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

Diff of /code/trunk/ChangeLog

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

revision 43 by nigel, Sat Feb 24 21:39:21 2007 UTC revision 553 by ph10, Fri Oct 22 15:57:50 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    
44    Version 8.10 25-Jun-2010
45    ------------------------
46    
47    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
48        THEN.
49    
50    2.  (*ACCEPT) was not working when inside an atomic group.
51    
52    3.  Inside a character class, \B is treated as a literal by default, but
53        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
54        causes the error). The code is unchanged, but I tidied the documentation.
55    
56    4.  Inside a character class, PCRE always treated \R and \X as literals,
57        whereas Perl faults them if its -w option is set. I have changed PCRE so
58        that it faults them when PCRE_EXTRA is set.
59    
60    5.  Added support for \N, which always matches any character other than
61        newline. (It is the same as "." when PCRE_DOTALL is not set.)
62    
63    6.  When compiling pcregrep with newer versions of gcc which may have
64        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
65        declared with attribute warn_unused_result" were given. Just casting the
66        result to (void) does not stop the warnings; a more elaborate fudge is
67        needed. I've used a macro to implement this.
68    
69    7.  Minor change to pcretest.c to avoid a compiler warning.
70    
71    8.  Added four artifical Unicode properties to help with an option to make
72        \s etc use properties (see next item). The new properties are: Xan
73        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
74    
75    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
76        use Unicode properties. (*UCP) at the start of a pattern can be used to set
77        this option. Modified pcretest to add /W to test this facility. Added
78        REG_UCP to make it available via the POSIX interface.
79    
80    10. Added --line-buffered to pcregrep.
81    
82    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
83        studied, and the match started with a letter with a code point greater than
84        127 whose first byte was different to the first byte of the other case of
85        the letter, the other case of this starting letter was not recognized
86        (#976).
87    
88    12. If a pattern that was studied started with a repeated Unicode property
89        test, for example, \p{Nd}+, there was the theoretical possibility of
90        setting up an incorrect bitmap of starting bytes, but fortunately it could
91        not have actually happened in practice until change 8 above was made (it
92        added property types that matched character-matching opcodes).
93    
94    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
95        possible starting bytes for non-anchored patterns.
96    
97    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
98        \R, and also a number of cases that involve Unicode properties, both
99        explicit and implicit when PCRE_UCP is set.
100    
101    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
102        input, it could crash or give wrong results if characters with values
103        greater than 0xc0 were present in the subject string. (Detail: it assumed
104        UTF-8 input when processing these items.)
105    
106    16. Added a lot of (int) casts to avoid compiler warnings in systems where
107        size_t is 64-bit (#991).
108    
109    17. Added a check for running out of memory when PCRE is compiled with
110        --disable-stack-for-recursion (#990).
111    
112    18. If the last data line in a file for pcretest does not have a newline on
113        the end, a newline was missing in the output.
114    
115    19. The default pcre_chartables.c file recognizes only ASCII characters (values
116        less than 128) in its various bitmaps. However, there is a facility for
117        generating tables according to the current locale when PCRE is compiled. It
118        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
119        characters, are recognized by isspace() and therefore were getting set in
120        these tables, and indeed these tables seem to approximate to ISO 8859. This
121        caused a problem in UTF-8 mode when pcre_study() was used to create a list
122        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
123        which of course cannot start UTF-8 characters. I have changed the code so
124        that only real ASCII characters (less than 128) and the correct starting
125        bytes for UTF-8 encodings are set for characters greater than 127 when in
126        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
127        altogether.)
128    
129    20. Added the /T option to pcretest so as to be able to run tests with non-
130        standard character tables, thus making it possible to include the tests
131        used for 19 above in the standard set of tests.
132    
133    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
134        reference to a subpattern the other side of a comment that contains an
135        opening parenthesis caused either an internal compiling error, or a
136        reference to the wrong subpattern.
137    
138    
139    Version 8.02 19-Mar-2010
140    ------------------------
141    
142    1.  The Unicode data tables have been updated to Unicode 5.2.0.
143    
144    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
145        configured.
146    
147    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
148        original author of that file, following a query about its status.
149    
150    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
151        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
152    
153    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
154        quantifier applied to a forward-referencing subroutine call, could compile
155        incorrect code or give the error "internal error: previously-checked
156        referenced subpattern not found".
157    
158    6.  Both MS Visual Studio and Symbian OS have problems with initializing
159        variables to point to external functions. For these systems, therefore,
160        pcre_malloc etc. are now initialized to local functions that call the
161        relevant global functions.
162    
163    7.  There were two entries missing in the vectors called coptable and poptable
164        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
165        I've fixed the data, and added a kludgy way of testing at compile time that
166        the lengths are correct (equal to the number of opcodes).
167    
168    8.  Following on from 7, I added a similar kludge to check the length of the
169        eint vector in pcreposix.c.
170    
171    9.  Error texts for pcre_compile() are held as one long string to avoid too
172        much relocation at load time. To find a text, the string is searched,
173        counting zeros. There was no check for running off the end of the string,
174        which could happen if a new error number was added without updating the
175        string.
176    
177    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
178    
179    11. \K was not working if it appeared in an atomic group or in a group that
180        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
181        \K is "not well defined" if used in an assertion. PCRE now accepts it if
182        the assertion is positive, but not if it is negative.
183    
184    12. Change 11 fortuitously reduced the size of the stack frame used in the
185        "match()" function of pcre_exec.c by one pointer. Forthcoming
186        implementation of support for (*MARK) will need an extra pointer on the
187        stack; I have reserved it now, so that the stack frame size does not
188        decrease.
189    
190    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
191        item in branch that calls a recursion is a subroutine call - as in the
192        second branch in the above example - was incorrectly given the compile-
193        time error "recursive call could loop indefinitely" because pcre_compile()
194        was not correctly checking the subroutine for matching a non-empty string.
195    
196    14. The checks for overrunning compiling workspace could trigger after an
197        overrun had occurred. This is a "should never occur" error, but it can be
198        triggered by pathological patterns such as hundreds of nested parentheses.
199        The checks now trigger 100 bytes before the end of the workspace.
200    
201    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
202    
203    
204    Version 8.01 19-Jan-2010
205    ------------------------
206    
207    1.  If a pattern contained a conditional subpattern with only one branch (in
208        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
209        computed the wrong minimum data length (which is of course zero for such
210        subpatterns). This could cause incorrect "no match" results.
211    
212    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
213        the pattern is reset in the first branch, pcre_compile() failed with
214        "internal error: code overflow at offset...". This happened only when
215        the reset was to the original external option setting. (An optimization
216        abstracts leading options settings into an external setting, which was the
217        cause of this.)
218    
219    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
220        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
221        assertion pattern did not match (meaning that the assertion was true), it
222        was incorrectly treated as false if the SKIP had been reached during the
223        matching. This also applied to assertions used as conditions.
224    
225    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
226        assertion subpattern, including such a pattern used as a condition,
227        unpredictable results occurred, instead of the error return
228        PCRE_ERROR_DFA_UITEM.
229    
230    5.  The C++ GlobalReplace function was not working like Perl for the special
231        situation when an empty string is matched. It now does the fancy magic
232        stuff that is necessary.
233    
234    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
235        removed. (These were left over from very, very early versions of PCRE.)
236    
237    7.  Some cosmetic changes to the code to make life easier when compiling it
238        as part of something else:
239    
240        (a) Change DEBUG to PCRE_DEBUG.
241    
242        (b) In pcre_compile(), rename the member of the "branch_chain" structure
243            called "current" as "current_branch", to prevent a collision with the
244            Linux macro when compiled as a kernel module.
245    
246        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
247            prevent a collision with the Linux macro when compiled as a kernel
248            module.
249    
250    8.  In pcre_compile() there are some checks for integer overflows that used to
251        cast potentially large values to (double). This has been changed to that
252        when building, a check for int64_t is made, and if it is found, it is used
253        instead, thus avoiding the use of floating point arithmetic. (There is no
254        other use of FP in PCRE.) If int64_t is not found, the fallback is to
255        double.
256    
257    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
258        2005 (difference between two addresses compared to an unsigned value).
259    
260    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
261        custom one, because of the following reported problem in Windows:
262    
263          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
264              under Win32.
265          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
266              therefore missing the function definition.
267          - The compiler thus generates a "C" signature for the test function.
268          - The linker fails to find the "C" function.
269          - PCRE fails to configure if asked to do so against libbz2.
270    
271    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
272        messages were output:
273    
274          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
275          rerunning libtoolize, to keep the correct libtool macros in-tree.
276          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
277    
278        I have done both of these things.
279    
280    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
281        most of the time, it *can* run out if it is given a pattern that contains a
282        runaway infinite recursion. I updated the discussion in the pcrestack man
283        page.
284    
285    13. Now that we have gone to the x.xx style of version numbers, the minor
286        version may start with zero. Using 08 or 09 is a bad idea because users
287        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
288        interpreted as invalid octal numbers. I've updated the previous comment in
289        configure.ac, and also added a check that gives an error if 08 or 09 are
290        used.
291    
292    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
293        causing partial matching to fail when the end of the subject matched \W
294        in a UTF-8 pattern where \W was quantified with a minimum of 3.
295    
296    15. There were some discrepancies between the declarations in pcre_internal.h
297        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
298        their definitions. The declarations used "const uschar *" and the
299        definitions used USPTR. Even though USPTR is normally defined as "const
300        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
301        reported that: "This difference in casting confuses some C++ compilers, for
302        example, SunCC recognizes above declarations as different functions and
303        generates broken code for hbpcre." I have changed the declarations to use
304        USPTR.
305    
306    16. GNU libtool is named differently on some systems. The autogen.sh script now
307        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
308        (FreeBSD).
309    
310    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
311        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
312        comment: "Figure out how to create a longlong from a string: strtoll and
313        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
314        instance, but it only takes 2 args instead of 3!"
315    
316    18. A subtle bug concerned with back references has been fixed by a change of
317        specification, with a corresponding code fix. A pattern such as
318        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
319        refers, was giving matches when it shouldn't. For example, xa=xaaa would
320        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
321        same bug. Such groups have to be quantified to be useful, or contained
322        inside another quantified group. (If there's no repetition, the reference
323        can never match.) The problem arises because, having left the group and
324        moved on to the rest of the pattern, a later failure that backtracks into
325        the group uses the captured value from the final iteration of the group
326        rather than the correct earlier one. I have fixed this in PCRE by forcing
327        any group that contains a reference to itself to be an atomic group; that
328        is, there cannot be any backtracking into it once it has completed. This is
329        similar to recursive and subroutine calls.
330    
331    
332    Version 8.00 19-Oct-09
333    ----------------------
334    
335    1.  The table for translating pcre_compile() error codes into POSIX error codes
336        was out-of-date, and there was no check on the pcre_compile() error code
337        being within the table. This could lead to an OK return being given in
338        error.
339    
340    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
341        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
342        in a Windows environment.
343    
344    3.  The pcregrep --count option prints the count for each file even when it is
345        zero, as does GNU grep. However, pcregrep was also printing all files when
346        --files-with-matches was added. Now, when both options are given, it prints
347        counts only for those files that have at least one match. (GNU grep just
348        prints the file name in this circumstance, but including the count seems
349        more useful - otherwise, why use --count?) Also ensured that the
350        combination -clh just lists non-zero counts, with no names.
351    
352    4.  The long form of the pcregrep -F option was incorrectly implemented as
353        --fixed_strings instead of --fixed-strings. This is an incompatible change,
354        but it seems right to fix it, and I didn't think it was worth preserving
355        the old behaviour.
356    
357    5.  The command line items --regex=pattern and --regexp=pattern were not
358        recognized by pcregrep, which required --regex pattern or --regexp pattern
359        (with a space rather than an '='). The man page documented the '=' forms,
360        which are compatible with GNU grep; these now work.
361    
362    6.  No libpcreposix.pc file was created for pkg-config; there was just
363        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
364    
365    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
366        when UCP support is not needed, by modifying the Python script that
367        generates it from Unicode data files. This should not matter if the module
368        is correctly used as a library, but I received one complaint about 50K of
369        unwanted data. My guess is that the person linked everything into his
370        program rather than using a library. Anyway, it does no harm.
371    
372    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
373        was a minimum greater than 1 for a wide character in a possessive
374        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
375        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
376        character. Chaos in the form of incorrect output or a compiling loop could
377        result.
378    
379    9.  The restrictions on what a pattern can contain when partial matching is
380        requested for pcre_exec() have been removed. All patterns can now be
381        partially matched by this function. In addition, if there are at least two
382        slots in the offset vector, the offset of the earliest inspected character
383        for the match and the offset of the end of the subject are set in them when
384        PCRE_ERROR_PARTIAL is returned.
385    
386    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
387        synonymous with PCRE_PARTIAL, for backwards compatibility, and
388        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
389        and may be more useful for multi-segment matching.
390    
391    11. Partial matching with pcre_exec() is now more intuitive. A partial match
392        used to be given if ever the end of the subject was reached; now it is
393        given only if matching could not proceed because another character was
394        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
395        string "Z", which now yields "no match" instead of "partial match". In the
396        case of pcre_dfa_exec(), "no match" is given if every matching path for the
397        final character ended with (*FAIL).
398    
399    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
400        if the pattern had a "must contain" character that was already found in the
401        earlier partial match, unless partial matching was again requested. For
402        example, with the pattern /dog.(body)?/, the "must contain" character is
403        "g". If the first part-match was for the string "dog", restarting with
404        "sbody" failed. This bug has been fixed.
405    
406    13. The string returned by pcre_dfa_exec() after a partial match has been
407        changed so that it starts at the first inspected character rather than the
408        first character of the match. This makes a difference only if the pattern
409        starts with a lookbehind assertion or \b or \B (\K is not supported by
410        pcre_dfa_exec()). It's an incompatible change, but it makes the two
411        matching functions compatible, and I think it's the right thing to do.
412    
413    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
414        so that the demonstration program is easily available in environments where
415        PCRE has not been installed from source.
416    
417    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
418        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
419        library.
420    
421    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
422        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
423        is not the first non-POSIX option to be added. Clearly some people find
424        these options useful.
425    
426    17. If a caller to the POSIX matching function regexec() passes a non-zero
427        value for nmatch with a NULL value for pmatch, the value of
428        nmatch is forced to zero.
429    
430    18. RunGrepTest did not have a test for the availability of the -u option of
431        the diff command, as RunTest does. It now checks in the same way as
432        RunTest, and also checks for the -b option.
433    
434    19. If an odd number of negated classes containing just a single character
435        interposed, within parentheses, between a forward reference to a named
436        subpattern and the definition of the subpattern, compilation crashed with
437        an internal error, complaining that it could not find the referenced
438        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
439        [The bug was that it was starting one character too far in when skipping
440        over the character class, thus treating the ] as data rather than
441        terminating the class. This meant it could skip too much.]
442    
443    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
444        /g option in pcretest when the pattern contains \K, which makes it possible
445        to have an empty string match not at the start, even when the pattern is
446        anchored. Updated pcretest and pcredemo to use this option.
447    
448    21. If the maximum number of capturing subpatterns in a recursion was greater
449        than the maximum at the outer level, the higher number was returned, but
450        with unset values at the outer level. The correct (outer level) value is
451        now given.
452    
453    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
454        PCRE did not set those parentheses (unlike Perl). I have now found a way to
455        make it do so. The string so far is captured, making this feature
456        compatible with Perl.
457    
458    23. The tests have been re-organized, adding tests 11 and 12, to make it
459        possible to check the Perl 5.10 features against Perl 5.10.
460    
461    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
462        pattern matches a fixed length string. PCRE did not allow this; now it
463        does. Neither allows recursion.
464    
465    25. I finally figured out how to implement a request to provide the minimum
466        length of subject string that was needed in order to match a given pattern.
467        (It was back references and recursion that I had previously got hung up
468        on.) This code has now been added to pcre_study(); it finds a lower bound
469        to the length of subject needed. It is not necessarily the greatest lower
470        bound, but using it to avoid searching strings that are too short does give
471        some useful speed-ups. The value is available to calling programs via
472        pcre_fullinfo().
473    
474    26. While implementing 25, I discovered to my embarrassment that pcretest had
475        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
476        study optimizations had never been tested with that matching function.
477        Oops. What is worse, even when it was passed study data, there was a bug in
478        pcre_dfa_exec() that meant it never actually used it. Double oops. There
479        were also very few tests of studied patterns with pcre_dfa_exec().
480    
481    27. If (?| is used to create subpatterns with duplicate numbers, they are now
482        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
483        on the other side of the coin, they are no longer allowed to have different
484        names, because these cannot be distinguished in PCRE, and this has caused
485        confusion. (This is a difference from Perl.)
486    
487    28. When duplicate subpattern names are present (necessarily with different
488        numbers, as required by 27 above), and a test is made by name in a
489        conditional pattern, either for a subpattern having been matched, or for
490        recursion in such a pattern, all the associated numbered subpatterns are
491        tested, and the overall condition is true if the condition is true for any
492        one of them. This is the way Perl works, and is also more like the way
493        testing by number works.
494    
495    
496    Version 7.9 11-Apr-09
497    ---------------------
498    
499    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
500        (pcretest), all targets were linked against these libraries. This included
501        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
502        libraries. This caused unwanted dependencies to be created. This problem
503        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
504        pcretest is linked with readline.
505    
506    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
507        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
508        moved outside it again, because FALSE and TRUE are already defined in AIX,
509        but BOOL is not.
510    
511    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
512        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
513    
514    4.  The pcregrep documentation said spaces were inserted as well as colons (or
515        hyphens) following file names and line numbers when outputting matching
516        lines. This is not true; no spaces are inserted. I have also clarified the
517        wording for the --colour (or --color) option.
518    
519    5.  In pcregrep, when --colour was used with -o, the list of matching strings
520        was not coloured; this is different to GNU grep, so I have changed it to be
521        the same.
522    
523    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
524        each matching line was coloured. Now it goes on to look for further matches
525        of any of the test patterns, which is the same behaviour as GNU grep.
526    
527    7.  A pattern that could match an empty string could cause pcregrep to loop; it
528        doesn't make sense to accept an empty string match in pcregrep, so I have
529        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
530        seems to be how GNU grep behaves.
531    
532    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
533        start or after a newline", because the conditional assertion was not being
534        correctly handled. The rule now is that both the assertion and what follows
535        in the first alternative must satisfy the test.
536    
537    9.  If auto-callout was enabled in a pattern with a conditional group whose
538        condition was an assertion, PCRE could crash during matching, both with
539        pcre_exec() and pcre_dfa_exec().
540    
541    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
542        used for matching.
543    
544    11. Unicode property support in character classes was not working for
545        characters (bytes) greater than 127 when not in UTF-8 mode.
546    
547    12. Added the -M command line option to pcretest.
548    
549    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
550    
551    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
552    
553    16. Added comments and documentation about mis-use of no_arg in the C++
554        wrapper.
555    
556    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
557        from Martin Jerabek that uses macro names for all relevant character and
558        string constants.
559    
560    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
561        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
562        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
563        these, but not everybody uses configure.
564    
565    19. A conditional group that had only one branch was not being correctly
566        recognized as an item that could match an empty string. This meant that an
567        enclosing group might also not be so recognized, causing infinite looping
568        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
569        with the subject "ab", where knowledge that the repeated group can match
570        nothing is needed in order to break the loop.
571    
572    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
573        exec(), but without supplying a callout function, matching went wrong.
574    
575    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
576        leak if the size of the offset vector was greater than 30. When the vector
577        is smaller, the saved offsets during recursion go onto a local stack
578        vector, but for larger vectors malloc() is used. It was failing to free
579        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
580        error, in fact).
581    
582    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
583        heapframe that is used only when UTF-8 support is enabled. This caused no
584        problem, but was untidy.
585    
586    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
587        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
588        included within another project.
589    
590    24. Steven Van Ingelgem's patches to add more options to the CMake support,
591        slightly modified by me:
592    
593          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
594              not building pcregrep.
595    
596          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
597              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
598    
599    25. Forward references, both numeric and by name, in patterns that made use of
600        duplicate group numbers, could behave incorrectly or give incorrect errors,
601        because when scanning forward to find the reference group, PCRE was not
602        taking into account the duplicate group numbers. A pattern such as
603        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
604    
605    26. Changed a few more instances of "const unsigned char *" to USPTR, making
606        the feature of a custom pointer more persuasive (as requested by a user).
607    
608    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
609        pcretest.c, inside #ifndefs, because it seems they are sometimes already
610        pre-defined.
611    
612    28. Added support for (*UTF8) at the start of a pattern.
613    
614    29. Arrange for flags added by the "release type" setting in CMake to be shown
615        in the configuration summary.
616    
617    
618    Version 7.8 05-Sep-08
619    ---------------------
620    
621    1.  Replaced UCP searching code with optimized version as implemented for Ad
622        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
623        stage table and inline lookup instead of a function, giving speed ups of 2
624        to 5 times on some simple patterns that I tested. Permission was given to
625        distribute the MultiStage2.py script that generates the tables (it's not in
626        the tarball, but is in the Subversion repository).
627    
628    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
629        scripts.
630    
631    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
632        a group with a zero qualifier. The result of the study could be incorrect,
633        or the function might crash, depending on the pattern.
634    
635    4.  Caseless matching was not working for non-ASCII characters in back
636        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
637        It now works when Unicode Property Support is available.
638    
639    5.  In pcretest, an escape such as \x{de} in the data was always generating
640        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
641        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
642        truncation.
643    
644    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
645    
646    7.  Added two (int) casts to pcregrep when printing the difference of two
647        pointers, in case they are 64-bit values.
648    
649    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
650        test 2 if it fails.
651    
652    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
653        and a #define of that name to empty if it is not externally set. This is to
654        allow users of MSVC to set it if necessary.
655    
656    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
657        the convenience functions in the pcre_get.c source file.
658    
659    11. An option change at the start of a pattern that had top-level alternatives
660        could cause overwriting and/or a crash. This command provoked a crash in
661        some environments:
662    
663          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
664    
665        This potential security problem was recorded as CVE-2008-2371.
666    
667    12. For a pattern where the match had to start at the beginning or immediately
668        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
669        pcre_dfa_exec() could read past the end of the passed subject if there was
670        no match. To help with detecting such bugs (e.g. with valgrind), I modified
671        pcretest so that it places the subject at the end of its malloc-ed buffer.
672    
673    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
674        exec() might read past the end of the data buffer in UTF-8 mode.
675    
676    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
677        the data contained the byte 0x85 as part of a UTF-8 character within its
678        first line. This applied both to normal and DFA matching.
679    
680    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
681        /^[^d]*?$/8 failed to match "abc".
682    
683    16. Added a missing copyright notice to pcrecpp_internal.h.
684    
685    17. Make it more clear in the documentation that values returned from
686        pcre_exec() in ovector are byte offsets, not character counts.
687    
688    18. Tidied a few places to stop certain compilers from issuing warnings.
689    
690    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
691        supplied by Stefan Weber. I made a further small update for 7.8 because
692        there is a change of source arrangements: the pcre_searchfuncs.c module is
693        replaced by pcre_ucd.c.
694    
695    
696    Version 7.7 07-May-08
697    ---------------------
698    
699    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
700        a string to a long long, pretend we don't even have a long long." This is
701        done by checking for the strtoq, strtoll, and _strtoi64 functions.
702    
703    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
704        pre-7.6 versions, which defined a global no_arg variable instead of putting
705        it in the RE class. (See also #8 below.)
706    
707    3.  Remove a line of dead code, identified by coverity and reported by Nuno
708        Lopes.
709    
710    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
711    
712        (1) The include/exclude patterns were being applied to the whole pathnames
713            of files, instead of just to the final components.
714    
715        (2) If there was more than one level of directory, the subdirectories were
716            skipped unless they satisfied the include/exclude conditions. This is
717            inconsistent with GNU grep (and could even be seen as contrary to the
718            pcregrep specification - which I improved to make it absolutely clear).
719            The action now is always to scan all levels of directory, and just
720            apply the include/exclude patterns to regular files.
721    
722    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
723        --exclude_dir in the tests to avoid scanning .svn directories.
724    
725    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
726        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
727        doesn't support NULs in patterns.
728    
729    7.  Added some missing "const"s to declarations of static tables in
730        pcre_compile.c and pcre_dfa_exec.c.
731    
732    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
733        caused by fix #2  above. (Subsequently also a second patch to fix the
734        first patch. And a third patch - this was a messy problem.)
735    
736    9.  Applied Craig's patch to remove the use of push_back().
737    
738    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
739        matching function regexec().
740    
741    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
742        which, however, unlike Perl's \g{...}, are subroutine calls, not back
743        references. PCRE supports relative numbers with this syntax (I don't think
744        Oniguruma does).
745    
746    12. Previously, a group with a zero repeat such as (...){0} was completely
747        omitted from the compiled regex. However, this means that if the group
748        was called as a subroutine from elsewhere in the pattern, things went wrong
749        (an internal error was given). Such groups are now left in the compiled
750        pattern, with a new opcode that causes them to be skipped at execution
751        time.
752    
753    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
754        to the way PCRE behaves:
755    
756        (a) A lone ] character is dis-allowed (Perl treats it as data).
757    
758        (b) A back reference to an unmatched subpattern matches an empty string
759            (Perl fails the current match path).
760    
761        (c) A data ] in a character class must be notated as \] because if the
762            first data character in a class is ], it defines an empty class. (In
763            Perl it is not possible to have an empty class.) The empty class []
764            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
765            The negative empty class [^] matches any one character, independently
766            of the DOTALL setting.
767    
768    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
769        non-existent subpattern following a character class starting with ']' and
770        containing () gave an internal compiling error instead of "reference to
771        non-existent subpattern". Fortunately, when the pattern did exist, the
772        compiled code was correct. (When scanning forwards to check for the
773        existencd of the subpattern, it was treating the data ']' as terminating
774        the class, so got the count wrong. When actually compiling, the reference
775        was subsequently set up correctly.)
776    
777    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
778        it was being rejected as not supported by pcre_dfa_exec(), even though
779        other assertions are supported. I have made pcre_dfa_exec() support
780        (*FAIL).
781    
782    16. The implementation of 13c above involved the invention of a new opcode,
783        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
784        cannot be changed at match time, I realized I could make a small
785        improvement to matching performance by compiling OP_ALLANY instead of
786        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
787        on the OP_ANY path.
788    
789    17. Compiling pcretest on Windows with readline support failed without the
790        following two fixes: (1) Make the unistd.h include conditional on
791        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
792    
793    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
794        ncurses library to be included for pcretest when ReadLine support is
795        requested, but also to allow for it to be overridden. This patch came from
796        Daniel Bergström.
797    
798    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
799        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
800        any errors with the current Unicode tables. Thanks to Peter Kankowski for
801        spotting this.
802    
803    
804    Version 7.6 28-Jan-08
805    ---------------------
806    
807    1.  A character class containing a very large number of characters with
808        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
809        overflow.
810    
811    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
812        HAVE_LONG_LONG is not defined.
813    
814    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
815        bring them up to date and include new features. This patch includes:
816    
817        - Fixed PH's badly added libz and libbz2 support.
818        - Fixed a problem with static linking.
819        - Added pcredemo. [But later removed - see 7 below.]
820        - Fixed dftables problem and added an option.
821        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
822            HAVE_LONG_LONG.
823        - Added readline support for pcretest.
824        - Added an listing of the option settings after cmake has run.
825    
826    4.  A user submitted a patch to Makefile that makes it easy to create
827        "pcre.dll" under mingw when using Configure/Make. I added stuff to
828        Makefile.am that cause it to include this special target, without
829        affecting anything else. Note that the same mingw target plus all
830        the other distribution libraries and programs are now supported
831        when configuring with CMake (see 6 below) instead of with
832        Configure/Make.
833    
834    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
835        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
836        exported in the Windows port". It has not yet been confirmed that the patch
837        solves the problem, but it does no harm.
838    
839    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
840        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
841        with CMake, and also correct the comment about stack recursion.
842    
843    7.  Remove the automatic building of pcredemo from the ./configure system and
844        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
845        of a program that users should build themselves after PCRE is installed, so
846        building it automatically is not really right. What is more, it gave
847        trouble in some build environments.
848    
849    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
850    
851    
852    Version 7.5 10-Jan-08
853    ---------------------
854    
855    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
856        values in parens when parsing an RE using the C++ wrapper."
857    
858    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
859        Characters greater than 255 were excluded from the class instead of being
860        included.
861    
862    3.  The same bug as (2) above applied to negated POSIX classes such as
863        [:^space:].
864    
865    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
866        defined or documented. It seems to have been a typo for PCRE_STATIC, so
867        I have changed it.
868    
869    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
870        first named subpattern) and a construct such as (?&a) would reference the
871        first named subpattern whose name started with "a" (in other words, the
872        length check was missing). Both these problems are fixed. "Subpattern name
873        expected" is now given for (?&) (a zero-length name), and this patch also
874        makes it give the same error for \k'' (previously it complained that that
875        was a reference to a non-existent subpattern).
876    
877    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
878        this is right because (?- can be followed by option settings as well as by
879        digits. I have, however, made the messages clearer.
880    
881    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
882        than the number used in the conditional) now cause a compile-time error.
883        This is actually not compatible with Perl, which accepts such patterns, but
884        treats the conditional as always being FALSE (as PCRE used to), but it
885        seems to me that giving a diagnostic is better.
886    
887    8.  Change "alphameric" to the more common word "alphanumeric" in comments
888        and messages.
889    
890    9.  Fix two occurrences of "backslash" in comments that should have been
891        "backspace".
892    
893    10. Remove two redundant lines of code that can never be obeyed (their function
894        was moved elsewhere).
895    
896    11. The program that makes PCRE's Unicode character property table had a bug
897        which caused it to generate incorrect table entries for sequences of
898        characters that have the same character type, but are in different scripts.
899        It amalgamated them into a single range, with the script of the first of
900        them. In other words, some characters were in the wrong script. There were
901        thirteen such cases, affecting characters in the following ranges:
902    
903          U+002b0 - U+002c1
904          U+0060c - U+0060d
905          U+0061e - U+00612
906          U+0064b - U+0065e
907          U+0074d - U+0076d
908          U+01800 - U+01805
909          U+01d00 - U+01d77
910          U+01d9b - U+01dbf
911          U+0200b - U+0200f
912          U+030fc - U+030fe
913          U+03260 - U+0327f
914          U+0fb46 - U+0fbb1
915          U+10450 - U+1049d
916    
917    12. The -o option (show only the matching part of a line) for pcregrep was not
918        compatible with GNU grep in that, if there was more than one match in a
919        line, it showed only the first of them. It now behaves in the same way as
920        GNU grep.
921    
922    13. If the -o and -v options were combined for pcregrep, it printed a blank
923        line for every non-matching line. GNU grep prints nothing, and pcregrep now
924        does the same. The return code can be used to tell if there were any
925        non-matching lines.
926    
927    14. Added --file-offsets and --line-offsets to pcregrep.
928    
929    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
930        infinitely looping recursion. The bug was that positive lookaheads were not
931        being skipped when checking for a possible empty match (negative lookaheads
932        and both kinds of lookbehind were skipped).
933    
934    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
935        inclusion of <windows.h> to before rather than after the definition of
936        INVALID_FILE_ATTRIBUTES (patch from David Byron).
937    
938    17. Specifying a possessive quantifier with a specific limit for a Unicode
939        character property caused pcre_compile() to compile bad code, which led at
940        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
941        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
942        caused the error; without that there was no problem.
943    
944    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
945    
946    19. Added --enable-pcretest-libreadline.
947    
948    20. In pcrecpp.cc, the variable 'count' was incremented twice in
949        RE::GlobalReplace(). As a result, the number of replacements returned was
950        double what it should be. I removed one of the increments, but Craig sent a
951        later patch that removed the other one (the right fix) and added unit tests
952        that check the return values (which was not done before).
953    
954    21. Several CMake things:
955    
956        (1) Arranged that, when cmake is used on Unix, the libraries end up with
957            the names libpcre and libpcreposix, not just pcre and pcreposix.
958    
959        (2) The above change means that pcretest and pcregrep are now correctly
960            linked with the newly-built libraries, not previously installed ones.
961    
962        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
963    
964    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
965        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
966        UTF-8 newline character). The key issue is that the pattern starts .*;
967        this means that the match must be either at the beginning, or after a
968        newline. The bug was in the code for advancing after a failed match and
969        checking that the new position followed a newline. It was not taking
970        account of UTF-8 characters correctly.
971    
972    23. PCRE was behaving differently from Perl in the way it recognized POSIX
973        character classes. PCRE was not treating the sequence [:...:] as a
974        character class unless the ... were all letters. Perl, however, seems to
975        allow any characters between [: and :], though of course it rejects as
976        unknown any "names" that contain non-letters, because all the known class
977        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
978        for example, whereas PCRE did not - it did not recognize a POSIX character
979        class. This seemed a bit dangerous, so the code has been changed to be
980        closer to Perl. The behaviour is not identical to Perl, because PCRE will
981        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
982        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
983        Perl does, and where it didn't before.
984    
985    24. Rewrite so as to remove the single use of %n from pcregrep because in some
986        Windows environments %n is disabled by default.
987    
988    
989    Version 7.4 21-Sep-07
990    ---------------------
991    
992    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
993        means that a class such as [\s] counted as "explicit reference to CR or
994        LF". That isn't really right - the whole point of the change was to try to
995        help when there was an actual mention of one of the two characters. So now
996        the change happens only if \r or \n (or a literal CR or LF) character is
997        encountered.
998    
999    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1000        of both had grown to the point where there were only 3 bits left.
1001        Fortunately, there was spare space in the data structure, and so I have
1002        moved the internal flags into a new 16-bit field to free up more option
1003        bits.
1004    
1005    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1006        but did not set the internal JCHANGED flag - either of these is enough to
1007        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1008        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1009        start sets both bits.
1010    
1011    4.  Added options (at build time, compile time, exec time) to change \R from
1012        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1013    
1014    5.  doc/pcresyntax.html was missing from the distribution.
1015    
1016    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1017        compatibility, even though it is no longer used.
1018    
1019    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1020        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1021        windows.h file is present (where different names are used). [This was
1022        reversed later after testing - see 16 below.]
1023    
1024    8.  Changed all #include <config.h> to #include "config.h". There were also
1025        some further <pcre.h> cases that I changed to "pcre.h".
1026    
1027    9.  When pcregrep was used with the --colour option, it missed the line ending
1028        sequence off the lines that it output.
1029    
1030    10. It was pointed out to me that arrays of string pointers cause lots of
1031        relocations when a shared library is dynamically loaded. A technique of
1032        using a single long string with a table of offsets can drastically reduce
1033        these. I have refactored PCRE in four places to do this. The result is
1034        dramatic:
1035    
1036          Originally:                          290
1037          After changing UCP table:            187
1038          After changing error message table:   43
1039          After changing table of "verbs"       36
1040          After changing table of Posix names   22
1041    
1042        Thanks to the folks working on Gregex for glib for this insight.
1043    
1044    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1045        unicode-properties was also set.
1046    
1047    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1048    
1049    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1050        checked only for CRLF.
1051    
1052    14. Added casts to pcretest.c to avoid compiler warnings.
1053    
1054    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1055    
1056    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1057        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1058        entirely. This removes changes made in 7 above.
1059    
1060    17. The CMake files have been updated, and there is now more information about
1061        building with CMake in the NON-UNIX-USE document.
1062    
1063    
1064    Version 7.3 28-Aug-07
1065    ---------------------
1066    
1067     1. In the rejigging of the build system that eventually resulted in 7.1, the
1068        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1069        brackets there is not right, since it causes compilers to look for an
1070        installed pcre.h, not the version that is in the source that is being
1071        compiled (which of course may be different). I have changed it back to:
1072    
1073          #include "pcre.h"
1074    
1075        I have a vague recollection that the change was concerned with compiling in
1076        different directories, but in the new build system, that is taken care of
1077        by the VPATH setting the Makefile.
1078    
1079     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1080        when the subject happened to end in the byte 0x85 (e.g. if the last
1081        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1082        characters but of course it shouldn't be taken as a newline when it is part
1083        of another character. The bug was that, for an unlimited repeat of . in
1084        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1085        characters when looking for a newline.
1086    
1087     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1088    
1089     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1090        in debug output.
1091    
1092     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1093        long printing in the pcrecpp unittest when running under MinGW.
1094    
1095     6. ESC_K was left out of the EBCDIC table.
1096    
1097     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1098        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1099        limit also applies to "virtual nesting" when a pattern is recursive, and in
1100        this case 1000 isn't so big. I have been able to remove this limit at the
1101        expense of backing off one optimization in certain circumstances. Normally,
1102        when pcre_exec() would call its internal match() function recursively and
1103        immediately return the result unconditionally, it uses a "tail recursion"
1104        feature to save stack. However, when a subpattern that can match an empty
1105        string has an unlimited repetition quantifier, it no longer makes this
1106        optimization. That gives it a stack frame in which to save the data for
1107        checking that an empty string has been matched. Previously this was taken
1108        from the 1000-entry workspace that had been reserved. So now there is no
1109        explicit limit, but more stack is used.
1110    
1111     8. Applied Daniel's patches to solve problems with the import/export magic
1112        syntax that is required for Windows, and which was going wrong for the
1113        pcreposix and pcrecpp parts of the library. These were overlooked when this
1114        problem was solved for the main library.
1115    
1116     9. There were some crude static tests to avoid integer overflow when computing
1117        the size of patterns that contain repeated groups with explicit upper
1118        limits. As the maximum quantifier is 65535, the maximum group length was
1119        set at 30,000 so that the product of these two numbers did not overflow a
1120        32-bit integer. However, it turns out that people want to use groups that
1121        are longer than 30,000 bytes (though not repeat them that many times).
1122        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1123        made it possible to implement the integer overflow checks in a much more
1124        dynamic way, which I have now done. The artificial limitation on group
1125        length has been removed - we now have only the limit on the total length of
1126        the compiled pattern, which depends on the LINK_SIZE setting.
1127    
1128    10. Fixed a bug in the documentation for get/copy named substring when
1129        duplicate names are permitted. If none of the named substrings are set, the
1130        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1131        empty string.
1132    
1133    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1134        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1135        because the ] is interpreted as the first data character and the
1136        terminating ] is not found. PCRE has been made compatible with Perl in this
1137        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1138        cause memory overwriting.
1139    
1140    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1141        string has been matched (to stop an infinite loop). It was not recognizing
1142        a conditional subpattern that could match an empty string if that
1143        subpattern was within another subpattern. For example, it looped when
1144        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1145        condition was not nested. This bug has been fixed.
1146    
1147    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1148        past the start of the subject in the presence of bytes with the top bit
1149        set, for example "\x8aBCD".
1150    
1151    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1152        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1153    
1154    14. Optimized (?!) to (*FAIL).
1155    
1156    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1157        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1158        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1159        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1160        does: it's just the validity check that is more restrictive.
1161    
1162    16. Inserted checks for integer overflows during escape sequence (backslash)
1163        processing, and also fixed erroneous offset values for syntax errors during
1164        backslash processing.
1165    
1166    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1167        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1168    
1169    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1170        caused an overrun.
1171    
1172    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1173        something other than just ASCII characters) inside a group that had an
1174        unlimited repeat caused a loop at compile time (while checking to see
1175        whether the group could match an empty string).
1176    
1177    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1178        [\P{Any}] did so. (Error in the code for printing property names.)
1179    
1180    21. An orphan \E inside a character class could cause a crash.
1181    
1182    22. A repeated capturing bracket such as (A)? could cause a wild memory
1183        reference during compilation.
1184    
1185    23. There are several functions in pcre_compile() that scan along a compiled
1186        expression for various reasons (e.g. to see if it's fixed length for look
1187        behind). There were bugs in these functions when a repeated \p or \P was
1188        present in the pattern. These operators have additional parameters compared
1189        with \d, etc, and these were not being taken into account when moving along
1190        the compiled data. Specifically:
1191    
1192        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1193            length.
1194    
1195        (b) An item such as \pL+ within a repeated group could cause crashes or
1196            loops.
1197    
1198        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1199            "reference to non-existent subpattern" error.
1200    
1201        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1202    
1203    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1204        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1205    
1206    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1207    
1208    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1209        character were causing crashes (broken optimization).
1210    
1211    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1212        \p or \P) caused a compile-time loop.
1213    
1214    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1215        break. For example, the unstudied pattern [\r\n]A does not match the string
1216        "\r\nA" because change 7.0/46 below moves the current point on by two
1217        characters after failing to match at the start. However, the pattern \nA
1218        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1219        the same is true. There doesn't seem any very clean way out of this, but
1220        what I have chosen to do makes the common cases work: PCRE now takes note
1221        of whether there can be an explicit match for \r or \n anywhere in the
1222        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1223        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1224        pattern has explicit CR or LF references.
1225    
1226    29. Added (*CR) etc for changing newline setting at start of pattern.
1227    
1228    
1229    Version 7.2 19-Jun-07
1230    ---------------------
1231    
1232     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1233        which is apparently normally available under Windows.
1234    
1235     2. Re-jig the pcregrep tests with different newline settings in an attempt
1236        to make them independent of the local environment's newline setting.
1237    
1238     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1239    
1240     4. Some of the "internals" tests were previously cut out when the link size
1241        was not 2, because the output contained actual offsets. The recent new
1242        "Z" feature of pcretest means that these can be cut out, making the tests
1243        usable with all link sizes.
1244    
1245     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1246        stack recursion. This gives a massive performance boost under BSD, but just
1247        a small improvement under Linux. However, it saves one field in the frame
1248        in all cases.
1249    
1250     6. Added more features from the forthcoming Perl 5.10:
1251    
1252        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1253            recursion call. It refers to the nth most recently opened parentheses.
1254    
1255        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1256            to be opened parentheses.
1257    
1258        (c) Conditions that refer to capturing parentheses can be specified
1259            relatively, for example, (?(-2)... or (?(+3)...
1260    
1261        (d) \K resets the start of the current match so that everything before
1262            is not part of it.
1263    
1264        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1265    
1266        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1267            reference syntax.
1268    
1269        (g) (?| introduces a group in which the numbering of parentheses in each
1270            alternative starts with the same number.
1271    
1272        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1273    
1274     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1275        PCRE_INFO_JCHANGED.
1276    
1277     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1278        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1279        for detecting groups that can match an empty string.
1280    
1281     9. A pattern with a very large number of alternatives (more than several
1282        hundred) was running out of internal workspace during the pre-compile
1283        phase, where pcre_compile() figures out how much memory will be needed. A
1284        bit of new cunning has reduced the workspace needed for groups with
1285        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1286        workspace instead of running out of the 4096 that are available.
1287    
1288    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1289    
1290    11. Applied patch from Google to remove an optimization that didn't quite work.
1291        The report of the bug said:
1292    
1293          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1294          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1295          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1296    
1297    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1298        it matched the wrong number of bytes.
1299    
1300    
1301    Version 7.1 24-Apr-07
1302    ---------------------
1303    
1304     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1305        that is more "standard", making use of automake and other Autotools. There
1306        is some re-arrangement of the files and adjustment of comments consequent
1307        on this.
1308    
1309     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1310        for recursive directory scanning broke on some systems because the files
1311        are not scanned in any specific order and on different systems the order
1312        was different. A call to "sort" has been inserted into RunGrepTest for the
1313        approprate test as a short-term fix. In the longer term there may be an
1314        alternative.
1315    
1316     3. I had an email from Eric Raymond about problems translating some of PCRE's
1317        man pages to HTML (despite the fact that I distribute HTML pages, some
1318        people do their own conversions for various reasons). The problems
1319        concerned the use of low-level troff macros .br and .in. I have therefore
1320        removed all such uses from the man pages (some were redundant, some could
1321        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1322        HTML has been updated to handle .nf/.fi and to complain if it encounters
1323        .br or .in.
1324    
1325     4. Updated comments in configure.ac that get placed in config.h.in and also
1326        arranged for config.h to be included in the distribution, with the name
1327        config.h.generic, for the benefit of those who have to compile without
1328        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1329    
1330     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1331        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1332        makevp.bat for the current PCRE, using the additional files
1333        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1334    
1335     6. A Windows user reported a minor discrepancy with test 2, which turned out
1336        to be caused by a trailing space on an input line that had got lost in his
1337        copy. The trailing space was an accident, so I've just removed it.
1338    
1339     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1340        that is needed.
1341    
1342     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1343        as "const" (a) because they are and (b) because it helps the PHP
1344        maintainers who have recently made a script to detect big data structures
1345        in the php code that should be moved to the .rodata section. I remembered
1346        to update Builducptable as well, so it won't revert if ucptable.h is ever
1347        re-created.
1348    
1349     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1350        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1351        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1352        support is not required. This saves 1.5-2K of code, which is important in
1353        some applications.
1354    
1355        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1356        so as not to refer to the tables, even though these functions will never be
1357        called when UTF-8 support is disabled. Otherwise there are problems with a
1358        shared library.
1359    
1360    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1361    
1362        (a) It was defining its arguments as char * instead of void *.
1363    
1364        (b) It was assuming that all moves were upwards in memory; this was true
1365            a long time ago when I wrote it, but is no longer the case.
1366    
1367        The emulated memove() is provided for those environments that have neither
1368        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1369        is clearly not the case, as these two bugs were recently reported.
1370    
1371    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1372        and Detrail to create the HTML documentation, the .txt form of the man
1373        pages, and it removes trailing spaces from listed files. It also creates
1374        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1375        case, it wraps all the #defines with #ifndefs. This script should be run
1376        before "make dist".
1377    
1378    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1379        with Unicode property support.
1380    
1381        (a) For a maximizing quantifier, if the two different cases of the
1382            character were of different lengths in their UTF-8 codings (there are
1383            some cases like this - I found 11), and the matching function had to
1384            back up over a mixture of the two cases, it incorrectly assumed they
1385            were both the same length.
1386    
1387        (b) When PCRE was configured to use the heap rather than the stack for
1388            recursion during matching, it was not correctly preserving the data for
1389            the other case of a UTF-8 character when checking ahead for a match
1390            while processing a minimizing repeat. If the check also involved
1391            matching a wide character, but failed, corruption could cause an
1392            erroneous result when trying to check for a repeat of the original
1393            character.
1394    
1395    13. Some tidying changes to the testing mechanism:
1396    
1397        (a) The RunTest script now detects the internal link size and whether there
1398            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1399            values substituted by "configure". (The RunGrepTest script already did
1400            this for UTF-8.) The configure.ac script no longer substitutes the
1401            relevant variables.
1402    
1403        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1404            with length and offset values. This means that the output is different
1405            for different internal link sizes. Test 2 is skipped for link sizes
1406            other than 2 because of this, bypassing the problem. Unfortunately,
1407            there was also a test in test 3 (the locale tests) that used /B and
1408            failed for link sizes other than 2. Rather than cut the whole test out,
1409            I have added a new /Z option to pcretest that replaces the length and
1410            offset values with spaces. This is now used to make test 3 independent
1411            of link size. (Test 2 will be tidied up later.)
1412    
1413    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1414        segmentation fault instead of returning the appropriate error message.
1415    
1416    15. In multiline mode when the newline sequence was set to "any", the pattern
1417        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1418        This doesn't seem right; it now treats the CRLF combination as the line
1419        ending, and so does not match in that case. It's only a pattern such as ^$
1420        that would hit this one: something like ^ABC$ would have failed after \r
1421        and then tried again after \r\n.
1422    
1423    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1424        in an attempt to make files that differ only in their line terminators
1425        compare equal. This works on Linux.
1426    
1427    17. Under certain error circumstances pcregrep might try to free random memory
1428        as it exited. This is now fixed, thanks to valgrind.
1429    
1430    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1431        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1432        was because its rules for how to advance for /g after matching an empty
1433        string at the end of a line did not allow for this case. They now check for
1434        it specially.
1435    
1436    20. pcretest is supposed to handle patterns and data of any length, by
1437        extending its buffers when necessary. It was getting this wrong when the
1438        buffer for a data line had to be extended.
1439    
1440    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1441        CRLF as a newline sequence.
1442    
1443    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1444        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1445        I have nevertheless tidied it up.
1446    
1447    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1448    
1449    24. Added a man page for pcre-config.
1450    
1451    
1452    Version 7.0 19-Dec-06
1453    ---------------------
1454    
1455     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1456        moving to gcc 4.1.1.
1457    
1458     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1459        sys/time.h, which is documented as needed for this function. It doesn't
1460        seem to matter on Linux, but it showed up on some releases of OS X.
1461    
1462     3. It seems that there are systems where bytes whose values are greater than
1463        127 match isprint() in the "C" locale. The "C" locale should be the
1464        default when a C program starts up. In most systems, only ASCII printing
1465        characters match isprint(). This difference caused the output from pcretest
1466        to vary, making some of the tests fail. I have changed pcretest so that:
1467    
1468        (a) When it is outputting text in the compiled version of a pattern, bytes
1469            other than 32-126 are always shown as hex escapes.
1470    
1471        (b) When it is outputting text that is a matched part of a subject string,
1472            it does the same, unless a different locale has been set for the match
1473            (using the /L modifier). In this case, it uses isprint() to decide.
1474    
1475     4. Fixed a major bug that caused incorrect computation of the amount of memory
1476        required for a compiled pattern when options that changed within the
1477        pattern affected the logic of the preliminary scan that determines the
1478        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1479        that the computed length was too small. The symptoms of this bug were
1480        either the PCRE error "internal error: code overflow" from pcre_compile(),
1481        or a glibc crash with a message such as "pcretest: free(): invalid next
1482        size (fast)". Examples of patterns that provoked this bug (shown in
1483        pcretest format) are:
1484    
1485          /(?-x: )/x
1486          /(?x)(?-x: \s*#\s*)/
1487          /((?i)[\x{c0}])/8
1488          /(?i:[\x{c0}])/8
1489    
1490        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1491        is now done differently.
1492    
1493     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1494        wrapper classes; (b) implement a new function in the C++ scanner that is
1495        more efficient than the old way of doing things because it avoids levels of
1496        recursion in the regex matching; (c) add a paragraph to the documentation
1497        for the FullMatch() function.
1498    
1499     6. The escape sequence \n was being treated as whatever was defined as
1500        "newline". Not only was this contrary to the documentation, which states
1501        that \n is character 10 (hex 0A), but it also went horribly wrong when
1502        "newline" was defined as CRLF. This has been fixed.
1503    
1504     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1505        was being set to -1 for the "end of line" case (supposedly a value that no
1506        character can have). Though this value is never used (the check for end of
1507        line is "zero bytes in current character"), it caused compiler complaints.
1508        I've changed it to 0xffffffff.
1509    
1510     8. In pcre_version.c, the version string was being built by a sequence of
1511        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1512        string (as it is for production releases) called a macro with an empty
1513        argument. The C standard says the result of this is undefined. The gcc
1514        compiler treats it as an empty string (which was what was wanted) but it is
1515        reported that Visual C gives an error. The source has been hacked around to
1516        avoid this problem.
1517    
1518     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1519        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1520        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1521        of them did).
1522    
1523    10. Originally, pcretest opened its input and output without "b"; then I was
1524        told that "b" was needed in some environments, so it was added for release
1525        5.0 to both the input and output. (It makes no difference on Unix-like
1526        systems.) Later I was told that it is wrong for the input on Windows. I've
1527        now abstracted the modes into two macros, to make it easier to fiddle with
1528        them, and removed "b" from the input mode under Windows.
1529    
1530    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1531    
1532    12. Added -help and --help to pcretest as an official way of being reminded
1533        of the options.
1534    
1535    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1536        and pcrecpp.cc because they annoy compilers at high warning levels.
1537    
1538    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1539    
1540    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1541        scripts are not C programs :-) and which was not noticed because it works
1542        on Linux.
1543    
1544    16. pcretest is supposed to handle any length of pattern and data line (as one
1545        line or as a continued sequence of lines) by extending its input buffer if
1546        necessary. This feature was broken for very long pattern lines, leading to
1547        a string of junk being passed to pcre_compile() if the pattern was longer
1548        than about 50K.
1549    
1550    17. I have done a major re-factoring of the way pcre_compile() computes the
1551        amount of memory needed for a compiled pattern. Previously, there was code
1552        that made a preliminary scan of the pattern in order to do this. That was
1553        OK when PCRE was new, but as the facilities have expanded, it has become
1554        harder and harder to keep it in step with the real compile phase, and there
1555        have been a number of bugs (see for example, 4 above). I have now found a
1556        cunning way of running the real compile function in a "fake" mode that
1557        enables it to compute how much memory it would need, while actually only
1558        ever using a few hundred bytes of working memory and without too many
1559        tests of the mode. This should make future maintenance and development
1560        easier. A side effect of this work is that the limit of 200 on the nesting
1561        depth of parentheses has been removed (though this was never a serious
1562        limitation, I suspect). However, there is a downside: pcre_compile() now
1563        runs more slowly than before (30% or more, depending on the pattern). I
1564        hope this isn't a big issue. There is no effect on runtime performance.
1565    
1566    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1567        newline (only possible for the last line of a file) and it was a
1568        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1569    
1570    19. Added additional timing features to pcretest. (1) The -tm option now times
1571        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1572        separate command line item, by a number that specifies the number of
1573        repeats to use when timing. The default is 50000; this gives better
1574        precision, but takes uncomfortably long for very large patterns.
1575    
1576    20. Extended pcre_study() to be more clever in cases where a branch of a
1577        subpattern has no definite first character. For example, (a*|b*)[cd] would
1578        previously give no result from pcre_study(). Now it recognizes that the
1579        first character must be a, b, c, or d.
1580    
1581    21. There was an incorrect error "recursive call could loop indefinitely" if
1582        a subpattern (or the entire pattern) that was being tested for matching an
1583        empty string contained only one non-empty item after a nested subpattern.
1584        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1585        incorrectly, because the \d was being skipped in the check.
1586    
1587    22. The pcretest program now has a new pattern option /B and a command line
1588        option -b, which is equivalent to adding /B to every pattern. This causes
1589        it to show the compiled bytecode, without the additional information that
1590        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1591        is the same as /B/I).
1592    
1593    23. A new optimization is now able automatically to treat some sequences such
1594        as a*b as a*+b. More specifically, if something simple (such as a character
1595        or a simple class like \d) has an unlimited quantifier, and is followed by
1596        something that cannot possibly match the quantified thing, the quantifier
1597        is automatically "possessified".
1598    
1599    24. A recursive reference to a subpattern whose number was greater than 39
1600        went wrong under certain circumstances in UTF-8 mode. This bug could also
1601        have affected the operation of pcre_study().
1602    
1603    25. Realized that a little bit of performance could be had by replacing
1604        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1605    
1606    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1607    
1608    27. Possessive quantifiers such as a++ were previously implemented by turning
1609        them into atomic groups such as ($>a+). Now they have their own opcodes,
1610        which improves performance. This includes the automatically created ones
1611        from 23 above.
1612    
1613    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1614        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1615        the first matched character to be a colon. This applied both to named and
1616        numbered groups.
1617    
1618    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1619    
1620    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1621        building PCRE on Windows easier, so I have included it in the distribution.
1622    
1623    31. There is now a check in pcretest against a ridiculously large number being
1624        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1625        loop, the loop is abandoned.
1626    
1627    32. Forward references to subpatterns in conditions such as (?(2)...) where
1628        subpattern 2 is defined later cause pcre_compile() to search forwards in
1629        the pattern for the relevant set of parentheses. This search went wrong
1630        when there were unescaped parentheses in a character class, parentheses
1631        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1632    
1633    33. "Subroutine" calls and backreferences were previously restricted to
1634        referencing subpatterns earlier in the regex. This restriction has now
1635        been removed.
1636    
1637    34. Added a number of extra features that are going to be in Perl 5.10. On the
1638        whole, these are just syntactic alternatives for features that PCRE had
1639        previously implemented using the Python syntax or my own invention. The
1640        other formats are all retained for compatibility.
1641    
1642        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1643            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1644            also .NET compatible.
1645    
1646        (b) A recursion or subroutine call to a named group can now be defined as
1647            (?&name) as well as (?P>name).
1648    
1649        (c) A backreference to a named group can now be defined as \k<name> or
1650            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1651            5.10, are also .NET compatible.
1652    
1653        (d) A conditional reference to a named group can now use the syntax
1654            (?(<name>) or (?('name') as well as (?(name).
1655    
1656        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1657            groups (named and numbered) that are never evaluated inline, but can be
1658            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1659            is always false. There may be only one alternative in such a group.
1660    
1661        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1662            as the simple (?(R). The condition is true only if the most recent
1663            recursion is that of the given number or name. It does not search out
1664            through the entire recursion stack.
1665    
1666        (g) The escape \gN or \g{N} has been added, where N is a positive or
1667            negative number, specifying an absolute or relative reference.
1668    
1669    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1670        some "unreachable code" warnings.
1671    
1672    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1673        things, this adds five new scripts.
1674    
1675    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1676        There were also incompatibilities regarding the handling of \Q..\E inside
1677        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1678        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1679    
1680    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1681        matches an empty string, and forcibly breaks the loop. There were bugs in
1682        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1683        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1684        separate and independent bugs (that affected different cases) have been
1685        fixed.
1686    
1687    39. Refactored the code to abolish the use of different opcodes for small
1688        capturing bracket numbers. This is a tidy that I avoided doing when I
1689        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1690        The new approach is not only tidier, it makes it possible to reduce the
1691        memory needed to fix the previous bug (38).
1692    
1693    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1694        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1695        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1696        mode.
1697    
1698    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1699        report.
1700    
1701    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1702        copy construction and assignment in the C++ wrapper.
1703    
1704    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1705        couple of bugs that could have given wrong results in the "--newline=crlf"
1706        case.
1707    
1708    44. Added a number of casts and did some reorganization of signed/unsigned int
1709        variables following suggestions from Dair Grant. Also renamed the variable
1710        "this" as "item" because it is a C++ keyword.
1711    
1712    45. Arranged for dftables to add
1713    
1714          #include "pcre_internal.h"
1715    
1716        to pcre_chartables.c because without it, gcc 4.x may remove the array
1717        definition from the final binary if PCRE is built into a static library and
1718        dead code stripping is activated.
1719    
1720    46. For an unanchored pattern, if a match attempt fails at the start of a
1721        newline sequence, and the newline setting is CRLF or ANY, and the next two
1722        characters are CRLF, advance by two characters instead of one.
1723    
1724    
1725    Version 6.7 04-Jul-06
1726    ---------------------
1727    
1728     1. In order to handle tests when input lines are enormously long, pcretest has
1729        been re-factored so that it automatically extends its buffers when
1730        necessary. The code is crude, but this _is_ just a test program. The
1731        default size has been increased from 32K to 50K.
1732    
1733     2. The code in pcre_study() was using the value of the re argument before
1734        testing it for NULL. (Of course, in any sensible call of the function, it
1735        won't be NULL.)
1736    
1737     3. The memmove() emulation function in pcre_internal.h, which is used on
1738        systems that lack both memmove() and bcopy() - that is, hardly ever -
1739        was missing a "static" storage class specifier.
1740    
1741     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1742        containing an extended class (one that cannot be represented by a bitmap
1743        because it contains high-valued characters or Unicode property items, e.g.
1744        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1745        pattern, but PCRE should not loop if you do not (it no longer does).
1746        [Detail: two cases were found: (a) a repeated subpattern containing an
1747        extended class; (b) a recursive reference to a subpattern that followed a
1748        previous extended class. It wasn't skipping over the extended class
1749        correctly when UTF-8 mode was not set.]
1750    
1751     5. A negated single-character class was not being recognized as fixed-length
1752        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1753        compile error "lookbehind assertion is not fixed length".
1754    
1755     6. The RunPerlTest auxiliary script was showing an unexpected difference
1756        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1757        write a Perl script that can interpret lines of an input file either as
1758        byte characters or as UTF-8, which is what "perltest" was being required to
1759        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1760        can't do is switch easily at run time between having the "use utf8;" pragma
1761        or not. In the end, I fudged it by using the RunPerlTest script to insert
1762        "use utf8;" explicitly for the UTF-8 tests.
1763    
1764     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1765        the end of the subject string, contrary to the documentation and to what
1766        Perl does. This was true of both matching functions. Now it matches only at
1767        the start of the subject and immediately after *internal* newlines.
1768    
1769     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1770        a pointer to an int instead of a pointer to an unsigned long int. This
1771        caused problems on 64-bit systems.
1772    
1773     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1774        instance of the 'standard' template library not being so standard".
1775    
1776    10. There was no check on the number of named subpatterns nor the maximum
1777        length of a subpattern name. The product of these values is used to compute
1778        the size of the memory block for a compiled pattern. By supplying a very
1779        long subpattern name and a large number of named subpatterns, the size
1780        computation could be caused to overflow. This is now prevented by limiting
1781        the length of names to 32 characters, and the number of named subpatterns
1782        to 10,000.
1783    
1784    11. Subpatterns that are repeated with specific counts have to be replicated in
1785        the compiled pattern. The size of memory for this was computed from the
1786        length of the subpattern and the repeat count. The latter is limited to
1787        65535, but there was no limit on the former, meaning that integer overflow
1788        could in principle occur. The compiled length of a repeated subpattern is
1789        now limited to 30,000 bytes in order to prevent this.
1790    
1791    12. Added the optional facility to have named substrings with the same name.
1792    
1793    13. Added the ability to use a named substring as a condition, using the
1794        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1795        are numbers (not recommended). Forward references are permitted.
1796    
1797    14. Added forward references in named backreferences (if you see what I mean).
1798    
1799    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1800        pattern could run off the end of the subject. For example, the pattern
1801        "(?s)(.{1,5})"8 did this with the subject "ab".
1802    
1803    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1804        PCRE_CASELESS was set when matching characters that were quantified with ?
1805        or *.
1806    
1807    17. A character class other than a single negated character that had a minimum
1808        but no maximum quantifier - for example [ab]{6,} - was not handled
1809        correctly by pce_dfa_exec(). It would match only one character.
1810    
1811    18. A valid (though odd) pattern that looked like a POSIX character
1812        class but used an invalid character after [ (for example [[,abc,]]) caused
1813        pcre_compile() to give the error "Failed: internal error: code overflow" or
1814        in some cases to crash with a glibc free() error. This could even happen if
1815        the pattern terminated after [[ but there just happened to be a sequence of
1816        letters, a binary zero, and a closing ] in the memory that followed.
1817    
1818    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1819        over the years. Originally (before any Unicode support), just the bottom 8
1820        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1821        output from "man perlunicode" includes this:
1822    
1823          The regular expression compiler produces polymorphic opcodes.  That
1824          is, the pattern adapts to the data and automatically switches to
1825          the Unicode character scheme when presented with Unicode data--or
1826          instead uses a traditional byte scheme when presented with byte
1827          data.
1828    
1829        Sadly, a wide octal escape does not cause a switch, and in a string with
1830        no other multibyte characters, these octal escapes are treated as before.
1831        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1832        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1833        Unicode string.
1834    
1835        I have not perpetrated such confusion in PCRE. Up till now, it took just
1836        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1837        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1838        translate to the appropriate multibyte character.
1839    
1840    29. Applied some refactoring to reduce the number of warnings from Microsoft
1841        and Borland compilers. This has included removing the fudge introduced
1842        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1843        a warning about an unused variable.
1844    
1845    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1846        characters since release 4.0, because Perl (from release 5.004) does not.
1847        [Or at least, is documented not to: some releases seem to be in conflict
1848        with the documentation.] However, when a pattern was studied with
1849        pcre_study() and all its branches started with \s, PCRE still included VT
1850        as a possible starting character. Of course, this did no harm; it just
1851        caused an unnecessary match attempt.
1852    
1853    22. Removed a now-redundant internal flag bit that recorded the fact that case
1854        dependency changed within the pattern. This was once needed for "required
1855        byte" processing, but is no longer used. This recovers a now-scarce options
1856        bit. Also moved the least significant internal flag bit to the most-
1857        significant bit of the word, which was not previously used (hangover from
1858        the days when it was an int rather than a uint) to free up another bit for
1859        the future.
1860    
1861    23. Added support for CRLF line endings as well as CR and LF. As well as the
1862        default being selectable at build time, it can now be changed at runtime
1863        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1864        specify that it is scanning data with non-default line endings.
1865    
1866    24. Changed the definition of CXXLINK to make it agree with the definition of
1867        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1868    
1869    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1870        recursions. This makes a big different to stack usage for some patterns.
1871    
1872    26. If a subpattern containing a named recursion or subroutine reference such
1873        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1874        the space required for the compiled pattern went wrong and gave too small a
1875        value. Depending on the environment, this could lead to "Failed: internal
1876        error: code overflow at offset 49" or "glibc detected double free or
1877        corruption" errors.
1878    
1879    27. Applied patches from Google (a) to support the new newline modes and (b) to
1880        advance over multibyte UTF-8 characters in GlobalReplace.
1881    
1882    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1883        difference for some implementation of PCRE in some Windows version.
1884    
1885    29. Added some extra testing facilities to pcretest:
1886    
1887        \q<number>   in a data line sets the "match limit" value
1888        \Q<number>   in a data line sets the "match recursion limt" value
1889        -S <number>  sets the stack size, where <number> is in megabytes
1890    
1891        The -S option isn't available for Windows.
1892    
1893    
1894    Version 6.6 06-Feb-06
1895    ---------------------
1896    
1897     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1898        in pcreposix.h. I have copied the definition from pcre.h.
1899    
1900     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1901        because pcre.h is no longer a built file.
1902    
1903     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1904        not normally included in the compiled code.
1905    
1906    
1907    Version 6.5 01-Feb-06
1908    ---------------------
1909    
1910     1. When using the partial match feature with pcre_dfa_exec(), it was not
1911        anchoring the second and subsequent partial matches at the new starting
1912        point. This could lead to incorrect results. For example, with the pattern
1913        /1234/, partially matching against "123" and then "a4" gave a match.
1914    
1915     2. Changes to pcregrep:
1916    
1917        (a) All non-match returns from pcre_exec() were being treated as failures
1918            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1919            error message is output. Some extra information is given for the
1920            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1921            probably the only errors that are likely to be caused by users (by
1922            specifying a regex that has nested indefinite repeats, for instance).
1923            If there are more than 20 of these errors, pcregrep is abandoned.
1924    
1925        (b) A binary zero was treated as data while matching, but terminated the
1926            output line if it was written out. This has been fixed: binary zeroes
1927            are now no different to any other data bytes.
1928    
1929        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1930            used to set a locale for matching. The --locale=xxxx long option has
1931            been added (no short equivalent) to specify a locale explicitly on the
1932            pcregrep command, overriding the environment variables.
1933    
1934        (d) When -B was used with -n, some line numbers in the output were one less
1935            than they should have been.
1936    
1937        (e) Added the -o (--only-matching) option.
1938    
1939        (f) If -A or -C was used with -c (count only), some lines of context were
1940            accidentally printed for the final match.
1941    
1942        (g) Added the -H (--with-filename) option.
1943    
1944        (h) The combination of options -rh failed to suppress file names for files
1945            that were found from directory arguments.
1946    
1947        (i) Added the -D (--devices) and -d (--directories) options.
1948    
1949        (j) Added the -F (--fixed-strings) option.
1950    
1951        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1952    
1953        (l) Added the --colo(u)r option.
1954    
1955        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1956            is not present by default.
1957    
1958     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1959        items such as (?R) or (?1), when the recursion could match a number of
1960        alternatives. If it matched one of the alternatives, but subsequently,
1961        outside the recursion, there was a failure, the code tried to back up into
1962        the recursion. However, because of the way PCRE is implemented, this is not
1963        possible, and the result was an incorrect result from the match.
1964    
1965        In order to prevent this happening, the specification of recursion has
1966        been changed so that all such subpatterns are automatically treated as
1967        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1968    
1969     4. I had overlooked the fact that, in some locales, there are characters for
1970        which isalpha() is true but neither isupper() nor islower() are true. In
1971        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1972        and ordfeminine) are like this. This affected the treatment of \w and \W
1973        when they appeared in character classes, but not when they appeared outside
1974        a character class. The bit map for "word" characters is now created
1975        separately from the results of isalnum() instead of just taking it from the
1976        upper, lower, and digit maps. (Plus the underscore character, of course.)
1977    
1978     5. The above bug also affected the handling of POSIX character classes such as
1979        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1980        permanent tables. Instead, the bit maps for such a class were previously
1981        created as the appropriate unions of the upper, lower, and digit bitmaps.
1982        Now they are created by subtraction from the [[:word:]] class, which has
1983        its own bitmap.
1984    
1985     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1986        It is created by subtracting the vertical space characters (\x09, \x0a,
1987        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1988        subtraction was done in the overall bitmap for a character class, meaning
1989        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1990        be recognized. This bug has been fixed.
1991    
1992     7. Patches from the folks at Google:
1993    
1994          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1995          real life, but is still worth protecting against".
1996    
1997          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1998          regular expressions".
1999    
2000          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2001          have it.
2002    
2003          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2004          "configure" and the latter not, in order to fix a problem somebody had
2005          with compiling the Arg class on HP-UX.
2006    
2007          (e) Improve the error-handling of the C++ wrapper a little bit.
2008    
2009          (f) New tests for checking recursion limiting.
2010    
2011     8. The pcre_memmove() function, which is used only if the environment does not
2012        have a standard memmove() function (and is therefore rarely compiled),
2013        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2014        returning a result (though PCRE never actually uses the result).
2015    
2016     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2017        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2018        returned instead of calling malloc() with an overflowing number that would
2019        most likely cause subsequent chaos.
2020    
2021    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2022    
2023    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2024        with this option is matched, the nmatch and pmatch options of regexec() are
2025        ignored.
2026    
2027    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2028        provided in case anyone wants to the the POSIX interface with UTF-8
2029        strings.
2030    
2031    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2032        C++ linking (needed for some HP-UX environments).
2033    
2034    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2035        (unused parameter) and in the pcre_printint() function (omitted "default"
2036        switch label when the default is to do nothing).
2037    
2038    15. Added some code to make it possible, when PCRE is compiled as a C++
2039        library, to replace subject pointers for pcre_exec() with a smart pointer
2040        class, thus making it possible to process discontinuous strings.
2041    
2042    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2043        much the same function. They were added by different people who were trying
2044        to make PCRE easy to compile on non-Unix systems. It has been suggested
2045        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2046        for compiling on Windows systems. I have therefore replaced it with
2047        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2048        defaults to "extern" for C or "extern C" for C++, which works fine on
2049        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2050        SCOPE with something explicit in config.h. In addition:
2051    
2052        (a) pcreposix.h still had just "extern" instead of either of these macros;
2053            I have replaced it with PCRE_DATA_SCOPE.
2054    
2055        (b) Functions such as _pcre_xclass(), which are internal to the library,
2056            but external in the C sense, all had PCRE_EXPORT in their definitions.
2057            This is apparently wrong for the Windows case, so I have removed it.
2058            (It makes no difference on Unix-like systems.)
2059    
2060    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2061        of recursive calls to match(). This is different to MATCH_LIMIT because
2062        that limits the total number of calls to match(), not all of which increase
2063        the depth of recursion. Limiting the recursion depth limits the amount of
2064        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2065        when PCRE is compiled, and changed at run time. A patch from Google adds
2066        this functionality to the C++ interface.
2067    
2068    18. Changes to the handling of Unicode character properties:
2069    
2070        (a) Updated the table to Unicode 4.1.0.
2071    
2072        (b) Recognize characters that are not in the table as "Cn" (undefined).
2073    
2074        (c) I revised the way the table is implemented to a much improved format
2075            which includes recognition of ranges. It now supports the ranges that
2076            are defined in UnicodeData.txt, and it also amalgamates other
2077            characters into ranges. This has reduced the number of entries in the
2078            table from around 16,000 to around 3,000, thus reducing its size
2079            considerably. I realized I did not need to use a tree structure after
2080            all - a binary chop search is just as efficient. Having reduced the
2081            number of entries, I extended their size from 6 bytes to 8 bytes to
2082            allow for more data.
2083    
2084        (d) Added support for Unicode script names via properties such as \p{Han}.
2085    
2086    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2087        matching that character.
2088    
2089    20. When matching a repeated Unicode property with a minimum greater than zero,
2090        (for example \pL{2,}), PCRE could look past the end of the subject if it
2091        reached it while seeking the minimum number of characters. This could
2092        happen only if some of the characters were more than one byte long, because
2093        there is a check for at least the minimum number of bytes.
2094    
2095    21. Refactored the implementation of \p and \P so as to be more general, to
2096        allow for more different types of property in future. This has changed the
2097        compiled form incompatibly. Anybody with saved compiled patterns that use
2098        \p or \P will have to recompile them.
2099    
2100    22. Added "Any" and "L&" to the supported property types.
2101    
2102    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2103        but give a compile time error if the value is greater than 0xff.
2104    
2105    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2106        accidentally not being installed or uninstalled.
2107    
2108    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2109        made were to insert the current release number. This seemed silly, because
2110        it made things harder for people building PCRE on systems that don't run
2111        "configure". I have turned pcre.h into a distributed file, no longer built
2112        by "configure", with the version identification directly included. There is
2113        no longer a pcre.h.in file.
2114    
2115        However, this change necessitated a change to the pcre-config script as
2116        well. It is built from pcre-config.in, and one of the substitutions was the
2117        release number. I have updated configure.ac so that ./configure now finds
2118        the release number by grepping pcre.h.
2119    
2120    26. Added the ability to run the tests under valgrind.
2121    
2122    
2123    Version 6.4 05-Sep-05
2124    ---------------------
2125    
2126     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2127        "--" to be printed when multiple files were scanned, even when none of the
2128        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2129        consider it to be a bug, and have restored the previous behaviour.
2130    
2131     2. A couple of code tidies to get rid of compiler warnings.
2132    
2133     3. The pcretest program used to cheat by referring to symbols in the library
2134        whose names begin with _pcre_. These are internal symbols that are not
2135        really supposed to be visible externally, and in some environments it is
2136        possible to suppress them. The cheating is now confined to including
2137        certain files from the library's source, which is a bit cleaner.
2138    
2139     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2140        file's purpose clearer.
2141    
2142     5. Reorganized pcre_ucp_findchar().
2143    
2144    
2145    Version 6.3 15-Aug-05
2146    ---------------------
2147    
2148     1. The file libpcre.pc.in did not have general read permission in the tarball.
2149    
2150     2. There were some problems when building without C++ support:
2151    
2152        (a) If C++ support was not built, "make install" and "make test" still
2153            tried to test it.
2154    
2155        (b) There were problems when the value of CXX was explicitly set. Some
2156            changes have been made to try to fix these, and ...
2157    
2158        (c) --disable-cpp can now be used to explicitly disable C++ support.
2159    
2160        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2161            backslash in a target when C++ was disabled. This confuses some
2162            versions of "make", apparently. Using an intermediate variable solves
2163            this. (Same for CPP_LOBJ.)
2164    
2165     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2166        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2167        necessary on certain architectures.
2168    
2169     4. Added a setting of -export-symbols-regex to the link command to remove
2170        those symbols that are exported in the C sense, but actually are local
2171        within the library, and not documented. Their names all begin with
2172        "_pcre_". This is not a perfect job, because (a) we have to except some
2173        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2174        available (and never for static libraries). I have made a note to try to
2175        find a way round (a) in the future.
2176    
2177    
2178    Version 6.2 01-Aug-05
2179    ---------------------
2180    
2181     1. There was no test for integer overflow of quantifier values. A construction
2182        such as {1111111111111111} would give undefined results. What is worse, if
2183        a minimum quantifier for a parenthesized subpattern overflowed and became
2184        negative, the calculation of the memory size went wrong. This could have
2185        led to memory overwriting.
2186    
2187     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2188    
2189     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2190        operating environments where this matters.
2191    
2192     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2193        PCRE options from within the C++ wrapper.
2194    
2195     5. Named capturing subpatterns were not being correctly counted when a pattern
2196        was compiled. This caused two problems: (a) If there were more than 100
2197        such subpatterns, the calculation of the memory needed for the whole
2198        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2199        back references of the form \12, where the number was greater than 9, were
2200        not recognized as back references, even though there were sufficient
2201        previous subpatterns.
2202    
2203     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2204        versions of gcc, e.g. 2.95.4.
2205    
2206    
2207    Version 6.1 21-Jun-05
2208    ---------------------
2209    
2210     1. There was one reference to the variable "posix" in pcretest.c that was not
2211        surrounded by "#if !defined NOPOSIX".
2212    
2213     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2214        the cross-check on the old pcre_info() function, for the benefit of the
2215        cut-down version of PCRE that is currently imported into Exim.
2216    
2217     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2218        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2219        patterns that start (?i) but I don't think that matters. The use of (?i) is
2220        just an example; this all applies to the other options as well.
2221    
2222     4. Since libtool seems to echo the compile commands it is issuing, the output
2223        from "make" can be reduced a bit by putting "@" in front of each libtool
2224        compile command.
2225    
2226     5. Patch from the folks at Google for configure.in to be a bit more thorough
2227        in checking for a suitable C++ installation before trying to compile the
2228        C++ stuff. This should fix a reported problem when a compiler was present,
2229        but no suitable headers.
2230    
2231     6. The man pages all had just "PCRE" as their title. I have changed them to
2232        be the relevant file name. I have also arranged that these names are
2233        retained in the file doc/pcre.txt, which is a concatenation in text format
2234        of all the man pages except the little individual ones for each function.
2235    
2236     7. The NON-UNIX-USE file had not been updated for the different set of source
2237        files that come with release 6. I also added a few comments about the C++
2238        wrapper.
2239    
2240    
2241    Version 6.0 07-Jun-05
2242    ---------------------
2243    
2244     1. Some minor internal re-organization to help with my DFA experiments.
2245    
2246     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2247        didn't matter for the library itself when fully configured, but did matter
2248        when compiling without UCP support, or within Exim, where the ucp files are
2249        not imported.
2250    
2251     3. Refactoring of the library code to split up the various functions into
2252        different source modules. The addition of the new DFA matching code (see
2253        below) to a single monolithic source would have made it really too
2254        unwieldy, quite apart from causing all the code to be include in a
2255        statically linked application, when only some functions are used. This is
2256        relevant even without the DFA addition now that patterns can be compiled in
2257        one application and matched in another.
2258    
2259        The downside of splitting up is that there have to be some external
2260        functions and data tables that are used internally in different modules of
2261        the library but which are not part of the API. These have all had their
2262        names changed to start with "_pcre_" so that they are unlikely to clash
2263        with other external names.
2264    
2265     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2266        a different (DFA) algorithm. Although it is slower than the original
2267        function, it does have some advantages for certain types of matching
2268        problem.
2269    
2270     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2271        including restarting after a partial match.
2272    
2273     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2274        defined when compiling for Windows was sent to me. I have put it into the
2275        code, though I have no means of testing or verifying it.
2276    
2277     7. Added the pcre_refcount() auxiliary function.
2278    
2279     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2280        match before or at the first newline in the subject string. In pcretest,
2281        the /f option on a pattern can be used to set this.
2282    
2283     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2284        would behave wrongly. This has been present in PCRE since release 4.0.
2285    
2286    10. A number of changes to the pcregrep command:
2287    
2288        (a) Refactored how -x works; insert ^(...)$ instead of setting
2289            PCRE_ANCHORED and checking the length, in preparation for adding
2290            something similar for -w.
2291    
2292        (b) Added the -w (match as a word) option.
2293    
2294        (c) Refactored the way lines are read and buffered so as to have more
2295            than one at a time available.
2296    
2297        (d) Implemented a pcregrep test script.
2298    
2299        (e) Added the -M (multiline match) option. This allows patterns to match
2300            over several lines of the subject. The buffering ensures that at least
2301            8K, or the rest of the document (whichever is the shorter) is available
2302            for matching (and similarly the previous 8K for lookbehind assertions).
2303    
2304        (f) Changed the --help output so that it now says
2305    
2306              -w, --word-regex(p)
2307    
2308            instead of two lines, one with "regex" and the other with "regexp"
2309            because that confused at least one person since the short forms are the
2310            same. (This required a bit of code, as the output is generated
2311            automatically from a table. It wasn't just a text change.)
2312    
2313        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2314            option but starts with a hyphen. Could be a pattern or a path name
2315            starting with a hyphen, for instance.
2316    
2317        (h) "-" can be given as a file name to represent stdin.
2318    
2319        (i) When file names are being printed, "(standard input)" is used for
2320            the standard input, for compatibility with GNU grep. Previously
2321            "<stdin>" was used.
2322    
2323        (j) The option --label=xxx can be used to supply a name to be used for
2324            stdin when file names are being printed. There is no short form.
2325    
2326        (k) Re-factored the options decoding logic because we are going to add
2327            two more options that take data. Such options can now be given in four
2328            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2329    
2330        (l) Added the -A, -B, and -C options for requesting that lines of context
2331            around matches be printed.
2332    
2333        (m) Added the -L option to print the names of files that do not contain
2334            any matching lines, that is, the complement of -l.
2335    
2336        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2337            continue to scan other files.
2338    
2339        (o) The -s option was incorrectly implemented. For compatibility with other
2340            greps, it now suppresses the error message for a non-existent or non-
2341            accessible file (but not the return code). There is a new option called
2342            -q that suppresses the output of matching lines, which was what -s was
2343            previously doing.
2344    
2345        (p) Added --include and --exclude options to specify files for inclusion
2346            and exclusion when recursing.
2347    
2348    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2349        Hopefully, it now does.
2350    
2351    12. Missing cast in pcre_study().
2352    
2353    13. Added an "uninstall" target to the makefile.
2354    
2355    14. Replaced "extern" in the function prototypes in Makefile.in with
2356        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2357        world, but is set differently for Windows.
2358    
2359    15. Added a second compiling function called pcre_compile2(). The only
2360        difference is that it has an extra argument, which is a pointer to an
2361        integer error code. When there is a compile-time failure, this is set
2362        non-zero, in addition to the error test pointer being set to point to an
2363        error message. The new argument may be NULL if no error number is required
2364        (but then you may as well call pcre_compile(), which is now just a
2365        wrapper). This facility is provided because some applications need a
2366        numeric error indication, but it has also enabled me to tidy up the way
2367        compile-time errors are handled in the POSIX wrapper.
2368    
2369    16. Added VPATH=.libs to the makefile; this should help when building with one
2370        prefix path and installing with another. (Or so I'm told by someone who
2371        knows more about this stuff than I do.)
2372    
2373    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2374        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2375        match everything, including newlines. This is not POSIX-compatible, but
2376        somebody wanted the feature. From pcretest it can be activated by using
2377        both the P and the s flags.
2378    
2379    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2380    
2381    19. libpcre.pc was being incorrectly installed as executable.
2382    
2383    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2384        it now also looks for '\r' so that it will work unmodified on Windows.
2385    
2386    21. Added Google's contributed C++ wrapper to the distribution.
2387    
2388    22. Added some untidy missing memory free() calls in pcretest, to keep
2389        Electric Fence happy when testing.
2390    
2391    
2392    
2393    Version 5.0 13-Sep-04
2394    ---------------------
2395    
2396     1. Internal change: literal characters are no longer packed up into items
2397        containing multiple characters in a single byte-string. Each character
2398        is now matched using a separate opcode. However, there may be more than one
2399        byte in the character in UTF-8 mode.
2400    
2401     2. The pcre_callout_block structure has two new fields: pattern_position and
2402        next_item_length. These contain the offset in the pattern to the next match
2403        item, and its length, respectively.
2404    
2405     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2406        insertion of callouts before each pattern item. Added the /C option to
2407        pcretest to make use of this.
2408    
2409     4. On the advice of a Windows user, the lines
2410    
2411          #if defined(_WIN32) || defined(WIN32)
2412          _setmode( _fileno( stdout ), 0x8000 );
2413          #endif  /* defined(_WIN32) || defined(WIN32) */
2414    
2415        have been added to the source of pcretest. This apparently does useful
2416        magic in relation to line terminators.
2417    
2418     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2419        for the benefit of those environments where the "b" makes a difference.
2420    
2421     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2422        to know about it. I have put a hack into configure.in that adds in code
2423        to set GCC=yes if CC=icc. This seems to end up at a point in the
2424        generated configure script that is early enough to affect the setting of
2425        compiler options, which is what is needed, but I have no means of testing
2426        whether it really works. (The user who reported this had patched the
2427        generated configure script, which of course I cannot do.)
2428    
2429        LATER: After change 22 below (new libtool files), the configure script
2430        seems to know about icc (and also ecc). Therefore, I have commented out
2431        this hack in configure.in.
2432    
2433     7. Added support for pkg-config (2 patches were sent in).
2434    
2435     8. Negated POSIX character classes that used a combination of internal tables
2436        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2437        [[:^ascii]]. Typically, they would match almost any characters. The other
2438        POSIX classes were not broken in this way.
2439    
2440     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2441        to find the match, as PCRE was deluded into thinking that the match had to
2442        start at the start point or following a newline. The same bug applied to
2443        patterns with negative forward assertions or any backward assertions
2444        preceding ".*" at the start, unless the pattern required a fixed first
2445        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2446    
2447    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2448        starting at the last subject character, bytes beyond the end of the subject
2449        string were read.
2450    
2451    11. Renamed the variable "class" as "classbits" to make life easier for C++
2452        users. (Previously there was a macro definition, but it apparently wasn't
2453        enough.)
2454    
2455    12. Added the new field "tables" to the extra data so that tables can be passed
2456        in at exec time, or the internal tables can be re-selected. This allows
2457        a compiled regex to be saved and re-used at a later time by a different
2458        program that might have everything at different addresses.
2459    
2460    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2461        -R library as well as a -L library.
2462    
2463    14. The debugging options of pcretest (-d on the command line or D on a
2464        pattern) showed incorrect output for anything following an extended class
2465        that contained multibyte characters and which was followed by a quantifier.
2466    
2467    15. Added optional support for general category Unicode character properties
2468        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2469        support. It adds about 90K to the size of the library. The meanings of the
2470        inbuilt class escapes such as \d and \s have NOT been changed.
2471    
2472    16. Updated pcredemo.c to include calls to free() to release the memory for the
2473        compiled pattern.
2474    
2475    17. The generated file chartables.c was being created in the source directory
2476        instead of in the building directory. This caused the build to fail if the
2477        source directory was different from the building directory, and was
2478        read-only.
2479    
2480    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2481        file. No doubt somebody will tell me if they don't make sense... Also added
2482        Dan Mooney's comments about building on OpenVMS.
2483    
2484    19. Added support for partial matching via the PCRE_PARTIAL option for
2485        pcre_exec() and the \P data escape in pcretest.
2486    
2487    20. Extended pcretest with 3 new pattern features:
2488    
2489        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2490              write the compiled pattern to the file whose name is "rest-of-line".
2491              This is a straight binary dump of the data, with the saved pointer to
2492              the character tables forced to be NULL. The study data, if any, is
2493              written too. After writing, pcretest reads a new pattern.
2494    
2495        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2496              compiled pattern from the given file. There must not be any
2497              occurrences of "<" in the file name (pretty unlikely); if there are,
2498              pcretest will instead treat the initial "<" as a pattern delimiter.
2499              After reading in the pattern, pcretest goes on to read data lines as
2500              usual.
2501    
2502        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2503              and 16-bit fields in a compiled pattern, to simulate a pattern that
2504              was compiled on a host of opposite endianness.
2505    
2506    21. The pcre-exec() function can now cope with patterns that were compiled on
2507        hosts of opposite endianness, with this restriction:
2508    
2509          As for any compiled expression that is saved and used later, the tables
2510          pointer field cannot be preserved; the extra_data field in the arguments
2511          to pcre_exec() should be used to pass in a tables address if a value
2512          other than the default internal tables were used at compile time.
2513    
2514    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2515        now diagnosed as an error. Previously, most of the time, a negative number
2516        would have been treated as zero, but if in addition "ovector" was passed as
2517        NULL, a crash could occur.
2518    
2519    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2520        new versions from the libtool 1.5 distribution (the last one is a copy of
2521        a file called libtool.m4). This seems to have fixed the need to patch
2522        "configure" to support Darwin 1.3 (which I used to do). However, I still
2523        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2524        workstation).
2525    
2526    24. Changed the PCRE licence to be the more standard "BSD" licence.
2527    
2528    
2529    Version 4.5 01-Dec-03
2530    ---------------------
2531    
2532     1. There has been some re-arrangement of the code for the match() function so
2533        that it can be compiled in a version that does not call itself recursively.
2534        Instead, it keeps those local variables that need separate instances for
2535        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2536        needs to "recurse". Keeping track of where control must go is done by means
2537        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2538        hide most of the details from the main code, and operates only if
2539        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2540        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2541        operating.
2542    
2543        To make it easier for callers to provide specially tailored get/free
2544        functions for this usage, two new functions, pcre_stack_malloc, and
2545        pcre_stack_free, are used. They are always called in strict stacking order,
2546        and the size of block requested is always the same.
2547    
2548        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2549        PCRE has been compiled to use the stack or the heap for recursion. The
2550        -C option of pcretest uses this to show which version is compiled.
2551    
2552        A new data escape \S, is added to pcretest; it causes the amounts of store
2553        obtained and freed by both kinds of malloc/free at match time to be added
2554        to the output.
2555    
2556     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2557        what's available on my current Linux desktop machine.
2558    
2559     3. When matching a UTF-8 string, the test for a valid string at the start has
2560        been extended. If start_offset is not zero, PCRE now checks that it points
2561        to a byte that is the start of a UTF-8 character. If not, it returns
2562        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2563        this is necessary because there may be backward assertions in the pattern.
2564        When matching the same subject several times, it may save resources to use
2565        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2566    
2567     4. The code for checking the validity of UTF-8 strings has been tightened so
2568        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2569        containing "overlong sequences".
2570    
2571     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2572        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2573        should have been "&", but it just so happened that all the cases this let
2574        through by mistake were picked up later in the function.
2575    
2576     6. I had used a variable called "isblank" - this is a C99 function, causing
2577        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2578    
2579     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2580        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2581        so the version is shown; (c) stop "make test" echoing "./RunTest".
2582    
2583     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2584    
2585     9. The prototype for memmove() for systems that don't have it was using
2586        size_t, but the inclusion of the header that defines size_t was later. I've
2587        moved the #includes for the C headers earlier to avoid this.
2588    
2589    10. Added some adjustments to the code to make it easier to compiler on certain
2590        special systems:
2591    
2592          (a) Some "const" qualifiers were missing.
2593          (b) Added the macro EXPORT before all exported functions; by default this
2594              is defined to be empty.
2595          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2596              that it reads its output file name as an argument instead of writing
2597              to the standard output and assuming this can be redirected.
2598    
2599    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2600        class containing characters with values greater than 255, PCRE compilation
2601        went into a loop.
2602    
2603    12. A recursive reference to a subpattern that was within another subpattern
2604        that had a minimum quantifier of zero caused PCRE to crash. For example,
2605        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2606        recursion. If the recursively-called subpattern itself had a zero repeat,
2607        that was OK.
2608    
2609    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2610        buffer into which it was copied (for escape processing) was still set at
2611        1024, so long lines caused crashes.
2612    
2613    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2614        "internal error: code overflow...". This applied to any character class
2615        that was followed by a possessive quantifier.
2616    
2617    15. Modified the Makefile to add libpcre.la as a prerequisite for
2618        libpcreposix.la because I was told this is needed for a parallel build to
2619        work.
2620    
2621    16. If a pattern that contained .* following optional items at the start was
2622        studied, the wrong optimizing data was generated, leading to matching
2623        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2624        matching string must start with a or b or c. The correct conclusion for
2625        this pattern is that a match can start with any character.
2626    
2627    
2628    Version 4.4 13-Aug-03
2629    ---------------------
2630    
2631     1. In UTF-8 mode, a character class containing characters with values between
2632        127 and 255 was not handled correctly if the compiled pattern was studied.
2633        In fixing this, I have also improved the studying algorithm for such
2634        classes (slightly).
2635    
2636     2. Three internal functions had redundant arguments passed to them. Removal
2637        might give a very teeny performance improvement.
2638    
2639     3. Documentation bug: the value of the capture_top field in a callout is *one
2640        more than* the number of the hightest numbered captured substring.
2641    
2642     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2643        in incorrectly linking with a previously installed version. They now link
2644        explicitly with libpcre.la.
2645    
2646     5. configure.in no longer needs to recognize Cygwin specially.
2647    
2648     6. A problem in pcre.in for Windows platforms is fixed.
2649    
2650     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2651        pcretest, it used to include the size of the study block as part of its
2652        output. Unfortunately, the structure contains a field that has a different
2653        size on different hardware architectures. This meant that the tests that
2654        showed this size failed. As the block is currently always of a fixed size,
2655        this information isn't actually particularly useful in pcretest output, so
2656        I have just removed it.
2657    
2658     8. Three pre-processor statements accidentally did not start in column 1.
2659        Sadly, there are *still* compilers around that complain, even though
2660        standard C has not required this for well over a decade. Sigh.
2661    
2662     9. In pcretest, the code for checking callouts passed small integers in the
2663        callout_data field, which is a void * field. However, some picky compilers
2664        complained about the casts involved for this on 64-bit systems. Now
2665        pcretest passes the address of the small integer instead, which should get
2666        rid of the warnings.
2667    
2668    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2669        both compile and run time, and gives an error if an invalid UTF-8 sequence
2670        is found. There is a option for disabling this check in cases where the
2671        string is known to be correct and/or the maximum performance is wanted.
2672    
2673    11. In response to a bug report, I changed one line in Makefile.in from
2674    
2675            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2676        to
2677            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2678    
2679        to look similar to other lines, but I have no way of telling whether this
2680        is the right thing to do, as I do not use Windows. No doubt I'll get told
2681        if it's wrong...
2682    
2683    
2684    Version 4.3 21-May-03
2685    ---------------------
2686    
2687    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2688       Makefile.
2689    
2690    2. Some refactoring to improve the quality of the code:
2691    
2692       (i)   The utf8_table... variables are now declared "const".
2693    
2694       (ii)  The code for \cx, which used the "case flipping" table to upper case
2695             lower case letters, now just substracts 32. This is ASCII-specific,
2696             but the whole concept of \cx is ASCII-specific, so it seems
2697             reasonable.
2698    
2699       (iii) PCRE was using its character types table to recognize decimal and
2700             hexadecimal digits in the pattern. This is silly, because it handles
2701             only 0-9, a-f, and A-F, but the character types table is locale-
2702             specific, which means strange things might happen. A private
2703             table is now used for this - though it costs 256 bytes, a table is
2704             much faster than multiple explicit tests. Of course, the standard
2705             character types table is still used for matching digits in subject
2706             strings against \d.
2707    
2708       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2709             ending in _t are). So I've renamed it as ESC_tee.
2710    
2711    3. The first argument for regexec() in the POSIX wrapper should have been
2712       defined as "const".
2713    
2714    4. Changed pcretest to use malloc() for its buffers so that they can be
2715       Electric Fenced for debugging.
2716    
2717    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2718       to read one or more bytes before the start of the subject string. Often this
2719       had no effect on PCRE's behaviour, but in some circumstances it could
2720       provoke a segmentation fault.
2721    
2722    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2723       to try to read one or more bytes before the start of the subject string.
2724    
2725    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2726       UTF-8 support could misbehave in various ways if the subject string
2727       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2728       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2729       back over UTF-8 characters.)
2730    
2731    
2732    Version 4.2 14-Apr-03
2733    ---------------------
2734    
2735    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2736    
2737    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2738         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2739         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2740         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2741         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2742           and BUILD_EXEEXT
2743         Note: automatic setting of the BUILD variables is not yet working
2744         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2745           compile-time but not at link-time
2746         [LINK]: use for linking executables only
2747         make different versions for Windows and non-Windows
2748         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2749           libraries
2750         [LINK_FOR_BUILD]: new variable
2751         [OBJEXT]: use throughout
2752         [EXEEXT]: use throughout
2753         <winshared>: new target
2754         <wininstall>: new target
2755         <dftables.o>: use native compiler
2756         <dftables>: use native linker
2757         <install>: handle Windows platform correctly
2758         <clean>: ditto
2759         <check>: ditto
2760         copy DLL to top builddir before testing
2761    
2762       As part of these changes, -no-undefined was removed again. This was reported
2763       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2764       in any case.
2765    
2766    3. Some tidies to get rid of compiler warnings:
2767    
2768       . In the match_data structure, match_limit was an unsigned long int, whereas
2769         match_call_count was an int. I've made them both unsigned long ints.
2770    
2771       . In pcretest the fact that a const uschar * doesn't automatically cast to
2772         a void * provoked a warning.
2773    
2774       . Turning on some more compiler warnings threw up some "shadow" variables
2775         and a few more missing casts.
2776    
2777    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2778       option, a class that contained a single character with a value between 128
2779       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2780    
2781    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2782       option, a class that contained several characters, but with at least one
2783       whose value was between 128 and 255 caused PCRE to crash.
2784    
2785    
2786    Version 4.1 12-Mar-03
2787    ---------------------
2788    
2789    1. Compiling with gcc -pedantic found a couple of places where casts were
2790    needed, and a string in dftables.c that was longer than standard compilers are
2791    required to support.
2792    
2793    2. Compiling with Sun's compiler found a few more places where the code could
2794    be tidied up in order to avoid warnings.
2795    
2796    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2797    first of these names is deprecated in the latest Autoconf in favour of the name
2798    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2799    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2800    analogy I have changed it to CFLAGS_FOR_BUILD.
2801    
2802    4. Added -no-undefined to the linking command in the Makefile, because this is
2803    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2804    linking step for the pcreposix library.
2805    
2806    5. PCRE was failing to diagnose the case of two named groups with the same
2807    name.
2808    
2809    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2810    literal character that is needed in the subject for a match, and scans along to
2811    ensure that it is present before embarking on the full matching process. This
2812    saves time in cases of nested unlimited repeats that are never going to match.
2813    Problem: the scan can take a lot of time if the subject is very long (e.g.
2814    megabytes), thus penalizing straightforward matches. It is now done only if the
2815    amount of subject to be scanned is less than 1000 bytes.
2816    
2817    7. A lesser problem with the same optimization is that it was recording the
2818    first character of an anchored pattern as "needed", thus provoking a search
2819    right along the subject, even when the first match of the pattern was going to
2820    fail. The "needed" character is now not set for anchored patterns, unless it
2821    follows something in the pattern that is of non-fixed length. Thus, it still
2822    fulfils its original purpose of finding quick non-matches in cases of nested
2823    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2824    
2825    
2826    Version 4.0 17-Feb-03
2827    ---------------------
2828    
2829    1. If a comment in an extended regex that started immediately after a meta-item
2830    extended to the end of string, PCRE compiled incorrect data. This could lead to
2831    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2832    
2833    2. Moved to autoconf 2.53 and libtool 1.4.2.
2834    
2835    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2836    the special perltest8 script is no longer needed - all the tests can be run
2837    from a single perltest script.
2838    
2839    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2840    by \s. It has now been removed in PCRE. This means it isn't recognized as
2841    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2842    class [:space:] *does* include VT, thereby creating a mess.
2843    
2844    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2845    space and tab.
2846    
2847    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2848    its new features into the main test script, reducing the number of scripts.
2849    
2850    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2851    were backward compatible, and made the (?i) apply to the whole pattern, as if
2852    /i were given. Now it behaves more logically, and applies the option setting
2853    only to what follows. PCRE has been changed to follow suit. However, if it
2854    finds options settings right at the start of the pattern, it extracts them into
2855    the global options, as before. Thus, they show up in the info data.
2856    
2857    8. Added support for the \Q...\E escape sequence. Characters in between are
2858    treated as literals. This is slightly different from Perl in that $ and @ are
2859    also handled as literals inside the quotes. In Perl, they will cause variable
2860    interpolation. Note the following examples:
2861    
2862        Pattern            PCRE matches      Perl matches
2863    
2864        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2865        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2866        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2867    
2868    For compatibility with Perl, \Q...\E sequences are recognized inside character
2869    classes as well as outside them.
2870    
2871    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2872    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2873    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2874    signed/unsigned warnings.
2875    
2876    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2877    option for pcretest, so I've replaced it by a simple function that does just
2878    that job.
2879    
2880    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2881    "pcregrep -".
2882    
2883    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2884    Java package. This provides some syntactic sugar for simple cases of what my
2885    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2886    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2887    item, you can use this simplified notation. Note that only makes sense with
2888    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2889    greediness, whatever the setting of the PCRE_UNGREEDY option.
2890    
2891    13. A change of greediness default within a pattern was not taking effect at
2892    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2893    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2894    was abstracted outside.
2895    
2896    14. PCRE now supports the \G assertion. It is true when the current matching
2897    position is at the start point of the match. This differs from \A when the
2898    starting offset is non-zero. Used with the /g option of pcretest (or similar
2899    code), it works in the same way as it does for Perl's /g option. If all
2900    alternatives of a regex begin with \G, the expression is anchored to the start
2901    match position, and the "anchored" flag is set in the compiled expression.
2902    
2903    15. Some bugs concerning the handling of certain option changes within patterns
2904    have been fixed. These applied to options other than (?ims). For example,
2905    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2906    been the other way round. Some of this was related to change 7 above.
2907    
2908    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2909    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2910    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2911    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2912    
2913    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2914    mode. Unlike ".", it always matches newline, whatever the setting of
2915    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2916    assertions. Perl allows it, but it doesn't (in general) work because it can't
2917    calculate the length of the lookbehind. At least, that's the case for Perl
2918    5.8.0 - I've been told they are going to document that it doesn't work in
2919    future.
2920    
2921    18. Added an error diagnosis for escapes that PCRE does not support: these are
2922    \L, \l, \N, \P, \p, \U, \u, and \X.
2923    
2924    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2925    reading past the end of the pattern in cases such as /[abcd/.
2926    
2927    20. PCRE was getting more memory than necessary for patterns with classes that
2928    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2929    
2930    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2931    compiling PCRE for use with Virtual Pascal.
2932    
2933    22. Small fix to the Makefile to make it work properly if the build is done
2934    outside the source tree.
2935    
2936    23. Added a new extension: a condition to go with recursion. If a conditional
2937    subpattern starts with (?(R) the "true" branch is used if recursion has
2938    happened, whereas the "false" branch is used only at the top level.
2939    
2940    24. When there was a very long string of literal characters (over 255 bytes
2941    without UTF support, over 250 bytes with UTF support), the computation of how
2942    much memory was required could be incorrect, leading to segfaults or other
2943    strange effects.
2944    
2945    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2946    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2947    there was a subsequent back reference to those brackets. This meant that, for
2948    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2949    possible to check for precisely this case. All we can do is abandon the
2950    optimization if .* occurs inside capturing brackets when there are any back
2951    references whatsoever. (See below for a better fix that came later.)
2952    
2953    26. The handling of the optimization for finding the first character of a
2954    non-anchored pattern, and for finding a character that is required later in the
2955    match were failing in some cases. This didn't break the matching; it just
2956    failed to optimize when it could. The way this is done has been re-implemented.
2957    
2958    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2959    
2960    28. Added a new feature that provides some of the functionality that Perl
2961    provides with (?{...}). The facility is termed a "callout". The way it is done
2962    in PCRE is for the caller to provide an optional function, by setting
2963    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2964    global variable. By default it is unset, which disables all calling out. To get
2965    the function called, the regex must include (?C) at appropriate points. This
2966    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2967    This provides a means of identifying different callout points. When PCRE
2968    reaches such a point in the regex, if pcre_callout has been set, the external
2969    function is called. It is provided with data in a structure called
2970    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2971    matching continues; if it returns a non-zero value, the match at the current
2972    point fails. However, backtracking will occur if possible. [This was changed
2973    later and other features added - see item 49 below.]
2974    
2975    29. pcretest is upgraded to test the callout functionality. It provides a
2976    callout function that displays information. By default, it shows the start of
2977    the match and the current position in the text. There are some new data escapes
2978    to vary what happens:
2979    
2980        \C+         in addition, show current contents of captured substrings
2981        \C-         do not supply a callout function
2982        \C!n        return 1 when callout number n is reached
2983        \C!n!m      return 1 when callout number n is reached for the mth time
2984    
2985    30. If pcregrep was called with the -l option and just a single file name, it
2986    output "<stdin>" if a match was found, instead of the file name.
2987    
2988    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2989    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2990    pcre_exec(). This saves a malloc/free per call. The default value of
2991    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2992    when configuring.
2993    
2994    32. The default maximum size of a compiled pattern is 64K. There have been a
2995    few cases of people hitting this limit. The code now uses macros to handle the
2996    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2997    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2998    configuring. Tests 2 and 5 work only with 2-byte links because they output
2999    debugging information about compiled patterns.
3000    
3001    33. Internal code re-arrangements:
3002    
3003    (a) Moved the debugging function for printing out a compiled regex into
3004        its own source file (printint.c) and used #include to pull it into
3005        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3006        separate copies.
3007    
3008    (b) Defined the list of op-code names for debugging as a macro in
3009        internal.h so that it is next to the definition of the opcodes.
3010    
3011    (c) Defined a table of op-code lengths for simpler skipping along compiled
3012        code. This is again a macro in internal.h so that it is next to the
3013        definition of the opcodes.
3014    
3015    34. Added support for recursive calls to individual subpatterns, along the
3016    lines of Robin Houston's patch (but implemented somewhat differently).
3017    
3018    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3019    allow it to read and process whole directories in Win32. This code was
3020    contributed by Lionel Fourquaux; it has not been tested by me.
3021    
3022    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3023    used to name a group. Names consist of alphanumerics and underscores, and must
3024    be unique. Back references use the syntax (?P=name) and recursive calls use
3025    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3026    numbers. The function pcre_fullinfo() can be used after compilation to extract
3027    a name/number map. There are three relevant calls:
3028    
3029      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3030      PCRE_INFO_NAMECOUNT            yields the number of entries
3031      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3032    
3033    The map is a vector of fixed-size entries. The size of each entry depends on
3034    the length of the longest name used. The first two bytes of each entry are the
3035    group number, most significant byte first. There follows the corresponding
3036    name, zero terminated. The names are in alphabetical order.
3037    
3038    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3039    case instead of 255. Making it the same both with and without UTF-8 support
3040    means that the same test output works with both.
3041    
3042    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3043    calling malloc() with a zero argument.
3044    
3045    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3046    optimization. I've improved things by keeping a bitmap of backreferences with
3047    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3048    fact referenced, the optimization can be applied. It is unlikely that a
3049    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3050    the match to follow \n) will appear inside brackets with a number greater than
3051    31, but if it does, any back reference > 31 suppresses the optimization.
3052    
3053    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3054    of disabling numbered capturing parentheses. Any opening parenthesis that is
3055    not followed by ? behaves as if it were followed by ?: but named parentheses
3056    can still be used for capturing (and they will acquire numbers in the usual
3057    way).
3058    
3059    41. Redesigned the return codes from the match() function into yes/no/error so
3060    that errors can be passed back from deep inside the nested calls. A malloc
3061    failure while inside a recursive subpattern call now causes the
3062    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3063    
3064    42. It is now possible to set a limit on the number of times the match()
3065    function is called in a call to pcre_exec(). This facility makes it possible to
3066    limit the amount of recursion and backtracking, though not in a directly
3067    obvious way, because the match() function is used in a number of different
3068    circumstances. The count starts from zero for each position in the subject
3069    string (for non-anchored patterns). The default limit is, for compatibility, a
3070    large number, namely 10 000 000. You can change this in two ways:
3071    
3072    (a) When configuring PCRE before making, you can use --with-match-limit=n
3073        to set a default value for the compiled library.
3074    
3075    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3076        a different value is set. See 45 below.
3077    
3078    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3079    
3080    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3081    of things that can be changed at compile time. The first argument specifies
3082    what is wanted and the second points to where the information is to be placed.
3083    The current list of available information is:
3084    
3085      PCRE_CONFIG_UTF8
3086    
3087    The output is an integer that is set to one if UTF-8 support is available;
3088    otherwise it is set to zero.
3089    
3090      PCRE_CONFIG_NEWLINE
3091    
3092    The output is an integer that it set to the value of the code that is used for
3093    newline. It is either LF (10) or CR (13).
3094    
3095      PCRE_CONFIG_LINK_SIZE
3096    
3097    The output is an integer that contains the number of bytes used for internal
3098    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3099    
3100      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3101    
3102    The output is an integer that contains the threshold above which the POSIX
3103    interface uses malloc() for output vectors. See item 31 above.
3104    
3105      PCRE_CONFIG_MATCH_LIMIT
3106    
3107    The output is an unsigned integer that contains the default limit of the number
3108    of match() calls in a pcre_exec() execution. See 42 above.
3109    
3110    44. pcretest has been upgraded by the addition of the -C option. This causes it
3111    to extract all the available output from the new pcre_config() function, and to
3112    output it. The program then exits immediately.
3113    
3114    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3115    order to support additional features. One way would have been to define
3116    pcre_exec2() (for example) with extra arguments, but this would not have been
3117    extensible, and would also have required all calls to the original function to
3118    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3119    is used for passing in "extra" data from pcre_study().
3120    
3121    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3122    contains the following fields:
3123    
3124      flags         a bitmap indicating which of the following fields are set
3125      study_data    opaque data from pcre_study()
3126      match_limit   a way of specifying a limit on match() calls for a specific
3127                      call to pcre_exec()
3128      callout_data  data for callouts (see 49 below)
3129    
3130    The flag bits are also defined in pcre.h, and are
3131    
3132      PCRE_EXTRA_STUDY_DATA
3133      PCRE_EXTRA_MATCH_LIMIT
3134      PCRE_EXTRA_CALLOUT_DATA
3135    
3136    The pcre_study() function now returns one of these new pcre_extra blocks, with
3137    the actual study data pointed to by the study_data field, and the
3138    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3139    before. That is, this change is entirely upwards-compatible and requires no
3140    change to existing code.
3141    
3142    If you want to pass in additional data to pcre_exec(), you can either place it
3143    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3144    block.
3145    
3146    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3147    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3148    times with different match limits, until it finds the minimum value needed for
3149    pcre_exec() to complete. The value is then output. This can be instructive; for
3150    most simple matches the number is quite small, but for pathological cases it
3151    gets very large very quickly.
3152    
3153    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3154    returns the size of the data block pointed to by the study_data field in a
3155    pcre_extra block, that is, the value that was passed as the argument to
3156    pcre_malloc() when PCRE was getting memory in which to place the information
3157    created by pcre_study(). The fourth argument should point to a size_t variable.
3158    pcretest has been extended so that this information is shown after a successful
3159    pcre_study() call when information about the compiled regex is being displayed.
3160    
3161    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3162    because what follows is always an absolute path. (Later: it turns out that this
3163    is more than cosmetic for MinGW, because it doesn't like empty path
3164    components.)
3165    
3166    49. Some changes have been made to the callout feature (see 28 above):
3167    
3168    (i)  A callout function now has three choices for what it returns:
3169    
3170           0  =>  success, carry on matching
3171         > 0  =>  failure at this point, but backtrack if possible
3172         < 0  =>  serious error, return this value from pcre_exec()
3173    
3174         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3175         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3176         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3177         use by callout functions. It will never be used by PCRE itself.
3178    
3179    (ii) The pcre_extra structure (see 45 above) has a void * field called
3180         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3181         pcre_callout_block structure has a field of the same name. The contents of
3182         the field passed in the pcre_extra structure are passed to the callout
3183         function in the corresponding field in the callout block. This makes it
3184         easier to use the same callout-containing regex from multiple threads. For
3185         testing, the pcretest program has a new data escape
3186    
3187           \C*n        pass the number n (may be negative) as callout_data
3188    
3189         If the callout function in pcretest receives a non-zero value as
3190         callout_data, it returns that value.
3191    
3192    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3193    there were some redundant $(CFLAGS) in commands that are now specified as
3194    $(LINK), which already includes $(CFLAGS).
3195    
3196    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3197    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3198    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3199    one-byte characters throughout. Note that case-insensitive matching applies
3200    only to characters whose values are less than 256. PCRE doesn't support the
3201    notion of cases for higher-valued characters.
3202    
3203    (i)   A character class whose characters are all within 0-255 is handled as
3204          a bit map, and the map is inverted for negative classes. Previously, a
3205          character > 255 always failed to match such a class; however it should
3206          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3207    
3208    (ii)  A negated character class with a single character < 255 is coded as
3209          "not this character" (OP_NOT). This wasn't working properly when the test
3210          character was multibyte, either singly or repeated.
3211    
3212    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3213          mode, for example: \x{100}{2,3}.
3214    
3215    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3216          singly or repeated) now correctly test multibyte characters. However,
3217          PCRE doesn't recognize any characters with values greater than 255 as
3218          digits, spaces, or word characters. Such characters always match \D, \S,
3219          and \W, and never match \d, \s, or \w.
3220    
3221    (v)   Classes may now contain characters and character ranges with values
3222          greater than 255. For example: [ab\x{100}-\x{400}].
3223    
3224    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3225          PCRE in UTF-8 mode.
3226    
3227    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3228    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3229    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3230    value.)
3231    
3232    53. The single man page has become too large. I have therefore split it up into
3233    a number of separate man pages. These also give rise to individual HTML pages;
3234    these are now put in a separate directory, and there is an index.html page that
3235    lists them all. Some hyperlinking between the pages has been installed.
3236    
3237    54. Added convenience functions for handling named capturing parentheses.
3238    
3239    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3240    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3241    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3242    are faulted.
3243    
3244    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3245    calling configure. These values are used when compiling the dftables.c program
3246    which is run to generate the source of the default character tables. They
3247    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3248    you will need to set these values.
3249    
3250    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3251    
3252    
3253    Version 3.9 02-Jan-02
3254    ---------------------
3255    
3256    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3257    
3258    2. If --disable-static was given, the building process failed when trying to
3259    build pcretest and pcregrep. (For some reason it was using libtool to compile
3260    them, which is not right, as they aren't part of the library.)
3261    
3262    
3263    Version 3.8 18-Dec-01
3264    ---------------------
3265    
3266    1. The experimental UTF-8 code was completely screwed up. It was packing the
3267    bytes in the wrong order. How dumb can you get?
3268    
3269    
3270    Version 3.7 29-Oct-01
3271    ---------------------
3272    
3273    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3274    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3275    this didn't happen under Solaris 8, where I normally test things.
3276    
3277    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3278    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3279    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3280    this fix an hour or so after the initial 3.7 release.)
3281    
3282    
3283    Version 3.6 23-Oct-01
3284    ---------------------
3285    
3286    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3287    offsets passed as NULL with zero offset count.
3288    
3289    2. The config.guess and config.sub files had not been updated when I moved to
3290    the latest autoconf.
3291    
3292    
3293    Version 3.5 15-Aug-01
3294    ---------------------
3295    
3296    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3297    had been forgotten.
3298    
3299    2. By using declared but undefined structures, we can avoid using "void"
3300    definitions in pcre.h while keeping the internal definitions of the structures
3301    private.
3302    
3303    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3304    user point of view, this means that both static and shared libraries are built
3305    by default, but this can be individually controlled. More of the work of
3306    handling this static/shared cases is now inside libtool instead of PCRE's make
3307    file.
3308    
3309    4. The pcretest utility is now installed along with pcregrep because it is
3310    useful for users (to test regexs) and by doing this, it automatically gets
3311    relinked by libtool. The documentation has been turned into a man page, so
3312    there are now .1, .txt, and .html versions in /doc.
3313    
3314    5. Upgrades to pcregrep:
3315       (i)   Added long-form option names like gnu grep.
3316       (ii)  Added --help to list all options with an explanatory phrase.
3317       (iii) Added -r, --recursive to recurse into sub-directories.
3318       (iv)  Added -f, --file to read patterns from a file.
3319    
3320    6. pcre_exec() was referring to its "code" argument before testing that
3321    argument for NULL (and giving an error if it was NULL).
3322    
3323    7. Upgraded Makefile.in to allow for compiling in a different directory from
3324    the source directory.
3325    
3326    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3327    options bits, the pointer it was passed was to an int instead of to an unsigned
3328    long int. This mattered only on 64-bit systems.
3329    
3330    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3331    generated) instead of pcre.in, which it its source. Also made the same change
3332    in several of the .c files.
3333    
3334    10. A new release of gcc defines printf() as a macro, which broke pcretest
3335    because it had an ifdef in the middle of a string argument for printf(). Fixed
3336    by using separate calls to printf().
3337    
3338    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3339    script, to force use of CR or LF instead of \n in the source. On non-Unix
3340    systems, the value can be set in config.h.
3341    
3342    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3343    absolute limit. Changed the text of the error message to make this clear, and
3344    likewise updated the man page.
3345    
3346    13. The limit of 99 on the number of capturing subpatterns has been removed.
3347    The new limit is 65535, which I hope will not be a "real" limit.
3348    
3349    
3350    Version 3.4 22-Aug-00
3351    ---------------------
3352    
3353    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3354    
3355    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3356    
3357    
3358    Version 3.3 01-Aug-00
3359    ---------------------
3360    
3361    1. If an octal character was given, but the value was greater than \377, it
3362    was not getting masked to the least significant bits, as documented. This could
3363    lead to crashes in some systems.
3364    
3365    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3366    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3367    
3368    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3369    These just pass their arguments on to (pcre_free)(), but they are provided
3370    because some uses of PCRE bind it to non-C systems that can call its functions,
3371    but cannot call free() or pcre_free() directly.
3372    
3373    4. Add "make test" as a synonym for "make check". Corrected some comments in
3374    the Makefile.
3375    
3376    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3377    Makefile.
3378    
3379    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3380    command called pgrep for grepping around the active processes.
3381    
3382    7. Added the beginnings of support for UTF-8 character strings.
3383    
3384    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3385    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3386    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3387    out for the ar command.)
3388    
3389    
3390    Version 3.2 12-May-00
3391    ---------------------
3392    
3393    This is purely a bug fixing release.
3394    
3395    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3396    of ZA. This was just one example of several cases that could provoke this bug,
3397    which was introduced by change 9 of version 2.00. The code for breaking
3398    infinite loops after an iteration that matches an empty string was't working
3399    correctly.
3400    
3401    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3402    when matched against abbab (for example). After matching an empty string, it
3403    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3404    caused it to match further down the string than it should.
3405    
3406    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3407    was there because it doesn't seem to be needed, and it causes trouble on some
3408    systems, as it is not a Standard C header. It has been removed.
3409    
3410    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3411    were reported on the Macintosh. The changes were from
3412    
3413      while ((c = *(++ptr)) != 0 && c != '\n');
3414    to
3415      while ((c = *(++ptr)) != 0 && c != '\n') ;
3416    
3417    Totally extraordinary, but if that's what it takes...
3418    
3419    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3420    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3421    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3422    assumes the way PCRE uses memmove() (always moving upwards).
3423    
3424    6. PCRE is being used in one environment where strchr() is not available. There
3425    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3426    faster code anyway.
3427    
3428    
3429    Version 3.1 09-Feb-00
3430    ---------------------
3431    
3432    The only change in this release is the fixing of some bugs in Makefile.in for
3433    the "install" target:
3434    
3435    (1) It was failing to install pcreposix.h.
3436    
3437    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3438    
3439    
3440  Version 3.0 01-Feb-00  Version 3.0 01-Feb-00
3441  ---------------------  ---------------------

Legend:
Removed from v.43  
changed lines
  Added in v.553

  ViewVC Help
Powered by ViewVC 1.1.5