's compiler found a few more places where the code could
2818    be tidied up in order to avoid warnings.
2819    
2820    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2821    first of these names is deprecated in the latest Autoconf in favour of the name
2822    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2823    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2824    analogy I have changed it to CFLAGS_FOR_BUILD.
2825    
2826    4. Added -no-undefined to the linking command in the Makefile, because this is
2827    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2828    linking step for the pcreposix library.
2829    
2830    5. PCRE was failing to diagnose the case of two named groups with the same
2831    name.
2832    
2833    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2834    literal character that is needed in the subject for a match, and scans along to
2835    ensure that it is present before embarking on the full matching process. This
2836    saves time in cases of nested unlimited repeats that are never going to match.
2837    Problem: the scan can take a lot of time if the subject is very long (e.g.
2838    megabytes), thus penalizing straightforward matches. It is now done only if the
2839    amount of subject to be scanned is less than 1000 bytes.
2840    
2841    7. A lesser problem with the same optimization is that it was recording the
2842    first character of an anchored pattern as "needed", thus provoking a search
2843    right along the subject, even when the first match of the pattern was going to
2844    fail. The "needed" character is now not set for anchored patterns, unless it
2845    follows something in the pattern that is of non-fixed length. Thus, it still
2846    fulfils its original purpose of finding quick non-matches in cases of nested
2847    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2848    
2849    
2850    Version 4.0 17-Feb-03
2851    ---------------------
2852    
2853    1. If a comment in an extended regex that started immediately after a meta-item
2854    extended to the end of string, PCRE compiled incorrect data. This could lead to
2855    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2856    
2857    2. Moved to autoconf 2.53 and libtool 1.4.2.
2858    
2859    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2860    the special perltest8 script is no longer needed - all the tests can be run
2861    from a single perltest script.
2862    
2863    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2864    by \s. It has now been removed in PCRE. This means it isn't recognized as
2865    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2866    class [:space:] *does* include VT, thereby creating a mess.
2867    
2868    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2869    space and tab.
2870    
2871    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2872    its new features into the main test script, reducing the number of scripts.
2873    
2874    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2875    were backward compatible, and made the (?i) apply to the whole pattern, as if
2876    /i were given. Now it behaves more logically, and applies the option setting
2877    only to what follows. PCRE has been changed to follow suit. However, if it
2878    finds options settings right at the start of the pattern, it extracts them into
2879    the global options, as before. Thus, they show up in the info data.
2880    
2881    8. Added support for the \Q...\E escape sequence. Characters in between are
2882    treated as literals. This is slightly different from Perl in that $ and @ are
2883    also handled as literals inside the quotes. In Perl, they will cause variable
2884    interpolation. Note the following examples:
2885    
2886        Pattern            PCRE matches      Perl matches
2887    
2888        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2889        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2890        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2891    
2892    For compatibility with Perl, \Q...\E sequences are recognized inside character
2893    classes as well as outside them.
2894    
2895    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2896    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2897    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2898    signed/unsigned warnings.
2899    
2900    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2901    option for pcretest, so I've replaced it by a simple function that does just
2902    that job.
2903    
2904    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2905    "pcregrep -".
2906    
2907    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2908    Java package. This provides some syntactic sugar for simple cases of what my
2909    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2910    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2911    item, you can use this simplified notation. Note that only makes sense with
2912    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2913    greediness, whatever the setting of the PCRE_UNGREEDY option.
2914    
2915    13. A change of greediness default within a pattern was not taking effect at
2916    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2917    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2918    was abstracted outside.
2919    
2920    14. PCRE now supports the \G assertion. It is true when the current matching
2921    position is at the start point of the match. This differs from \A when the
2922    starting offset is non-zero. Used with the /g option of pcretest (or similar
2923    code), it works in the same way as it does for Perl's /g option. If all
2924    alternatives of a regex begin with \G, the expression is anchored to the start
2925    match position, and the "anchored" flag is set in the compiled expression.
2926    
2927    15. Some bugs concerning the handling of certain option changes within patterns
2928    have been fixed. These applied to options other than (?ims). For example,
2929    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2930    been the other way round. Some of this was related to change 7 above.
2931    
2932    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2933    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2934    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2935    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2936    
2937    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2938    mode. Unlike ".", it always matches newline, whatever the setting of
2939    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2940    assertions. Perl allows it, but it doesn't (in general) work because it can't
2941    calculate the length of the lookbehind. At least, that's the case for Perl
2942    5.8.0 - I've been told they are going to document that it doesn't work in
2943    future.
2944    
2945    18. Added an error diagnosis for escapes that PCRE does not support: these are
2946    \L, \l, \N, \P, \p, \U, \u, and \X.
2947    
2948    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2949    reading past the end of the pattern in cases such as /[abcd/.
2950    
2951    20. PCRE was getting more memory than necessary for patterns with classes that
2952    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2953    
2954    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2955    compiling PCRE for use with Virtual Pascal.
2956    
2957    22. Small fix to the Makefile to make it work properly if the build is done
2958    outside the source tree.
2959    
2960    23. Added a new extension: a condition to go with recursion. If a conditional
2961    subpattern starts with (?(R) the "true" branch is used if recursion has
2962    happened, whereas the "false" branch is used only at the top level.
2963    
2964    24. When there was a very long string of literal characters (over 255 bytes
2965    without UTF support, over 250 bytes with UTF support), the computation of how
2966    much memory was required could be incorrect, leading to segfaults or other
2967    strange effects.
2968    
2969    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2970    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2971    there was a subsequent back reference to those brackets. This meant that, for
2972    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2973    possible to check for precisely this case. All we can do is abandon the
2974    optimization if .* occurs inside capturing brackets when there are any back
2975    references whatsoever. (See below for a better fix that came later.)
2976    
2977    26. The handling of the optimization for finding the first character of a
2978    non-anchored pattern, and for finding a character that is required later in the
2979    match were failing in some cases. This didn't break the matching; it just
2980    failed to optimize when it could. The way this is done has been re-implemented.
2981    
2982    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2983    
2984    28. Added a new feature that provides some of the functionality that Perl
2985    provides with (?{...}). The facility is termed a "callout". The way it is done
2986    in PCRE is for the caller to provide an optional function, by setting
2987    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2988    global variable. By default it is unset, which disables all calling out. To get
2989    the function called, the regex must include (?C) at appropriate points. This
2990    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2991    This provides a means of identifying different callout points. When PCRE
2992    reaches such a point in the regex, if pcre_callout has been set, the external
2993    function is called. It is provided with data in a structure called
2994    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2995    matching continues; if it returns a non-zero value, the match at the current
2996    point fails. However, backtracking will occur if possible. [This was changed
2997    later and other features added - see item 49 below.]
2998    
2999    29. pcretest is upgraded to test the callout functionality. It provides a
3000    callout function that displays information. By default, it shows the start of
3001    the match and the current position in the text. There are some new data escapes
3002    to vary what happens:
3003    
3004        \C+         in addition, show current contents of captured substrings
3005        \C-         do not supply a callout function
3006        \C!n        return 1 when callout number n is reached
3007        \C!n!m      return 1 when callout number n is reached for the mth time
3008    
3009    30. If pcregrep was called with the -l option and just a single file name, it
3010    output "<stdin>" if a match was found, instead of the file name.
3011    
3012    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3013    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3014    pcre_exec(). This saves a malloc/free per call. The default value of
3015    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3016    when configuring.
3017    
3018    32. The default maximum size of a compiled pattern is 64K. There have been a
3019    few cases of people hitting this limit. The code now uses macros to handle the
3020    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3021    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3022    configuring. Tests 2 and 5 work only with 2-byte links because they output
3023    debugging information about compiled patterns.
3024    
3025    33. Internal code re-arrangements:
3026    
3027    (a) Moved the debugging function for printing out a compiled regex into
3028        its own source file (printint.c) and used #include to pull it into
3029        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3030        separate copies.
3031    
3032    (b) Defined the list of op-code names for debugging as a macro in
3033        internal.h so that it is next to the definition of the opcodes.
3034    
3035    (c) Defined a table of op-code lengths for simpler skipping along compiled
3036        code. This is again a macro in internal.h so that it is next to the
3037        definition of the opcodes.
3038    
3039    34. Added support for recursive calls to individual subpatterns, along the
3040    lines of Robin Houston's patch (but implemented somewhat differently).
3041    
3042    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3043    allow it to read and process whole directories in Win32. This code was
3044    contributed by Lionel Fourquaux; it has not been tested by me.
3045    
3046    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3047    used to name a group. Names consist of alphanumerics and underscores, and must
3048    be unique. Back references use the syntax (?P=name) and recursive calls use
3049    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3050    numbers. The function pcre_fullinfo() can be used after compilation to extract
3051    a name/number map. There are three relevant calls:
3052    
3053      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3054      PCRE_INFO_NAMECOUNT            yields the number of entries
3055      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3056    
3057    The map is a vector of fixed-size entries. The size of each entry depends on
3058    the length of the longest name used. The first two bytes of each entry are the
3059    group number, most significant byte first. There follows the corresponding
3060    name, zero terminated. The names are in alphabetical order.
3061    
3062    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3063    case instead of 255. Making it the same both with and without UTF-8 support
3064    means that the same test output works with both.
3065    
3066    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3067    calling malloc() with a zero argument.
3068    
3069    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3070    optimization. I've improved things by keeping a bitmap of backreferences with
3071    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3072    fact referenced, the optimization can be applied. It is unlikely that a
3073    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3074    the match to follow \n) will appear inside brackets with a number greater than
3075    31, but if it does, any back reference > 31 suppresses the optimization.
3076    
3077    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3078    of disabling numbered capturing parentheses. Any opening parenthesis that is
3079    not followed by ? behaves as if it were followed by ?: but named parentheses
3080    can still be used for capturing (and they will acquire numbers in the usual
3081    way).
3082    
3083    41. Redesigned the return codes from the match() function into yes/no/error so
3084    that errors can be passed back from deep inside the nested calls. A malloc
3085    failure while inside a recursive subpattern call now causes the
3086    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3087    
3088    42. It is now possible to set a limit on the number of times the match()
3089    function is called in a call to pcre_exec(). This facility makes it possible to
3090    limit the amount of recursion and backtracking, though not in a directly
3091    obvious way, because the match() function is used in a number of different
3092    circumstances. The count starts from zero for each position in the subject
3093    string (for non-anchored patterns). The default limit is, for compatibility, a
3094    large number, namely 10 000 000. You can change this in two ways:
3095    
3096    (a) When configuring PCRE before making, you can use --with-match-limit=n
3097        to set a default value for the compiled library.
3098    
3099    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3100        a different value is set. See 45 below.
3101    
3102    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3103    
3104    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3105    of things that can be changed at compile time. The first argument specifies
3106    what is wanted and the second points to where the information is to be placed.
3107    The current list of available information is:
3108    
3109      PCRE_CONFIG_UTF8
3110    
3111    The output is an integer that is set to one if UTF-8 support is available;
3112    otherwise it is set to zero.
3113    
3114      PCRE_CONFIG_NEWLINE
3115    
3116    The output is an integer that it set to the value of the code that is used for
3117    newline. It is either LF (10) or CR (13).
3118    
3119      PCRE_CONFIG_LINK_SIZE
3120    
3121    The output is an integer that contains the number of bytes used for internal
3122    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3123    
3124      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3125    
3126    The output is an integer that contains the threshold above which the POSIX
3127    interface uses malloc() for output vectors. See item 31 above.
3128    
3129      PCRE_CONFIG_MATCH_LIMIT
3130    
3131    The output is an unsigned integer that contains the default limit of the number
3132    of match() calls in a pcre_exec() execution. See 42 above.
3133    
3134    44. pcretest has been upgraded by the addition of the -C option. This causes it
3135    to extract all the available output from the new pcre_config() function, and to
3136    output it. The program then exits immediately.
3137    
3138    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3139    order to support additional features. One way would have been to define
3140    pcre_exec2() (for example) with extra arguments, but this would not have been
3141    extensible, and would also have required all calls to the original function to
3142    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3143    is used for passing in "extra" data from pcre_study().
3144    
3145    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3146    contains the following fields:
3147    
3148      flags         a bitmap indicating which of the following fields are set
3149      study_data    opaque data from pcre_study()
3150      match_limit   a way of specifying a limit on match() calls for a specific
3151                      call to pcre_exec()
3152      callout_data  data for callouts (see 49 below)
3153    
3154    The flag bits are also defined in pcre.h, and are
3155    
3156      PCRE_EXTRA_STUDY_DATA
3157      PCRE_EXTRA_MATCH_LIMIT
3158      PCRE_EXTRA_CALLOUT_DATA
3159    
3160    The pcre_study() function now returns one of these new pcre_extra blocks, with
3161    the actual study data pointed to by the study_data field, and the
3162    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3163    before. That is, this change is entirely upwards-compatible and requires no
3164    change to existing code.
3165    
3166    If you want to pass in additional data to pcre_exec(), you can either place it
3167    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3168    block.
3169    
3170    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3171    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3172    times with different match limits, until it finds the minimum value needed for
3173    pcre_exec() to complete. The value is then output. This can be instructive; for
3174    most simple matches the number is quite small, but for pathological cases it
3175    gets very large very quickly.
3176    
3177    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3178    returns the size of the data block pointed to by the study_data field in a
3179    pcre_extra block, that is, the value that was passed as the argument to
3180    pcre_malloc() when PCRE was getting memory in which to place the information
3181    created by pcre_study(). The fourth argument should point to a size_t variable.
3182    pcretest has been extended so that this information is shown after a successful
3183    pcre_study() call when information about the compiled regex is being displayed.
3184    
3185    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3186    because what follows is always an absolute path. (Later: it turns out that this
3187    is more than cosmetic for MinGW, because it doesn't like empty path
3188    components.)
3189    
3190    49. Some changes have been made to the callout feature (see 28 above):
3191    
3192    (i)  A callout function now has three choices for what it returns:
3193    
3194           0  =>  success, carry on matching
3195         > 0  =>  failure at this point, but backtrack if possible
3196         < 0  =>  serious error, return this value from pcre_exec()
3197    
3198         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3199         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3200         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3201         use by callout functions. It will never be used by PCRE itself.
3202    
3203    (ii) The pcre_extra structure (see 45 above) has a void * field called
3204         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3205         pcre_callout_block structure has a field of the same name. The contents of
3206         the field passed in the pcre_extra structure are passed to the callout
3207         function in the corresponding field in the callout block. This makes it
3208         easier to use the same callout-containing regex from multiple threads. For
3209         testing, the pcretest program has a new data escape
3210    
3211           \C*n        pass the number n (may be negative) as callout_data
3212    
3213         If the callout function in pcretest receives a non-zero value as
3214         callout_data, it returns that value.
3215    
3216    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3217    there were some redundant $(CFLAGS) in commands that are now specified as
3218    $(LINK), which already includes $(CFLAGS).
3219    
3220    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3221    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3222    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3223    one-byte characters throughout. Note that case-insensitive matching applies
3224    only to characters whose values are less than 256. PCRE doesn't support the
3225    notion of cases for higher-valued characters.
3226    
3227    (i)   A character class whose characters are all within 0-255 is handled as
3228          a bit map, and the map is inverted for negative classes. Previously, a
3229          character > 255 always failed to match such a class; however it should
3230          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3231    
3232    (ii)  A negated character class with a single character < 255 is coded as
3233          "not this character" (OP_NOT). This wasn't working properly when the test
3234          character was multibyte, either singly or repeated.
3235    
3236    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3237          mode, for example: \x{100}{2,3}.
3238    
3239    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3240          singly or repeated) now correctly test multibyte characters. However,
3241          PCRE doesn't recognize any characters with values greater than 255 as
3242          digits, spaces, or word characters. Such characters always match \D, \S,
3243          and \W, and never match \d, \s, or \w.
3244    
3245    (v)   Classes may now contain characters and character ranges with values
3246          greater than 255. For example: [ab\x{100}-\x{400}].
3247    
3248    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3249          PCRE in UTF-8 mode.
3250    
3251    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3252    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3253    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3254    value.)
3255    
3256    53. The single man page has become too large. I have therefore split it up into
3257    a number of separate man pages. These also give rise to individual HTML pages;
3258    these are now put in a separate directory, and there is an index.html page that
3259    lists them all. Some hyperlinking between the pages has been installed.
3260    
3261    54. Added convenience functions for handling named capturing parentheses.
3262    
3263    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3264    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3265    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3266    are faulted.
3267    
3268    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3269    calling configure. These values are used when compiling the dftables.c program
3270    which is run to generate the source of the default character tables. They
3271    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3272    you will need to set these values.
3273    
3274    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3275    
3276    
3277    Version 3.9 02-Jan-02
3278    ---------------------
3279    
3280    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3281    
3282    2. If --disable-static was given, the building process failed when trying to
3283    build pcretest and pcregrep. (For some reason it was using libtool to compile
3284    them, which is not right, as they aren't part of the library.)
3285    
3286    
3287    Version 3.8 18-Dec-01
3288    ---------------------
3289    
3290    1. The experimental UTF-8 code was completely screwed up. It was packing the
3291    bytes in the wrong order. How dumb can you get?
3292    
3293    
3294    Version 3.7 29-Oct-01
3295    ---------------------
3296    
3297    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3298    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3299    this didn't happen under Solaris 8, where I normally test things.
3300    
3301    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3302    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3303    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3304    this fix an hour or so after the initial 3.7 release.)
3305    
3306    
3307    Version 3.6 23-Oct-01
3308    ---------------------
3309    
3310    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3311    offsets passed as NULL with zero offset count.
3312    
3313    2. The config.guess and config.sub files had not been updated when I moved to
3314    the latest autoconf.
3315    
3316    
3317  Version 3.5 15-Aug-01  Version 3.5 15-Aug-01
3318  ---------------------  ---------------------
3319    

Legend:
Removed from v.53  
changed lines
  Added in v.564

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12