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

Diff of /code/trunk/ChangeLog

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

revision 23 by nigel, Sat Feb 24 21:38:41 2007 UTC revision 539 by ph10, Sun Jun 13 21:35:04 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.10 03-Jun-2010
5    ------------------------
6    
7    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8        THEN.
9    
10    2.  (*ACCEPT) was not working when inside an atomic group.
11    
12    3.  Inside a character class, \B is treated as a literal by default, but
13        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14        causes the error). The code is unchanged, but I tidied the documentation.
15    
16    4.  Inside a character class, PCRE always treated \R and \X as literals,
17        whereas Perl faults them if its -w option is set. I have changed PCRE so
18        that it faults them when PCRE_EXTRA is set.
19    
20    5.  Added support for \N, which always matches any character other than
21        newline. (It is the same as "." when PCRE_DOTALL is not set.)
22    
23    6.  When compiling pcregrep with newer versions of gcc which may have
24        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25        declared with attribute warn_unused_result" were given. Just casting the
26        result to (void) does not stop the warnings; a more elaborate fudge is
27        needed. I've used a macro to implement this.
28    
29    7.  Minor change to pcretest.c to avoid a compiler warning.
30    
31    8.  Added four artifical Unicode properties to help with an option to make
32        \s etc use properties (see next item). The new properties are: Xan
33        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34    
35    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36        use Unicode properties. (*UCP) at the start of a pattern can be used to set
37        this option. Modified pcretest to add /W to test this facility. Added
38        REG_UCP to make it available via the POSIX interface.
39    
40    10. Added --line-buffered to pcregrep.
41    
42    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
43        studied, and the match started with a letter with a code point greater than
44        127 whose first byte was different to the first byte of the other case of
45        the letter, the other case of this starting letter was not recognized
46        (#976).
47    
48    12. If a pattern that was studied started with a repeated Unicode property
49        test, for example, \p{Nd}+, there was the theoretical possibility of
50        setting up an incorrect bitmap of starting bytes, but fortunately it could
51        not have actually happened in practice until change 8 above was made (it
52        added property types that matched character-matching opcodes).
53    
54    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
55        possible starting bytes for non-anchored patterns.
56    
57    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
58        \R, and also a number of cases that involve Unicode properties, both
59        explicit and implicit when PCRE_UCP is set.
60    
61    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
62        input, it could crash or give wrong results if characters with values
63        greater than 0xc0 were present in the subject string. (Detail: it assumed
64        UTF-8 input when processing these items.)
65    
66    16. Added a lot of (int) casts to avoid compiler warnings in systems where
67        size_t is 64-bit (#991).
68    
69    17. Added a check for running out of memory when PCRE is compiled with
70        --disable-stack-for-recursion (#990).
71    
72    18. If the last data line in a file for pcretest does not have a newline on
73        the end, a newline was missing in the output.
74    
75    19. The default pcre_chartables.c file recognizes only ASCII characters (values
76        less than 128) in its various bitmaps. However, there is a facility for
77        generating tables according to the current locale when PCRE is compiled. It
78        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
79        characters, are recognized by isspace() and therefore were getting set in
80        these tables. This caused a problem in UTF-8 mode when pcre_study() was
81        used to create a list of bytes that can start a match. For \s, it was
82        including 0x85 and 0xa0, which of course cannot start UTF-8 characters. I
83        have changed the code so that only real ASCII characters (less than 128)
84        and the correct starting bytes for UTF-8 encodings are set in this case.
85        (When PCRE_UCP is set - see 9 above - the code is different altogether.)
86    
87    
88    Version 8.02 19-Mar-2010
89    ------------------------
90    
91    1.  The Unicode data tables have been updated to Unicode 5.2.0.
92    
93    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
94        configured.
95    
96    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
97        original author of that file, following a query about its status.
98    
99    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
100        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
101    
102    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
103        quantifier applied to a forward-referencing subroutine call, could compile
104        incorrect code or give the error "internal error: previously-checked
105        referenced subpattern not found".
106    
107    6.  Both MS Visual Studio and Symbian OS have problems with initializing
108        variables to point to external functions. For these systems, therefore,
109        pcre_malloc etc. are now initialized to local functions that call the
110        relevant global functions.
111    
112    7.  There were two entries missing in the vectors called coptable and poptable
113        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
114        I've fixed the data, and added a kludgy way of testing at compile time that
115        the lengths are correct (equal to the number of opcodes).
116    
117    8.  Following on from 7, I added a similar kludge to check the length of the
118        eint vector in pcreposix.c.
119    
120    9.  Error texts for pcre_compile() are held as one long string to avoid too
121        much relocation at load time. To find a text, the string is searched,
122        counting zeros. There was no check for running off the end of the string,
123        which could happen if a new error number was added without updating the
124        string.
125    
126    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
127    
128    11. \K was not working if it appeared in an atomic group or in a group that
129        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
130        \K is "not well defined" if used in an assertion. PCRE now accepts it if
131        the assertion is positive, but not if it is negative.
132    
133    12. Change 11 fortuitously reduced the size of the stack frame used in the
134        "match()" function of pcre_exec.c by one pointer. Forthcoming
135        implementation of support for (*MARK) will need an extra pointer on the
136        stack; I have reserved it now, so that the stack frame size does not
137        decrease.
138    
139    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
140        item in branch that calls a recursion is a subroutine call - as in the
141        second branch in the above example - was incorrectly given the compile-
142        time error "recursive call could loop indefinitely" because pcre_compile()
143        was not correctly checking the subroutine for matching a non-empty string.
144    
145    14. The checks for overrunning compiling workspace could trigger after an
146        overrun had occurred. This is a "should never occur" error, but it can be
147        triggered by pathological patterns such as hundreds of nested parentheses.
148        The checks now trigger 100 bytes before the end of the workspace.
149    
150    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
151    
152    
153    Version 8.01 19-Jan-2010
154    ------------------------
155    
156    1.  If a pattern contained a conditional subpattern with only one branch (in
157        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
158        computed the wrong minimum data length (which is of course zero for such
159        subpatterns). This could cause incorrect "no match" results.
160    
161    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
162        the pattern is reset in the first branch, pcre_compile() failed with
163        "internal error: code overflow at offset...". This happened only when
164        the reset was to the original external option setting. (An optimization
165        abstracts leading options settings into an external setting, which was the
166        cause of this.)
167    
168    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
169        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
170        assertion pattern did not match (meaning that the assertion was true), it
171        was incorrectly treated as false if the SKIP had been reached during the
172        matching. This also applied to assertions used as conditions.
173    
174    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
175        assertion subpattern, including such a pattern used as a condition,
176        unpredictable results occurred, instead of the error return
177        PCRE_ERROR_DFA_UITEM.
178    
179    5.  The C++ GlobalReplace function was not working like Perl for the special
180        situation when an empty string is matched. It now does the fancy magic
181        stuff that is necessary.
182    
183    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
184        removed. (These were left over from very, very early versions of PCRE.)
185    
186    7.  Some cosmetic changes to the code to make life easier when compiling it
187        as part of something else:
188    
189        (a) Change DEBUG to PCRE_DEBUG.
190    
191        (b) In pcre_compile(), rename the member of the "branch_chain" structure
192            called "current" as "current_branch", to prevent a collision with the
193            Linux macro when compiled as a kernel module.
194    
195        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
196            prevent a collision with the Linux macro when compiled as a kernel
197            module.
198    
199    8.  In pcre_compile() there are some checks for integer overflows that used to
200        cast potentially large values to (double). This has been changed to that
201        when building, a check for int64_t is made, and if it is found, it is used
202        instead, thus avoiding the use of floating point arithmetic. (There is no
203        other use of FP in PCRE.) If int64_t is not found, the fallback is to
204        double.
205    
206    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
207        2005 (difference between two addresses compared to an unsigned value).
208    
209    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
210        custom one, because of the following reported problem in Windows:
211    
212          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
213              under Win32.
214          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
215              therefore missing the function definition.
216          - The compiler thus generates a "C" signature for the test function.
217          - The linker fails to find the "C" function.
218          - PCRE fails to configure if asked to do so against libbz2.
219    
220    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
221        messages were output:
222    
223          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
224          rerunning libtoolize, to keep the correct libtool macros in-tree.
225          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
226    
227        I have done both of these things.
228    
229    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
230        most of the time, it *can* run out if it is given a pattern that contains a
231        runaway infinite recursion. I updated the discussion in the pcrestack man
232        page.
233    
234    13. Now that we have gone to the x.xx style of version numbers, the minor
235        version may start with zero. Using 08 or 09 is a bad idea because users
236        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
237        interpreted as invalid octal numbers. I've updated the previous comment in
238        configure.ac, and also added a check that gives an error if 08 or 09 are
239        used.
240    
241    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
242        causing partial matching to fail when the end of the subject matched \W
243        in a UTF-8 pattern where \W was quantified with a minimum of 3.
244    
245    15. There were some discrepancies between the declarations in pcre_internal.h
246        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
247        their definitions. The declarations used "const uschar *" and the
248        definitions used USPTR. Even though USPTR is normally defined as "const
249        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
250        reported that: "This difference in casting confuses some C++ compilers, for
251        example, SunCC recognizes above declarations as different functions and
252        generates broken code for hbpcre." I have changed the declarations to use
253        USPTR.
254    
255    16. GNU libtool is named differently on some systems. The autogen.sh script now
256        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
257        (FreeBSD).
258    
259    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
260        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
261        comment: "Figure out how to create a longlong from a string: strtoll and
262        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
263        instance, but it only takes 2 args instead of 3!"
264    
265    18. A subtle bug concerned with back references has been fixed by a change of
266        specification, with a corresponding code fix. A pattern such as
267        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
268        refers, was giving matches when it shouldn't. For example, xa=xaaa would
269        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
270        same bug. Such groups have to be quantified to be useful, or contained
271        inside another quantified group. (If there's no repetition, the reference
272        can never match.) The problem arises because, having left the group and
273        moved on to the rest of the pattern, a later failure that backtracks into
274        the group uses the captured value from the final iteration of the group
275        rather than the correct earlier one. I have fixed this in PCRE by forcing
276        any group that contains a reference to itself to be an atomic group; that
277        is, there cannot be any backtracking into it once it has completed. This is
278        similar to recursive and subroutine calls.
279    
280    
281    Version 8.00 19-Oct-09
282    ----------------------
283    
284    1.  The table for translating pcre_compile() error codes into POSIX error codes
285        was out-of-date, and there was no check on the pcre_compile() error code
286        being within the table. This could lead to an OK return being given in
287        error.
288    
289    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
290        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
291        in a Windows environment.
292    
293    3.  The pcregrep --count option prints the count for each file even when it is
294        zero, as does GNU grep. However, pcregrep was also printing all files when
295        --files-with-matches was added. Now, when both options are given, it prints
296        counts only for those files that have at least one match. (GNU grep just
297        prints the file name in this circumstance, but including the count seems
298        more useful - otherwise, why use --count?) Also ensured that the
299        combination -clh just lists non-zero counts, with no names.
300    
301    4.  The long form of the pcregrep -F option was incorrectly implemented as
302        --fixed_strings instead of --fixed-strings. This is an incompatible change,
303        but it seems right to fix it, and I didn't think it was worth preserving
304        the old behaviour.
305    
306    5.  The command line items --regex=pattern and --regexp=pattern were not
307        recognized by pcregrep, which required --regex pattern or --regexp pattern
308        (with a space rather than an '='). The man page documented the '=' forms,
309        which are compatible with GNU grep; these now work.
310    
311    6.  No libpcreposix.pc file was created for pkg-config; there was just
312        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
313    
314    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
315        when UCP support is not needed, by modifying the Python script that
316        generates it from Unicode data files. This should not matter if the module
317        is correctly used as a library, but I received one complaint about 50K of
318        unwanted data. My guess is that the person linked everything into his
319        program rather than using a library. Anyway, it does no harm.
320    
321    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
322        was a minimum greater than 1 for a wide character in a possessive
323        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
324        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
325        character. Chaos in the form of incorrect output or a compiling loop could
326        result.
327    
328    9.  The restrictions on what a pattern can contain when partial matching is
329        requested for pcre_exec() have been removed. All patterns can now be
330        partially matched by this function. In addition, if there are at least two
331        slots in the offset vector, the offset of the earliest inspected character
332        for the match and the offset of the end of the subject are set in them when
333        PCRE_ERROR_PARTIAL is returned.
334    
335    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
336        synonymous with PCRE_PARTIAL, for backwards compatibility, and
337        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
338        and may be more useful for multi-segment matching.
339    
340    11. Partial matching with pcre_exec() is now more intuitive. A partial match
341        used to be given if ever the end of the subject was reached; now it is
342        given only if matching could not proceed because another character was
343        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
344        string "Z", which now yields "no match" instead of "partial match". In the
345        case of pcre_dfa_exec(), "no match" is given if every matching path for the
346        final character ended with (*FAIL).
347    
348    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
349        if the pattern had a "must contain" character that was already found in the
350        earlier partial match, unless partial matching was again requested. For
351        example, with the pattern /dog.(body)?/, the "must contain" character is
352        "g". If the first part-match was for the string "dog", restarting with
353        "sbody" failed. This bug has been fixed.
354    
355    13. The string returned by pcre_dfa_exec() after a partial match has been
356        changed so that it starts at the first inspected character rather than the
357        first character of the match. This makes a difference only if the pattern
358        starts with a lookbehind assertion or \b or \B (\K is not supported by
359        pcre_dfa_exec()). It's an incompatible change, but it makes the two
360        matching functions compatible, and I think it's the right thing to do.
361    
362    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
363        so that the demonstration program is easily available in environments where
364        PCRE has not been installed from source.
365    
366    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
367        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
368        library.
369    
370    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
371        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
372        is not the first non-POSIX option to be added. Clearly some people find
373        these options useful.
374    
375    17. If a caller to the POSIX matching function regexec() passes a non-zero
376        value for nmatch with a NULL value for pmatch, the value of
377        nmatch is forced to zero.
378    
379    18. RunGrepTest did not have a test for the availability of the -u option of
380        the diff command, as RunTest does. It now checks in the same way as
381        RunTest, and also checks for the -b option.
382    
383    19. If an odd number of negated classes containing just a single character
384        interposed, within parentheses, between a forward reference to a named
385        subpattern and the definition of the subpattern, compilation crashed with
386        an internal error, complaining that it could not find the referenced
387        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
388        [The bug was that it was starting one character too far in when skipping
389        over the character class, thus treating the ] as data rather than
390        terminating the class. This meant it could skip too much.]
391    
392    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
393        /g option in pcretest when the pattern contains \K, which makes it possible
394        to have an empty string match not at the start, even when the pattern is
395        anchored. Updated pcretest and pcredemo to use this option.
396    
397    21. If the maximum number of capturing subpatterns in a recursion was greater
398        than the maximum at the outer level, the higher number was returned, but
399        with unset values at the outer level. The correct (outer level) value is
400        now given.
401    
402    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
403        PCRE did not set those parentheses (unlike Perl). I have now found a way to
404        make it do so. The string so far is captured, making this feature
405        compatible with Perl.
406    
407    23. The tests have been re-organized, adding tests 11 and 12, to make it
408        possible to check the Perl 5.10 features against Perl 5.10.
409    
410    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
411        pattern matches a fixed length string. PCRE did not allow this; now it
412        does. Neither allows recursion.
413    
414    25. I finally figured out how to implement a request to provide the minimum
415        length of subject string that was needed in order to match a given pattern.
416        (It was back references and recursion that I had previously got hung up
417        on.) This code has now been added to pcre_study(); it finds a lower bound
418        to the length of subject needed. It is not necessarily the greatest lower
419        bound, but using it to avoid searching strings that are too short does give
420        some useful speed-ups. The value is available to calling programs via
421        pcre_fullinfo().
422    
423    26. While implementing 25, I discovered to my embarrassment that pcretest had
424        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
425        study optimizations had never been tested with that matching function.
426        Oops. What is worse, even when it was passed study data, there was a bug in
427        pcre_dfa_exec() that meant it never actually used it. Double oops. There
428        were also very few tests of studied patterns with pcre_dfa_exec().
429    
430    27. If (?| is used to create subpatterns with duplicate numbers, they are now
431        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
432        on the other side of the coin, they are no longer allowed to have different
433        names, because these cannot be distinguished in PCRE, and this has caused
434        confusion. (This is a difference from Perl.)
435    
436    28. When duplicate subpattern names are present (necessarily with different
437        numbers, as required by 27 above), and a test is made by name in a
438        conditional pattern, either for a subpattern having been matched, or for
439        recursion in such a pattern, all the associated numbered subpatterns are
440        tested, and the overall condition is true if the condition is true for any
441        one of them. This is the way Perl works, and is also more like the way
442        testing by number works.
443    
444    
445    Version 7.9 11-Apr-09
446    ---------------------
447    
448    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
449        (pcretest), all targets were linked against these libraries. This included
450        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
451        libraries. This caused unwanted dependencies to be created. This problem
452        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
453        pcretest is linked with readline.
454    
455    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
456        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
457        moved outside it again, because FALSE and TRUE are already defined in AIX,
458        but BOOL is not.
459    
460    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
461        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
462    
463    4.  The pcregrep documentation said spaces were inserted as well as colons (or
464        hyphens) following file names and line numbers when outputting matching
465        lines. This is not true; no spaces are inserted. I have also clarified the
466        wording for the --colour (or --color) option.
467    
468    5.  In pcregrep, when --colour was used with -o, the list of matching strings
469        was not coloured; this is different to GNU grep, so I have changed it to be
470        the same.
471    
472    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
473        each matching line was coloured. Now it goes on to look for further matches
474        of any of the test patterns, which is the same behaviour as GNU grep.
475    
476    7.  A pattern that could match an empty string could cause pcregrep to loop; it
477        doesn't make sense to accept an empty string match in pcregrep, so I have
478        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
479        seems to be how GNU grep behaves.
480    
481    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
482        start or after a newline", because the conditional assertion was not being
483        correctly handled. The rule now is that both the assertion and what follows
484        in the first alternative must satisfy the test.
485    
486    9.  If auto-callout was enabled in a pattern with a conditional group whose
487        condition was an assertion, PCRE could crash during matching, both with
488        pcre_exec() and pcre_dfa_exec().
489    
490    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
491        used for matching.
492    
493    11. Unicode property support in character classes was not working for
494        characters (bytes) greater than 127 when not in UTF-8 mode.
495    
496    12. Added the -M command line option to pcretest.
497    
498    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
499    
500    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
501    
502    16. Added comments and documentation about mis-use of no_arg in the C++
503        wrapper.
504    
505    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
506        from Martin Jerabek that uses macro names for all relevant character and
507        string constants.
508    
509    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
510        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
511        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
512        these, but not everybody uses configure.
513    
514    19. A conditional group that had only one branch was not being correctly
515        recognized as an item that could match an empty string. This meant that an
516        enclosing group might also not be so recognized, causing infinite looping
517        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
518        with the subject "ab", where knowledge that the repeated group can match
519        nothing is needed in order to break the loop.
520    
521    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
522        exec(), but without supplying a callout function, matching went wrong.
523    
524    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
525        leak if the size of the offset vector was greater than 30. When the vector
526        is smaller, the saved offsets during recursion go onto a local stack
527        vector, but for larger vectors malloc() is used. It was failing to free
528        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
529        error, in fact).
530    
531    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
532        heapframe that is used only when UTF-8 support is enabled. This caused no
533        problem, but was untidy.
534    
535    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
536        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
537        included within another project.
538    
539    24. Steven Van Ingelgem's patches to add more options to the CMake support,
540        slightly modified by me:
541    
542          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
543              not building pcregrep.
544    
545          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
546              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
547    
548    25. Forward references, both numeric and by name, in patterns that made use of
549        duplicate group numbers, could behave incorrectly or give incorrect errors,
550        because when scanning forward to find the reference group, PCRE was not
551        taking into account the duplicate group numbers. A pattern such as
552        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
553    
554    26. Changed a few more instances of "const unsigned char *" to USPTR, making
555        the feature of a custom pointer more persuasive (as requested by a user).
556    
557    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
558        pcretest.c, inside #ifndefs, because it seems they are sometimes already
559        pre-defined.
560    
561    28. Added support for (*UTF8) at the start of a pattern.
562    
563    29. Arrange for flags added by the "release type" setting in CMake to be shown
564        in the configuration summary.
565    
566    
567    Version 7.8 05-Sep-08
568    ---------------------
569    
570    1.  Replaced UCP searching code with optimized version as implemented for Ad
571        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
572        stage table and inline lookup instead of a function, giving speed ups of 2
573        to 5 times on some simple patterns that I tested. Permission was given to
574        distribute the MultiStage2.py script that generates the tables (it's not in
575        the tarball, but is in the Subversion repository).
576    
577    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
578        scripts.
579    
580    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
581        a group with a zero qualifier. The result of the study could be incorrect,
582        or the function might crash, depending on the pattern.
583    
584    4.  Caseless matching was not working for non-ASCII characters in back
585        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
586        It now works when Unicode Property Support is available.
587    
588    5.  In pcretest, an escape such as \x{de} in the data was always generating
589        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
590        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
591        truncation.
592    
593    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
594    
595    7.  Added two (int) casts to pcregrep when printing the difference of two
596        pointers, in case they are 64-bit values.
597    
598    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
599        test 2 if it fails.
600    
601    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
602        and a #define of that name to empty if it is not externally set. This is to
603        allow users of MSVC to set it if necessary.
604    
605    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
606        the convenience functions in the pcre_get.c source file.
607    
608    11. An option change at the start of a pattern that had top-level alternatives
609        could cause overwriting and/or a crash. This command provoked a crash in
610        some environments:
611    
612          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
613    
614        This potential security problem was recorded as CVE-2008-2371.
615    
616    12. For a pattern where the match had to start at the beginning or immediately
617        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
618        pcre_dfa_exec() could read past the end of the passed subject if there was
619        no match. To help with detecting such bugs (e.g. with valgrind), I modified
620        pcretest so that it places the subject at the end of its malloc-ed buffer.
621    
622    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
623        exec() might read past the end of the data buffer in UTF-8 mode.
624    
625    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
626        the data contained the byte 0x85 as part of a UTF-8 character within its
627        first line. This applied both to normal and DFA matching.
628    
629    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
630        /^[^d]*?$/8 failed to match "abc".
631    
632    16. Added a missing copyright notice to pcrecpp_internal.h.
633    
634    17. Make it more clear in the documentation that values returned from
635        pcre_exec() in ovector are byte offsets, not character counts.
636    
637    18. Tidied a few places to stop certain compilers from issuing warnings.
638    
639    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
640        supplied by Stefan Weber. I made a further small update for 7.8 because
641        there is a change of source arrangements: the pcre_searchfuncs.c module is
642        replaced by pcre_ucd.c.
643    
644    
645    Version 7.7 07-May-08
646    ---------------------
647    
648    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
649        a string to a long long, pretend we don't even have a long long." This is
650        done by checking for the strtoq, strtoll, and _strtoi64 functions.
651    
652    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
653        pre-7.6 versions, which defined a global no_arg variable instead of putting
654        it in the RE class. (See also #8 below.)
655    
656    3.  Remove a line of dead code, identified by coverity and reported by Nuno
657        Lopes.
658    
659    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
660    
661        (1) The include/exclude patterns were being applied to the whole pathnames
662            of files, instead of just to the final components.
663    
664        (2) If there was more than one level of directory, the subdirectories were
665            skipped unless they satisfied the include/exclude conditions. This is
666            inconsistent with GNU grep (and could even be seen as contrary to the
667            pcregrep specification - which I improved to make it absolutely clear).
668            The action now is always to scan all levels of directory, and just
669            apply the include/exclude patterns to regular files.
670    
671    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
672        --exclude_dir in the tests to avoid scanning .svn directories.
673    
674    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
675        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
676        doesn't support NULs in patterns.
677    
678    7.  Added some missing "const"s to declarations of static tables in
679        pcre_compile.c and pcre_dfa_exec.c.
680    
681    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
682        caused by fix #2  above. (Subsequently also a second patch to fix the
683        first patch. And a third patch - this was a messy problem.)
684    
685    9.  Applied Craig's patch to remove the use of push_back().
686    
687    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
688        matching function regexec().
689    
690    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
691        which, however, unlike Perl's \g{...}, are subroutine calls, not back
692        references. PCRE supports relative numbers with this syntax (I don't think
693        Oniguruma does).
694    
695    12. Previously, a group with a zero repeat such as (...){0} was completely
696        omitted from the compiled regex. However, this means that if the group
697        was called as a subroutine from elsewhere in the pattern, things went wrong
698        (an internal error was given). Such groups are now left in the compiled
699        pattern, with a new opcode that causes them to be skipped at execution
700        time.
701    
702    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
703        to the way PCRE behaves:
704    
705        (a) A lone ] character is dis-allowed (Perl treats it as data).
706    
707        (b) A back reference to an unmatched subpattern matches an empty string
708            (Perl fails the current match path).
709    
710        (c) A data ] in a character class must be notated as \] because if the
711            first data character in a class is ], it defines an empty class. (In
712            Perl it is not possible to have an empty class.) The empty class []
713            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
714            The negative empty class [^] matches any one character, independently
715            of the DOTALL setting.
716    
717    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
718        non-existent subpattern following a character class starting with ']' and
719        containing () gave an internal compiling error instead of "reference to
720        non-existent subpattern". Fortunately, when the pattern did exist, the
721        compiled code was correct. (When scanning forwards to check for the
722        existencd of the subpattern, it was treating the data ']' as terminating
723        the class, so got the count wrong. When actually compiling, the reference
724        was subsequently set up correctly.)
725    
726    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
727        it was being rejected as not supported by pcre_dfa_exec(), even though
728        other assertions are supported. I have made pcre_dfa_exec() support
729        (*FAIL).
730    
731    16. The implementation of 13c above involved the invention of a new opcode,
732        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
733        cannot be changed at match time, I realized I could make a small
734        improvement to matching performance by compiling OP_ALLANY instead of
735        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
736        on the OP_ANY path.
737    
738    17. Compiling pcretest on Windows with readline support failed without the
739        following two fixes: (1) Make the unistd.h include conditional on
740        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
741    
742    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
743        ncurses library to be included for pcretest when ReadLine support is
744        requested, but also to allow for it to be overridden. This patch came from
745        Daniel Bergström.
746    
747    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
748        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
749        any errors with the current Unicode tables. Thanks to Peter Kankowski for
750        spotting this.
751    
752    
753    Version 7.6 28-Jan-08
754    ---------------------
755    
756    1.  A character class containing a very large number of characters with
757        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
758        overflow.
759    
760    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
761        HAVE_LONG_LONG is not defined.
762    
763    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
764        bring them up to date and include new features. This patch includes:
765    
766        - Fixed PH's badly added libz and libbz2 support.
767        - Fixed a problem with static linking.
768        - Added pcredemo. [But later removed - see 7 below.]
769        - Fixed dftables problem and added an option.
770        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
771            HAVE_LONG_LONG.
772        - Added readline support for pcretest.
773        - Added an listing of the option settings after cmake has run.
774    
775    4.  A user submitted a patch to Makefile that makes it easy to create
776        "pcre.dll" under mingw when using Configure/Make. I added stuff to
777        Makefile.am that cause it to include this special target, without
778        affecting anything else. Note that the same mingw target plus all
779        the other distribution libraries and programs are now supported
780        when configuring with CMake (see 6 below) instead of with
781        Configure/Make.
782    
783    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
784        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
785        exported in the Windows port". It has not yet been confirmed that the patch
786        solves the problem, but it does no harm.
787    
788    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
789        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
790        with CMake, and also correct the comment about stack recursion.
791    
792    7.  Remove the automatic building of pcredemo from the ./configure system and
793        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
794        of a program that users should build themselves after PCRE is installed, so
795        building it automatically is not really right. What is more, it gave
796        trouble in some build environments.
797    
798    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
799    
800    
801    Version 7.5 10-Jan-08
802    ---------------------
803    
804    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
805        values in parens when parsing an RE using the C++ wrapper."
806    
807    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
808        Characters greater than 255 were excluded from the class instead of being
809        included.
810    
811    3.  The same bug as (2) above applied to negated POSIX classes such as
812        [:^space:].
813    
814    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
815        defined or documented. It seems to have been a typo for PCRE_STATIC, so
816        I have changed it.
817    
818    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
819        first named subpattern) and a construct such as (?&a) would reference the
820        first named subpattern whose name started with "a" (in other words, the
821        length check was missing). Both these problems are fixed. "Subpattern name
822        expected" is now given for (?&) (a zero-length name), and this patch also
823        makes it give the same error for \k'' (previously it complained that that
824        was a reference to a non-existent subpattern).
825    
826    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
827        this is right because (?- can be followed by option settings as well as by
828        digits. I have, however, made the messages clearer.
829    
830    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
831        than the number used in the conditional) now cause a compile-time error.
832        This is actually not compatible with Perl, which accepts such patterns, but
833        treats the conditional as always being FALSE (as PCRE used to), but it
834        seems to me that giving a diagnostic is better.
835    
836    8.  Change "alphameric" to the more common word "alphanumeric" in comments
837        and messages.
838    
839    9.  Fix two occurrences of "backslash" in comments that should have been
840        "backspace".
841    
842    10. Remove two redundant lines of code that can never be obeyed (their function
843        was moved elsewhere).
844    
845    11. The program that makes PCRE's Unicode character property table had a bug
846        which caused it to generate incorrect table entries for sequences of
847        characters that have the same character type, but are in different scripts.
848        It amalgamated them into a single range, with the script of the first of
849        them. In other words, some characters were in the wrong script. There were
850        thirteen such cases, affecting characters in the following ranges:
851    
852          U+002b0 - U+002c1
853          U+0060c - U+0060d
854          U+0061e - U+00612
855          U+0064b - U+0065e
856          U+0074d - U+0076d
857          U+01800 - U+01805
858          U+01d00 - U+01d77
859          U+01d9b - U+01dbf
860          U+0200b - U+0200f
861          U+030fc - U+030fe
862          U+03260 - U+0327f
863          U+0fb46 - U+0fbb1
864          U+10450 - U+1049d
865    
866    12. The -o option (show only the matching part of a line) for pcregrep was not
867        compatible with GNU grep in that, if there was more than one match in a
868        line, it showed only the first of them. It now behaves in the same way as
869        GNU grep.
870    
871    13. If the -o and -v options were combined for pcregrep, it printed a blank
872        line for every non-matching line. GNU grep prints nothing, and pcregrep now
873        does the same. The return code can be used to tell if there were any
874        non-matching lines.
875    
876    14. Added --file-offsets and --line-offsets to pcregrep.
877    
878    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
879        infinitely looping recursion. The bug was that positive lookaheads were not
880        being skipped when checking for a possible empty match (negative lookaheads
881        and both kinds of lookbehind were skipped).
882    
883    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
884        inclusion of <windows.h> to before rather than after the definition of
885        INVALID_FILE_ATTRIBUTES (patch from David Byron).
886    
887    17. Specifying a possessive quantifier with a specific limit for a Unicode
888        character property caused pcre_compile() to compile bad code, which led at
889        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
890        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
891        caused the error; without that there was no problem.
892    
893    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
894    
895    19. Added --enable-pcretest-libreadline.
896    
897    20. In pcrecpp.cc, the variable 'count' was incremented twice in
898        RE::GlobalReplace(). As a result, the number of replacements returned was
899        double what it should be. I removed one of the increments, but Craig sent a
900        later patch that removed the other one (the right fix) and added unit tests
901        that check the return values (which was not done before).
902    
903    21. Several CMake things:
904    
905        (1) Arranged that, when cmake is used on Unix, the libraries end up with
906            the names libpcre and libpcreposix, not just pcre and pcreposix.
907    
908        (2) The above change means that pcretest and pcregrep are now correctly
909            linked with the newly-built libraries, not previously installed ones.
910    
911        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
912    
913    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
914        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
915        UTF-8 newline character). The key issue is that the pattern starts .*;
916        this means that the match must be either at the beginning, or after a
917        newline. The bug was in the code for advancing after a failed match and
918        checking that the new position followed a newline. It was not taking
919        account of UTF-8 characters correctly.
920    
921    23. PCRE was behaving differently from Perl in the way it recognized POSIX
922        character classes. PCRE was not treating the sequence [:...:] as a
923        character class unless the ... were all letters. Perl, however, seems to
924        allow any characters between [: and :], though of course it rejects as
925        unknown any "names" that contain non-letters, because all the known class
926        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
927        for example, whereas PCRE did not - it did not recognize a POSIX character
928        class. This seemed a bit dangerous, so the code has been changed to be
929        closer to Perl. The behaviour is not identical to Perl, because PCRE will
930        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
931        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
932        Perl does, and where it didn't before.
933    
934    24. Rewrite so as to remove the single use of %n from pcregrep because in some
935        Windows environments %n is disabled by default.
936    
937    
938    Version 7.4 21-Sep-07
939    ---------------------
940    
941    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
942        means that a class such as [\s] counted as "explicit reference to CR or
943        LF". That isn't really right - the whole point of the change was to try to
944        help when there was an actual mention of one of the two characters. So now
945        the change happens only if \r or \n (or a literal CR or LF) character is
946        encountered.
947    
948    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
949        of both had grown to the point where there were only 3 bits left.
950        Fortunately, there was spare space in the data structure, and so I have
951        moved the internal flags into a new 16-bit field to free up more option
952        bits.
953    
954    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
955        but did not set the internal JCHANGED flag - either of these is enough to
956        control the way the "get" function works - but the PCRE_INFO_JCHANGED
957        facility is supposed to tell if (?J) was ever used, so now (?J) at the
958        start sets both bits.
959    
960    4.  Added options (at build time, compile time, exec time) to change \R from
961        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
962    
963    5.  doc/pcresyntax.html was missing from the distribution.
964    
965    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
966        compatibility, even though it is no longer used.
967    
968    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
969        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
970        windows.h file is present (where different names are used). [This was
971        reversed later after testing - see 16 below.]
972    
973    8.  Changed all #include <config.h> to #include "config.h". There were also
974        some further <pcre.h> cases that I changed to "pcre.h".
975    
976    9.  When pcregrep was used with the --colour option, it missed the line ending
977        sequence off the lines that it output.
978    
979    10. It was pointed out to me that arrays of string pointers cause lots of
980        relocations when a shared library is dynamically loaded. A technique of
981        using a single long string with a table of offsets can drastically reduce
982        these. I have refactored PCRE in four places to do this. The result is
983        dramatic:
984    
985          Originally:                          290
986          After changing UCP table:            187
987          After changing error message table:   43
988          After changing table of "verbs"       36
989          After changing table of Posix names   22
990    
991        Thanks to the folks working on Gregex for glib for this insight.
992    
993    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
994        unicode-properties was also set.
995    
996    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
997    
998    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
999        checked only for CRLF.
1000    
1001    14. Added casts to pcretest.c to avoid compiler warnings.
1002    
1003    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1004    
1005    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1006        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1007        entirely. This removes changes made in 7 above.
1008    
1009    17. The CMake files have been updated, and there is now more information about
1010        building with CMake in the NON-UNIX-USE document.
1011    
1012    
1013    Version 7.3 28-Aug-07
1014    ---------------------
1015    
1016     1. In the rejigging of the build system that eventually resulted in 7.1, the
1017        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1018        brackets there is not right, since it causes compilers to look for an
1019        installed pcre.h, not the version that is in the source that is being
1020        compiled (which of course may be different). I have changed it back to:
1021    
1022          #include "pcre.h"
1023    
1024        I have a vague recollection that the change was concerned with compiling in
1025        different directories, but in the new build system, that is taken care of
1026        by the VPATH setting the Makefile.
1027    
1028     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1029        when the subject happened to end in the byte 0x85 (e.g. if the last
1030        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1031        characters but of course it shouldn't be taken as a newline when it is part
1032        of another character. The bug was that, for an unlimited repeat of . in
1033        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1034        characters when looking for a newline.
1035    
1036     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1037    
1038     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1039        in debug output.
1040    
1041     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1042        long printing in the pcrecpp unittest when running under MinGW.
1043    
1044     6. ESC_K was left out of the EBCDIC table.
1045    
1046     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1047        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1048        limit also applies to "virtual nesting" when a pattern is recursive, and in
1049        this case 1000 isn't so big. I have been able to remove this limit at the
1050        expense of backing off one optimization in certain circumstances. Normally,
1051        when pcre_exec() would call its internal match() function recursively and
1052        immediately return the result unconditionally, it uses a "tail recursion"
1053        feature to save stack. However, when a subpattern that can match an empty
1054        string has an unlimited repetition quantifier, it no longer makes this
1055        optimization. That gives it a stack frame in which to save the data for
1056        checking that an empty string has been matched. Previously this was taken
1057        from the 1000-entry workspace that had been reserved. So now there is no
1058        explicit limit, but more stack is used.
1059    
1060     8. Applied Daniel's patches to solve problems with the import/export magic
1061        syntax that is required for Windows, and which was going wrong for the
1062        pcreposix and pcrecpp parts of the library. These were overlooked when this
1063        problem was solved for the main library.
1064    
1065     9. There were some crude static tests to avoid integer overflow when computing
1066        the size of patterns that contain repeated groups with explicit upper
1067        limits. As the maximum quantifier is 65535, the maximum group length was
1068        set at 30,000 so that the product of these two numbers did not overflow a
1069        32-bit integer. However, it turns out that people want to use groups that
1070        are longer than 30,000 bytes (though not repeat them that many times).
1071        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1072        made it possible to implement the integer overflow checks in a much more
1073        dynamic way, which I have now done. The artificial limitation on group
1074        length has been removed - we now have only the limit on the total length of
1075        the compiled pattern, which depends on the LINK_SIZE setting.
1076    
1077    10. Fixed a bug in the documentation for get/copy named substring when
1078        duplicate names are permitted. If none of the named substrings are set, the
1079        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1080        empty string.
1081    
1082    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1083        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1084        because the ] is interpreted as the first data character and the
1085        terminating ] is not found. PCRE has been made compatible with Perl in this
1086        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1087        cause memory overwriting.
1088    
1089    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1090        string has been matched (to stop an infinite loop). It was not recognizing
1091        a conditional subpattern that could match an empty string if that
1092        subpattern was within another subpattern. For example, it looped when
1093        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1094        condition was not nested. This bug has been fixed.
1095    
1096    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1097        past the start of the subject in the presence of bytes with the top bit
1098        set, for example "\x8aBCD".
1099    
1100    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1101        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1102    
1103    14. Optimized (?!) to (*FAIL).
1104    
1105    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1106        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1107        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1108        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1109        does: it's just the validity check that is more restrictive.
1110    
1111    16. Inserted checks for integer overflows during escape sequence (backslash)
1112        processing, and also fixed erroneous offset values for syntax errors during
1113        backslash processing.
1114    
1115    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1116        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1117    
1118    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1119        caused an overrun.
1120    
1121    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1122        something other than just ASCII characters) inside a group that had an
1123        unlimited repeat caused a loop at compile time (while checking to see
1124        whether the group could match an empty string).
1125    
1126    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1127        [\P{Any}] did so. (Error in the code for printing property names.)
1128    
1129    21. An orphan \E inside a character class could cause a crash.
1130    
1131    22. A repeated capturing bracket such as (A)? could cause a wild memory
1132        reference during compilation.
1133    
1134    23. There are several functions in pcre_compile() that scan along a compiled
1135        expression for various reasons (e.g. to see if it's fixed length for look
1136        behind). There were bugs in these functions when a repeated \p or \P was
1137        present in the pattern. These operators have additional parameters compared
1138        with \d, etc, and these were not being taken into account when moving along
1139        the compiled data. Specifically:
1140    
1141        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1142            length.
1143    
1144        (b) An item such as \pL+ within a repeated group could cause crashes or
1145            loops.
1146    
1147        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1148            "reference to non-existent subpattern" error.
1149    
1150        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1151    
1152    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1153        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1154    
1155    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1156    
1157    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1158        character were causing crashes (broken optimization).
1159    
1160    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1161        \p or \P) caused a compile-time loop.
1162    
1163    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1164        break. For example, the unstudied pattern [\r\n]A does not match the string
1165        "\r\nA" because change 7.0/46 below moves the current point on by two
1166        characters after failing to match at the start. However, the pattern \nA
1167        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1168        the same is true. There doesn't seem any very clean way out of this, but
1169        what I have chosen to do makes the common cases work: PCRE now takes note
1170        of whether there can be an explicit match for \r or \n anywhere in the
1171        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1172        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1173        pattern has explicit CR or LF references.
1174    
1175    29. Added (*CR) etc for changing newline setting at start of pattern.
1176    
1177    
1178    Version 7.2 19-Jun-07
1179    ---------------------
1180    
1181     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1182        which is apparently normally available under Windows.
1183    
1184     2. Re-jig the pcregrep tests with different newline settings in an attempt
1185        to make them independent of the local environment's newline setting.
1186    
1187     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1188    
1189     4. Some of the "internals" tests were previously cut out when the link size
1190        was not 2, because the output contained actual offsets. The recent new
1191        "Z" feature of pcretest means that these can be cut out, making the tests
1192        usable with all link sizes.
1193    
1194     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1195        stack recursion. This gives a massive performance boost under BSD, but just
1196        a small improvement under Linux. However, it saves one field in the frame
1197        in all cases.
1198    
1199     6. Added more features from the forthcoming Perl 5.10:
1200    
1201        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1202            recursion call. It refers to the nth most recently opened parentheses.
1203    
1204        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1205            to be opened parentheses.
1206    
1207        (c) Conditions that refer to capturing parentheses can be specified
1208            relatively, for example, (?(-2)... or (?(+3)...
1209    
1210        (d) \K resets the start of the current match so that everything before
1211            is not part of it.
1212    
1213        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1214    
1215        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1216            reference syntax.
1217    
1218        (g) (?| introduces a group in which the numbering of parentheses in each
1219            alternative starts with the same number.
1220    
1221        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1222    
1223     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1224        PCRE_INFO_JCHANGED.
1225    
1226     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1227        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1228        for detecting groups that can match an empty string.
1229    
1230     9. A pattern with a very large number of alternatives (more than several
1231        hundred) was running out of internal workspace during the pre-compile
1232        phase, where pcre_compile() figures out how much memory will be needed. A
1233        bit of new cunning has reduced the workspace needed for groups with
1234        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1235        workspace instead of running out of the 4096 that are available.
1236    
1237    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1238    
1239    11. Applied patch from Google to remove an optimization that didn't quite work.
1240        The report of the bug said:
1241    
1242          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1243          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1244          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1245    
1246    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1247        it matched the wrong number of bytes.
1248    
1249    
1250    Version 7.1 24-Apr-07
1251    ---------------------
1252    
1253     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1254        that is more "standard", making use of automake and other Autotools. There
1255        is some re-arrangement of the files and adjustment of comments consequent
1256        on this.
1257    
1258     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1259        for recursive directory scanning broke on some systems because the files
1260        are not scanned in any specific order and on different systems the order
1261        was different. A call to "sort" has been inserted into RunGrepTest for the
1262        approprate test as a short-term fix. In the longer term there may be an
1263        alternative.
1264    
1265     3. I had an email from Eric Raymond about problems translating some of PCRE's
1266        man pages to HTML (despite the fact that I distribute HTML pages, some
1267        people do their own conversions for various reasons). The problems
1268        concerned the use of low-level troff macros .br and .in. I have therefore
1269        removed all such uses from the man pages (some were redundant, some could
1270        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1271        HTML has been updated to handle .nf/.fi and to complain if it encounters
1272        .br or .in.
1273    
1274     4. Updated comments in configure.ac that get placed in config.h.in and also
1275        arranged for config.h to be included in the distribution, with the name
1276        config.h.generic, for the benefit of those who have to compile without
1277        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1278    
1279     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1280        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1281        makevp.bat for the current PCRE, using the additional files
1282        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1283    
1284     6. A Windows user reported a minor discrepancy with test 2, which turned out
1285        to be caused by a trailing space on an input line that had got lost in his
1286        copy. The trailing space was an accident, so I've just removed it.
1287    
1288     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1289        that is needed.
1290    
1291     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1292        as "const" (a) because they are and (b) because it helps the PHP
1293        maintainers who have recently made a script to detect big data structures
1294        in the php code that should be moved to the .rodata section. I remembered
1295        to update Builducptable as well, so it won't revert if ucptable.h is ever
1296        re-created.
1297    
1298     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1299        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1300        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1301        support is not required. This saves 1.5-2K of code, which is important in
1302        some applications.
1303    
1304        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1305        so as not to refer to the tables, even though these functions will never be
1306        called when UTF-8 support is disabled. Otherwise there are problems with a
1307        shared library.
1308    
1309    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1310    
1311        (a) It was defining its arguments as char * instead of void *.
1312    
1313        (b) It was assuming that all moves were upwards in memory; this was true
1314            a long time ago when I wrote it, but is no longer the case.
1315    
1316        The emulated memove() is provided for those environments that have neither
1317        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1318        is clearly not the case, as these two bugs were recently reported.
1319    
1320    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1321        and Detrail to create the HTML documentation, the .txt form of the man
1322        pages, and it removes trailing spaces from listed files. It also creates
1323        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1324        case, it wraps all the #defines with #ifndefs. This script should be run
1325        before "make dist".
1326    
1327    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1328        with Unicode property support.
1329    
1330        (a) For a maximizing quantifier, if the two different cases of the
1331            character were of different lengths in their UTF-8 codings (there are
1332            some cases like this - I found 11), and the matching function had to
1333            back up over a mixture of the two cases, it incorrectly assumed they
1334            were both the same length.
1335    
1336        (b) When PCRE was configured to use the heap rather than the stack for
1337            recursion during matching, it was not correctly preserving the data for
1338            the other case of a UTF-8 character when checking ahead for a match
1339            while processing a minimizing repeat. If the check also involved
1340            matching a wide character, but failed, corruption could cause an
1341            erroneous result when trying to check for a repeat of the original
1342            character.
1343    
1344    13. Some tidying changes to the testing mechanism:
1345    
1346        (a) The RunTest script now detects the internal link size and whether there
1347            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1348            values substituted by "configure". (The RunGrepTest script already did
1349            this for UTF-8.) The configure.ac script no longer substitutes the
1350            relevant variables.
1351    
1352        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1353            with length and offset values. This means that the output is different
1354            for different internal link sizes. Test 2 is skipped for link sizes
1355            other than 2 because of this, bypassing the problem. Unfortunately,
1356            there was also a test in test 3 (the locale tests) that used /B and
1357            failed for link sizes other than 2. Rather than cut the whole test out,
1358            I have added a new /Z option to pcretest that replaces the length and
1359            offset values with spaces. This is now used to make test 3 independent
1360            of link size. (Test 2 will be tidied up later.)
1361    
1362    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1363        segmentation fault instead of returning the appropriate error message.
1364    
1365    15. In multiline mode when the newline sequence was set to "any", the pattern
1366        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1367        This doesn't seem right; it now treats the CRLF combination as the line
1368        ending, and so does not match in that case. It's only a pattern such as ^$
1369        that would hit this one: something like ^ABC$ would have failed after \r
1370        and then tried again after \r\n.
1371    
1372    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1373        in an attempt to make files that differ only in their line terminators
1374        compare equal. This works on Linux.
1375    
1376    17. Under certain error circumstances pcregrep might try to free random memory
1377        as it exited. This is now fixed, thanks to valgrind.
1378    
1379    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1380        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1381        was because its rules for how to advance for /g after matching an empty
1382        string at the end of a line did not allow for this case. They now check for
1383        it specially.
1384    
1385    20. pcretest is supposed to handle patterns and data of any length, by
1386        extending its buffers when necessary. It was getting this wrong when the
1387        buffer for a data line had to be extended.
1388    
1389    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1390        CRLF as a newline sequence.
1391    
1392    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1393        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1394        I have nevertheless tidied it up.
1395    
1396    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1397    
1398    24. Added a man page for pcre-config.
1399    
1400    
1401    Version 7.0 19-Dec-06
1402    ---------------------
1403    
1404     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1405        moving to gcc 4.1.1.
1406    
1407     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1408        sys/time.h, which is documented as needed for this function. It doesn't
1409        seem to matter on Linux, but it showed up on some releases of OS X.
1410    
1411     3. It seems that there are systems where bytes whose values are greater than
1412        127 match isprint() in the "C" locale. The "C" locale should be the
1413        default when a C program starts up. In most systems, only ASCII printing
1414        characters match isprint(). This difference caused the output from pcretest
1415        to vary, making some of the tests fail. I have changed pcretest so that:
1416    
1417        (a) When it is outputting text in the compiled version of a pattern, bytes
1418            other than 32-126 are always shown as hex escapes.
1419    
1420        (b) When it is outputting text that is a matched part of a subject string,
1421            it does the same, unless a different locale has been set for the match
1422            (using the /L modifier). In this case, it uses isprint() to decide.
1423    
1424     4. Fixed a major bug that caused incorrect computation of the amount of memory
1425        required for a compiled pattern when options that changed within the
1426        pattern affected the logic of the preliminary scan that determines the
1427        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1428        that the computed length was too small. The symptoms of this bug were
1429        either the PCRE error "internal error: code overflow" from pcre_compile(),
1430        or a glibc crash with a message such as "pcretest: free(): invalid next
1431        size (fast)". Examples of patterns that provoked this bug (shown in
1432        pcretest format) are:
1433    
1434          /(?-x: )/x
1435          /(?x)(?-x: \s*#\s*)/
1436          /((?i)[\x{c0}])/8
1437          /(?i:[\x{c0}])/8
1438    
1439        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1440        is now done differently.
1441    
1442     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1443        wrapper classes; (b) implement a new function in the C++ scanner that is
1444        more efficient than the old way of doing things because it avoids levels of
1445        recursion in the regex matching; (c) add a paragraph to the documentation
1446        for the FullMatch() function.
1447    
1448     6. The escape sequence \n was being treated as whatever was defined as
1449        "newline". Not only was this contrary to the documentation, which states
1450        that \n is character 10 (hex 0A), but it also went horribly wrong when
1451        "newline" was defined as CRLF. This has been fixed.
1452    
1453     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1454        was being set to -1 for the "end of line" case (supposedly a value that no
1455        character can have). Though this value is never used (the check for end of
1456        line is "zero bytes in current character"), it caused compiler complaints.
1457        I've changed it to 0xffffffff.
1458    
1459     8. In pcre_version.c, the version string was being built by a sequence of
1460        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1461        string (as it is for production releases) called a macro with an empty
1462        argument. The C standard says the result of this is undefined. The gcc
1463        compiler treats it as an empty string (which was what was wanted) but it is
1464        reported that Visual C gives an error. The source has been hacked around to
1465        avoid this problem.
1466    
1467     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1468        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1469        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1470        of them did).
1471    
1472    10. Originally, pcretest opened its input and output without "b"; then I was
1473        told that "b" was needed in some environments, so it was added for release
1474        5.0 to both the input and output. (It makes no difference on Unix-like
1475        systems.) Later I was told that it is wrong for the input on Windows. I've
1476        now abstracted the modes into two macros, to make it easier to fiddle with
1477        them, and removed "b" from the input mode under Windows.
1478    
1479    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1480    
1481    12. Added -help and --help to pcretest as an official way of being reminded
1482        of the options.
1483    
1484    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1485        and pcrecpp.cc because they annoy compilers at high warning levels.
1486    
1487    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1488    
1489    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1490        scripts are not C programs :-) and which was not noticed because it works
1491        on Linux.
1492    
1493    16. pcretest is supposed to handle any length of pattern and data line (as one
1494        line or as a continued sequence of lines) by extending its input buffer if
1495        necessary. This feature was broken for very long pattern lines, leading to
1496        a string of junk being passed to pcre_compile() if the pattern was longer
1497        than about 50K.
1498    
1499    17. I have done a major re-factoring of the way pcre_compile() computes the
1500        amount of memory needed for a compiled pattern. Previously, there was code
1501        that made a preliminary scan of the pattern in order to do this. That was
1502        OK when PCRE was new, but as the facilities have expanded, it has become
1503        harder and harder to keep it in step with the real compile phase, and there
1504        have been a number of bugs (see for example, 4 above). I have now found a
1505        cunning way of running the real compile function in a "fake" mode that
1506        enables it to compute how much memory it would need, while actually only
1507        ever using a few hundred bytes of working memory and without too many
1508        tests of the mode. This should make future maintenance and development
1509        easier. A side effect of this work is that the limit of 200 on the nesting
1510        depth of parentheses has been removed (though this was never a serious
1511        limitation, I suspect). However, there is a downside: pcre_compile() now
1512        runs more slowly than before (30% or more, depending on the pattern). I
1513        hope this isn't a big issue. There is no effect on runtime performance.
1514    
1515    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1516        newline (only possible for the last line of a file) and it was a
1517        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1518    
1519    19. Added additional timing features to pcretest. (1) The -tm option now times
1520        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1521        separate command line item, by a number that specifies the number of
1522        repeats to use when timing. The default is 50000; this gives better
1523        precision, but takes uncomfortably long for very large patterns.
1524    
1525    20. Extended pcre_study() to be more clever in cases where a branch of a
1526        subpattern has no definite first character. For example, (a*|b*)[cd] would
1527        previously give no result from pcre_study(). Now it recognizes that the
1528        first character must be a, b, c, or d.
1529    
1530    21. There was an incorrect error "recursive call could loop indefinitely" if
1531        a subpattern (or the entire pattern) that was being tested for matching an
1532        empty string contained only one non-empty item after a nested subpattern.
1533        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1534        incorrectly, because the \d was being skipped in the check.
1535    
1536    22. The pcretest program now has a new pattern option /B and a command line
1537        option -b, which is equivalent to adding /B to every pattern. This causes
1538        it to show the compiled bytecode, without the additional information that
1539        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1540        is the same as /B/I).
1541    
1542    23. A new optimization is now able automatically to treat some sequences such
1543        as a*b as a*+b. More specifically, if something simple (such as a character
1544        or a simple class like \d) has an unlimited quantifier, and is followed by
1545        something that cannot possibly match the quantified thing, the quantifier
1546        is automatically "possessified".
1547    
1548    24. A recursive reference to a subpattern whose number was greater than 39
1549        went wrong under certain circumstances in UTF-8 mode. This bug could also
1550        have affected the operation of pcre_study().
1551    
1552    25. Realized that a little bit of performance could be had by replacing
1553        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1554    
1555    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1556    
1557    27. Possessive quantifiers such as a++ were previously implemented by turning
1558        them into atomic groups such as ($>a+). Now they have their own opcodes,
1559        which improves performance. This includes the automatically created ones
1560        from 23 above.
1561    
1562    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1563        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1564        the first matched character to be a colon. This applied both to named and
1565        numbered groups.
1566    
1567    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1568    
1569    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1570        building PCRE on Windows easier, so I have included it in the distribution.
1571    
1572    31. There is now a check in pcretest against a ridiculously large number being
1573        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1574        loop, the loop is abandoned.
1575    
1576    32. Forward references to subpatterns in conditions such as (?(2)...) where
1577        subpattern 2 is defined later cause pcre_compile() to search forwards in
1578        the pattern for the relevant set of parentheses. This search went wrong
1579        when there were unescaped parentheses in a character class, parentheses
1580        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1581    
1582    33. "Subroutine" calls and backreferences were previously restricted to
1583        referencing subpatterns earlier in the regex. This restriction has now
1584        been removed.
1585    
1586    34. Added a number of extra features that are going to be in Perl 5.10. On the
1587        whole, these are just syntactic alternatives for features that PCRE had
1588        previously implemented using the Python syntax or my own invention. The
1589        other formats are all retained for compatibility.
1590    
1591        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1592            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1593            also .NET compatible.
1594    
1595        (b) A recursion or subroutine call to a named group can now be defined as
1596            (?&name) as well as (?P>name).
1597    
1598        (c) A backreference to a named group can now be defined as \k<name> or
1599            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1600            5.10, are also .NET compatible.
1601    
1602        (d) A conditional reference to a named group can now use the syntax
1603            (?(<name>) or (?('name') as well as (?(name).
1604    
1605        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1606            groups (named and numbered) that are never evaluated inline, but can be
1607            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1608            is always false. There may be only one alternative in such a group.
1609    
1610        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1611            as the simple (?(R). The condition is true only if the most recent
1612            recursion is that of the given number or name. It does not search out
1613            through the entire recursion stack.
1614    
1615        (g) The escape \gN or \g{N} has been added, where N is a positive or
1616            negative number, specifying an absolute or relative reference.
1617    
1618    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1619        some "unreachable code" warnings.
1620    
1621    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1622        things, this adds five new scripts.
1623    
1624    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1625        There were also incompatibilities regarding the handling of \Q..\E inside
1626        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1627        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1628    
1629    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1630        matches an empty string, and forcibly breaks the loop. There were bugs in
1631        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1632        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1633        separate and independent bugs (that affected different cases) have been
1634        fixed.
1635    
1636    39. Refactored the code to abolish the use of different opcodes for small
1637        capturing bracket numbers. This is a tidy that I avoided doing when I
1638        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1639        The new approach is not only tidier, it makes it possible to reduce the
1640        memory needed to fix the previous bug (38).
1641    
1642    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1643        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1644        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1645        mode.
1646    
1647    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1648        report.
1649    
1650    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1651        copy construction and assignment in the C++ wrapper.
1652    
1653    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1654        couple of bugs that could have given wrong results in the "--newline=crlf"
1655        case.
1656    
1657    44. Added a number of casts and did some reorganization of signed/unsigned int
1658        variables following suggestions from Dair Grant. Also renamed the variable
1659        "this" as "item" because it is a C++ keyword.
1660    
1661    45. Arranged for dftables to add
1662    
1663          #include "pcre_internal.h"
1664    
1665        to pcre_chartables.c because without it, gcc 4.x may remove the array
1666        definition from the final binary if PCRE is built into a static library and
1667        dead code stripping is activated.
1668    
1669    46. For an unanchored pattern, if a match attempt fails at the start of a
1670        newline sequence, and the newline setting is CRLF or ANY, and the next two
1671        characters are CRLF, advance by two characters instead of one.
1672    
1673    
1674    Version 6.7 04-Jul-06
1675    ---------------------
1676    
1677     1. In order to handle tests when input lines are enormously long, pcretest has
1678        been re-factored so that it automatically extends its buffers when
1679        necessary. The code is crude, but this _is_ just a test program. The
1680        default size has been increased from 32K to 50K.
1681    
1682     2. The code in pcre_study() was using the value of the re argument before
1683        testing it for NULL. (Of course, in any sensible call of the function, it
1684        won't be NULL.)
1685    
1686     3. The memmove() emulation function in pcre_internal.h, which is used on
1687        systems that lack both memmove() and bcopy() - that is, hardly ever -
1688        was missing a "static" storage class specifier.
1689    
1690     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1691        containing an extended class (one that cannot be represented by a bitmap
1692        because it contains high-valued characters or Unicode property items, e.g.
1693        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1694        pattern, but PCRE should not loop if you do not (it no longer does).
1695        [Detail: two cases were found: (a) a repeated subpattern containing an
1696        extended class; (b) a recursive reference to a subpattern that followed a
1697        previous extended class. It wasn't skipping over the extended class
1698        correctly when UTF-8 mode was not set.]
1699    
1700     5. A negated single-character class was not being recognized as fixed-length
1701        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1702        compile error "lookbehind assertion is not fixed length".
1703    
1704     6. The RunPerlTest auxiliary script was showing an unexpected difference
1705        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1706        write a Perl script that can interpret lines of an input file either as
1707        byte characters or as UTF-8, which is what "perltest" was being required to
1708        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1709        can't do is switch easily at run time between having the "use utf8;" pragma
1710        or not. In the end, I fudged it by using the RunPerlTest script to insert
1711        "use utf8;" explicitly for the UTF-8 tests.
1712    
1713     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1714        the end of the subject string, contrary to the documentation and to what
1715        Perl does. This was true of both matching functions. Now it matches only at
1716        the start of the subject and immediately after *internal* newlines.
1717    
1718     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1719        a pointer to an int instead of a pointer to an unsigned long int. This
1720        caused problems on 64-bit systems.
1721    
1722     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1723        instance of the 'standard' template library not being so standard".
1724    
1725    10. There was no check on the number of named subpatterns nor the maximum
1726        length of a subpattern name. The product of these values is used to compute
1727        the size of the memory block for a compiled pattern. By supplying a very
1728        long subpattern name and a large number of named subpatterns, the size
1729        computation could be caused to overflow. This is now prevented by limiting
1730        the length of names to 32 characters, and the number of named subpatterns
1731        to 10,000.
1732    
1733    11. Subpatterns that are repeated with specific counts have to be replicated in
1734        the compiled pattern. The size of memory for this was computed from the
1735        length of the subpattern and the repeat count. The latter is limited to
1736        65535, but there was no limit on the former, meaning that integer overflow
1737        could in principle occur. The compiled length of a repeated subpattern is
1738        now limited to 30,000 bytes in order to prevent this.
1739    
1740    12. Added the optional facility to have named substrings with the same name.
1741    
1742    13. Added the ability to use a named substring as a condition, using the
1743        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1744        are numbers (not recommended). Forward references are permitted.
1745    
1746    14. Added forward references in named backreferences (if you see what I mean).
1747    
1748    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1749        pattern could run off the end of the subject. For example, the pattern
1750        "(?s)(.{1,5})"8 did this with the subject "ab".
1751    
1752    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1753        PCRE_CASELESS was set when matching characters that were quantified with ?
1754        or *.
1755    
1756    17. A character class other than a single negated character that had a minimum
1757        but no maximum quantifier - for example [ab]{6,} - was not handled
1758        correctly by pce_dfa_exec(). It would match only one character.
1759    
1760    18. A valid (though odd) pattern that looked like a POSIX character
1761        class but used an invalid character after [ (for example [[,abc,]]) caused
1762        pcre_compile() to give the error "Failed: internal error: code overflow" or
1763        in some cases to crash with a glibc free() error. This could even happen if
1764        the pattern terminated after [[ but there just happened to be a sequence of
1765        letters, a binary zero, and a closing ] in the memory that followed.
1766    
1767    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1768        over the years. Originally (before any Unicode support), just the bottom 8
1769        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1770        output from "man perlunicode" includes this:
1771    
1772          The regular expression compiler produces polymorphic opcodes.  That
1773          is, the pattern adapts to the data and automatically switches to
1774          the Unicode character scheme when presented with Unicode data--or
1775          instead uses a traditional byte scheme when presented with byte
1776          data.
1777    
1778        Sadly, a wide octal escape does not cause a switch, and in a string with
1779        no other multibyte characters, these octal escapes are treated as before.
1780        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1781        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1782        Unicode string.
1783    
1784        I have not perpetrated such confusion in PCRE. Up till now, it took just
1785        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1786        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1787        translate to the appropriate multibyte character.
1788    
1789    29. Applied some refactoring to reduce the number of warnings from Microsoft
1790        and Borland compilers. This has included removing the fudge introduced
1791        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1792        a warning about an unused variable.
1793    
1794    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1795        characters since release 4.0, because Perl (from release 5.004) does not.
1796        [Or at least, is documented not to: some releases seem to be in conflict
1797        with the documentation.] However, when a pattern was studied with
1798        pcre_study() and all its branches started with \s, PCRE still included VT
1799        as a possible starting character. Of course, this did no harm; it just
1800        caused an unnecessary match attempt.
1801    
1802    22. Removed a now-redundant internal flag bit that recorded the fact that case
1803        dependency changed within the pattern. This was once needed for "required
1804        byte" processing, but is no longer used. This recovers a now-scarce options
1805        bit. Also moved the least significant internal flag bit to the most-
1806        significant bit of the word, which was not previously used (hangover from
1807        the days when it was an int rather than a uint) to free up another bit for
1808        the future.
1809    
1810    23. Added support for CRLF line endings as well as CR and LF. As well as the
1811        default being selectable at build time, it can now be changed at runtime
1812        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1813        specify that it is scanning data with non-default line endings.
1814    
1815    24. Changed the definition of CXXLINK to make it agree with the definition of
1816        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1817    
1818    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1819        recursions. This makes a big different to stack usage for some patterns.
1820    
1821    26. If a subpattern containing a named recursion or subroutine reference such
1822        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1823        the space required for the compiled pattern went wrong and gave too small a
1824        value. Depending on the environment, this could lead to "Failed: internal
1825        error: code overflow at offset 49" or "glibc detected double free or
1826        corruption" errors.
1827    
1828    27. Applied patches from Google (a) to support the new newline modes and (b) to
1829        advance over multibyte UTF-8 characters in GlobalReplace.
1830    
1831    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1832        difference for some implementation of PCRE in some Windows version.
1833    
1834    29. Added some extra testing facilities to pcretest:
1835    
1836        \q<number>   in a data line sets the "match limit" value
1837        \Q<number>   in a data line sets the "match recursion limt" value
1838        -S <number>  sets the stack size, where <number> is in megabytes
1839    
1840        The -S option isn't available for Windows.
1841    
1842    
1843    Version 6.6 06-Feb-06
1844    ---------------------
1845    
1846     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1847        in pcreposix.h. I have copied the definition from pcre.h.
1848    
1849     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1850        because pcre.h is no longer a built file.
1851    
1852     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1853        not normally included in the compiled code.
1854    
1855    
1856    Version 6.5 01-Feb-06
1857    ---------------------
1858    
1859     1. When using the partial match feature with pcre_dfa_exec(), it was not
1860        anchoring the second and subsequent partial matches at the new starting
1861        point. This could lead to incorrect results. For example, with the pattern
1862        /1234/, partially matching against "123" and then "a4" gave a match.
1863    
1864     2. Changes to pcregrep:
1865    
1866        (a) All non-match returns from pcre_exec() were being treated as failures
1867            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1868            error message is output. Some extra information is given for the
1869            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1870            probably the only errors that are likely to be caused by users (by
1871            specifying a regex that has nested indefinite repeats, for instance).
1872            If there are more than 20 of these errors, pcregrep is abandoned.
1873    
1874        (b) A binary zero was treated as data while matching, but terminated the
1875            output line if it was written out. This has been fixed: binary zeroes
1876            are now no different to any other data bytes.
1877    
1878        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1879            used to set a locale for matching. The --locale=xxxx long option has
1880            been added (no short equivalent) to specify a locale explicitly on the
1881            pcregrep command, overriding the environment variables.
1882    
1883        (d) When -B was used with -n, some line numbers in the output were one less
1884            than they should have been.
1885    
1886        (e) Added the -o (--only-matching) option.
1887    
1888        (f) If -A or -C was used with -c (count only), some lines of context were
1889            accidentally printed for the final match.
1890    
1891        (g) Added the -H (--with-filename) option.
1892    
1893        (h) The combination of options -rh failed to suppress file names for files
1894            that were found from directory arguments.
1895    
1896        (i) Added the -D (--devices) and -d (--directories) options.
1897    
1898        (j) Added the -F (--fixed-strings) option.
1899    
1900        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1901    
1902        (l) Added the --colo(u)r option.
1903    
1904        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1905            is not present by default.
1906    
1907     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1908        items such as (?R) or (?1), when the recursion could match a number of
1909        alternatives. If it matched one of the alternatives, but subsequently,
1910        outside the recursion, there was a failure, the code tried to back up into
1911        the recursion. However, because of the way PCRE is implemented, this is not
1912        possible, and the result was an incorrect result from the match.
1913    
1914        In order to prevent this happening, the specification of recursion has
1915        been changed so that all such subpatterns are automatically treated as
1916        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1917    
1918     4. I had overlooked the fact that, in some locales, there are characters for
1919        which isalpha() is true but neither isupper() nor islower() are true. In
1920        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1921        and ordfeminine) are like this. This affected the treatment of \w and \W
1922        when they appeared in character classes, but not when they appeared outside
1923        a character class. The bit map for "word" characters is now created
1924        separately from the results of isalnum() instead of just taking it from the
1925        upper, lower, and digit maps. (Plus the underscore character, of course.)
1926    
1927     5. The above bug also affected the handling of POSIX character classes such as
1928        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1929        permanent tables. Instead, the bit maps for such a class were previously
1930        created as the appropriate unions of the upper, lower, and digit bitmaps.
1931        Now they are created by subtraction from the [[:word:]] class, which has
1932        its own bitmap.
1933    
1934     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1935        It is created by subtracting the vertical space characters (\x09, \x0a,
1936        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1937        subtraction was done in the overall bitmap for a character class, meaning
1938        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1939        be recognized. This bug has been fixed.
1940    
1941     7. Patches from the folks at Google:
1942    
1943          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1944          real life, but is still worth protecting against".
1945    
1946          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1947          regular expressions".
1948    
1949          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1950          have it.
1951    
1952          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1953          "configure" and the latter not, in order to fix a problem somebody had
1954          with compiling the Arg class on HP-UX.
1955    
1956          (e) Improve the error-handling of the C++ wrapper a little bit.
1957    
1958          (f) New tests for checking recursion limiting.
1959    
1960     8. The pcre_memmove() function, which is used only if the environment does not
1961        have a standard memmove() function (and is therefore rarely compiled),
1962        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1963        returning a result (though PCRE never actually uses the result).
1964    
1965     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1966        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1967        returned instead of calling malloc() with an overflowing number that would
1968        most likely cause subsequent chaos.
1969    
1970    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1971    
1972    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1973        with this option is matched, the nmatch and pmatch options of regexec() are
1974        ignored.
1975    
1976    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1977        provided in case anyone wants to the the POSIX interface with UTF-8
1978        strings.
1979    
1980    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1981        C++ linking (needed for some HP-UX environments).
1982    
1983    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1984        (unused parameter) and in the pcre_printint() function (omitted "default"
1985        switch label when the default is to do nothing).
1986    
1987    15. Added some code to make it possible, when PCRE is compiled as a C++
1988        library, to replace subject pointers for pcre_exec() with a smart pointer
1989        class, thus making it possible to process discontinuous strings.
1990    
1991    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1992        much the same function. They were added by different people who were trying
1993        to make PCRE easy to compile on non-Unix systems. It has been suggested
1994        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1995        for compiling on Windows systems. I have therefore replaced it with
1996        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1997        defaults to "extern" for C or "extern C" for C++, which works fine on
1998        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1999        SCOPE with something explicit in config.h. In addition:
2000    
2001        (a) pcreposix.h still had just "extern" instead of either of these macros;
2002            I have replaced it with PCRE_DATA_SCOPE.
2003    
2004        (b) Functions such as _pcre_xclass(), which are internal to the library,
2005            but external in the C sense, all had PCRE_EXPORT in their definitions.
2006            This is apparently wrong for the Windows case, so I have removed it.
2007            (It makes no difference on Unix-like systems.)
2008    
2009    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2010        of recursive calls to match(). This is different to MATCH_LIMIT because
2011        that limits the total number of calls to match(), not all of which increase
2012        the depth of recursion. Limiting the recursion depth limits the amount of
2013        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2014        when PCRE is compiled, and changed at run time. A patch from Google adds
2015        this functionality to the C++ interface.
2016    
2017    18. Changes to the handling of Unicode character properties:
2018    
2019        (a) Updated the table to Unicode 4.1.0.
2020    
2021        (b) Recognize characters that are not in the table as "Cn" (undefined).
2022    
2023        (c) I revised the way the table is implemented to a much improved format
2024            which includes recognition of ranges. It now supports the ranges that
2025            are defined in UnicodeData.txt, and it also amalgamates other
2026            characters into ranges. This has reduced the number of entries in the
2027            table from around 16,000 to around 3,000, thus reducing its size
2028            considerably. I realized I did not need to use a tree structure after
2029            all - a binary chop search is just as efficient. Having reduced the
2030            number of entries, I extended their size from 6 bytes to 8 bytes to
2031            allow for more data.
2032    
2033        (d) Added support for Unicode script names via properties such as \p{Han}.
2034    
2035    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2036        matching that character.
2037    
2038    20. When matching a repeated Unicode property with a minimum greater than zero,
2039        (for example \pL{2,}), PCRE could look past the end of the subject if it
2040        reached it while seeking the minimum number of characters. This could
2041        happen only if some of the characters were more than one byte long, because
2042        there is a check for at least the minimum number of bytes.
2043    
2044    21. Refactored the implementation of \p and \P so as to be more general, to
2045        allow for more different types of property in future. This has changed the
2046        compiled form incompatibly. Anybody with saved compiled patterns that use
2047        \p or \P will have to recompile them.
2048    
2049    22. Added "Any" and "L&" to the supported property types.
2050    
2051    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2052        but give a compile time error if the value is greater than 0xff.
2053    
2054    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2055        accidentally not being installed or uninstalled.
2056    
2057    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2058        made were to insert the current release number. This seemed silly, because
2059        it made things harder for people building PCRE on systems that don't run
2060        "configure". I have turned pcre.h into a distributed file, no longer built
2061        by "configure", with the version identification directly included. There is
2062        no longer a pcre.h.in file.
2063    
2064        However, this change necessitated a change to the pcre-config script as
2065        well. It is built from pcre-config.in, and one of the substitutions was the
2066        release number. I have updated configure.ac so that ./configure now finds
2067        the release number by grepping pcre.h.
2068    
2069    26. Added the ability to run the tests under valgrind.
2070    
2071    
2072    Version 6.4 05-Sep-05
2073    ---------------------
2074    
2075     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2076        "--" to be printed when multiple files were scanned, even when none of the
2077        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2078        consider it to be a bug, and have restored the previous behaviour.
2079    
2080     2. A couple of code tidies to get rid of compiler warnings.
2081    
2082     3. The pcretest program used to cheat by referring to symbols in the library
2083        whose names begin with _pcre_. These are internal symbols that are not
2084        really supposed to be visible externally, and in some environments it is
2085        possible to suppress them. The cheating is now confined to including
2086        certain files from the library's source, which is a bit cleaner.
2087    
2088     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2089        file's purpose clearer.
2090    
2091     5. Reorganized pcre_ucp_findchar().
2092    
2093    
2094    Version 6.3 15-Aug-05
2095    ---------------------
2096    
2097     1. The file libpcre.pc.in did not have general read permission in the tarball.
2098    
2099     2. There were some problems when building without C++ support:
2100    
2101        (a) If C++ support was not built, "make install" and "make test" still
2102            tried to test it.
2103    
2104        (b) There were problems when the value of CXX was explicitly set. Some
2105            changes have been made to try to fix these, and ...
2106    
2107        (c) --disable-cpp can now be used to explicitly disable C++ support.
2108    
2109        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2110            backslash in a target when C++ was disabled. This confuses some
2111            versions of "make", apparently. Using an intermediate variable solves
2112            this. (Same for CPP_LOBJ.)
2113    
2114     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2115        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2116        necessary on certain architectures.
2117    
2118     4. Added a setting of -export-symbols-regex to the link command to remove
2119        those symbols that are exported in the C sense, but actually are local
2120        within the library, and not documented. Their names all begin with
2121        "_pcre_". This is not a perfect job, because (a) we have to except some
2122        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2123        available (and never for static libraries). I have made a note to try to
2124        find a way round (a) in the future.
2125    
2126    
2127    Version 6.2 01-Aug-05
2128    ---------------------
2129    
2130     1. There was no test for integer overflow of quantifier values. A construction
2131        such as {1111111111111111} would give undefined results. What is worse, if
2132        a minimum quantifier for a parenthesized subpattern overflowed and became
2133        negative, the calculation of the memory size went wrong. This could have
2134        led to memory overwriting.
2135    
2136     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2137    
2138     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2139        operating environments where this matters.
2140    
2141     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2142        PCRE options from within the C++ wrapper.
2143    
2144     5. Named capturing subpatterns were not being correctly counted when a pattern
2145        was compiled. This caused two problems: (a) If there were more than 100
2146        such subpatterns, the calculation of the memory needed for the whole
2147        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2148        back references of the form \12, where the number was greater than 9, were
2149        not recognized as back references, even though there were sufficient
2150        previous subpatterns.
2151    
2152     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2153        versions of gcc, e.g. 2.95.4.
2154    
2155    
2156    Version 6.1 21-Jun-05
2157    ---------------------
2158    
2159     1. There was one reference to the variable "posix" in pcretest.c that was not
2160        surrounded by "#if !defined NOPOSIX".
2161    
2162     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2163        the cross-check on the old pcre_info() function, for the benefit of the
2164        cut-down version of PCRE that is currently imported into Exim.
2165    
2166     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2167        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2168        patterns that start (?i) but I don't think that matters. The use of (?i) is
2169        just an example; this all applies to the other options as well.
2170    
2171     4. Since libtool seems to echo the compile commands it is issuing, the output
2172        from "make" can be reduced a bit by putting "@" in front of each libtool
2173        compile command.
2174    
2175     5. Patch from the folks at Google for configure.in to be a bit more thorough
2176        in checking for a suitable C++ installation before trying to compile the
2177        C++ stuff. This should fix a reported problem when a compiler was present,
2178        but no suitable headers.
2179    
2180     6. The man pages all had just "PCRE" as their title. I have changed them to
2181        be the relevant file name. I have also arranged that these names are
2182        retained in the file doc/pcre.txt, which is a concatenation in text format
2183        of all the man pages except the little individual ones for each function.
2184    
2185     7. The NON-UNIX-USE file had not been updated for the different set of source
2186        files that come with release 6. I also added a few comments about the C++
2187        wrapper.
2188    
2189    
2190    Version 6.0 07-Jun-05
2191    ---------------------
2192    
2193     1. Some minor internal re-organization to help with my DFA experiments.
2194    
2195     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2196        didn't matter for the library itself when fully configured, but did matter
2197        when compiling without UCP support, or within Exim, where the ucp files are
2198        not imported.
2199    
2200     3. Refactoring of the library code to split up the various functions into
2201        different source modules. The addition of the new DFA matching code (see
2202        below) to a single monolithic source would have made it really too
2203        unwieldy, quite apart from causing all the code to be include in a
2204        statically linked application, when only some functions are used. This is
2205        relevant even without the DFA addition now that patterns can be compiled in
2206        one application and matched in another.
2207    
2208        The downside of splitting up is that there have to be some external
2209        functions and data tables that are used internally in different modules of
2210        the library but which are not part of the API. These have all had their
2211        names changed to start with "_pcre_" so that they are unlikely to clash
2212        with other external names.
2213    
2214     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2215        a different (DFA) algorithm. Although it is slower than the original
2216        function, it does have some advantages for certain types of matching
2217        problem.
2218    
2219     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2220        including restarting after a partial match.
2221    
2222     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2223        defined when compiling for Windows was sent to me. I have put it into the
2224        code, though I have no means of testing or verifying it.
2225    
2226     7. Added the pcre_refcount() auxiliary function.
2227    
2228     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2229        match before or at the first newline in the subject string. In pcretest,
2230        the /f option on a pattern can be used to set this.
2231    
2232     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2233        would behave wrongly. This has been present in PCRE since release 4.0.
2234    
2235    10. A number of changes to the pcregrep command:
2236    
2237        (a) Refactored how -x works; insert ^(...)$ instead of setting
2238            PCRE_ANCHORED and checking the length, in preparation for adding
2239            something similar for -w.
2240    
2241        (b) Added the -w (match as a word) option.
2242    
2243        (c) Refactored the way lines are read and buffered so as to have more
2244            than one at a time available.
2245    
2246        (d) Implemented a pcregrep test script.
2247    
2248        (e) Added the -M (multiline match) option. This allows patterns to match
2249            over several lines of the subject. The buffering ensures that at least
2250            8K, or the rest of the document (whichever is the shorter) is available
2251            for matching (and similarly the previous 8K for lookbehind assertions).
2252    
2253        (f) Changed the --help output so that it now says
2254    
2255              -w, --word-regex(p)
2256    
2257            instead of two lines, one with "regex" and the other with "regexp"
2258            because that confused at least one person since the short forms are the
2259            same. (This required a bit of code, as the output is generated
2260            automatically from a table. It wasn't just a text change.)
2261    
2262        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2263            option but starts with a hyphen. Could be a pattern or a path name
2264            starting with a hyphen, for instance.
2265    
2266        (h) "-" can be given as a file name to represent stdin.
2267    
2268        (i) When file names are being printed, "(standard input)" is used for
2269            the standard input, for compatibility with GNU grep. Previously
2270            "<stdin>" was used.
2271    
2272        (j) The option --label=xxx can be used to supply a name to be used for
2273            stdin when file names are being printed. There is no short form.
2274    
2275        (k) Re-factored the options decoding logic because we are going to add
2276            two more options that take data. Such options can now be given in four
2277            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2278    
2279        (l) Added the -A, -B, and -C options for requesting that lines of context
2280            around matches be printed.
2281    
2282        (m) Added the -L option to print the names of files that do not contain
2283            any matching lines, that is, the complement of -l.
2284    
2285        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2286            continue to scan other files.
2287    
2288        (o) The -s option was incorrectly implemented. For compatibility with other
2289            greps, it now suppresses the error message for a non-existent or non-
2290            accessible file (but not the return code). There is a new option called
2291            -q that suppresses the output of matching lines, which was what -s was
2292            previously doing.
2293    
2294        (p) Added --include and --exclude options to specify files for inclusion
2295            and exclusion when recursing.
2296    
2297    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2298        Hopefully, it now does.
2299    
2300    12. Missing cast in pcre_study().
2301    
2302    13. Added an "uninstall" target to the makefile.
2303    
2304    14. Replaced "extern" in the function prototypes in Makefile.in with
2305        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2306        world, but is set differently for Windows.
2307    
2308    15. Added a second compiling function called pcre_compile2(). The only
2309        difference is that it has an extra argument, which is a pointer to an
2310        integer error code. When there is a compile-time failure, this is set
2311        non-zero, in addition to the error test pointer being set to point to an
2312        error message. The new argument may be NULL if no error number is required
2313        (but then you may as well call pcre_compile(), which is now just a
2314        wrapper). This facility is provided because some applications need a
2315        numeric error indication, but it has also enabled me to tidy up the way
2316        compile-time errors are handled in the POSIX wrapper.
2317    
2318    16. Added VPATH=.libs to the makefile; this should help when building with one
2319        prefix path and installing with another. (Or so I'm told by someone who
2320        knows more about this stuff than I do.)
2321    
2322    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2323        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2324        match everything, including newlines. This is not POSIX-compatible, but
2325        somebody wanted the feature. From pcretest it can be activated by using
2326        both the P and the s flags.
2327    
2328    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2329    
2330    19. libpcre.pc was being incorrectly installed as executable.
2331    
2332    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2333        it now also looks for '\r' so that it will work unmodified on Windows.
2334    
2335    21. Added Google's contributed C++ wrapper to the distribution.
2336    
2337    22. Added some untidy missing memory free() calls in pcretest, to keep
2338        Electric Fence happy when testing.
2339    
2340    
2341    
2342    Version 5.0 13-Sep-04
2343    ---------------------
2344    
2345     1. Internal change: literal characters are no longer packed up into items
2346        containing multiple characters in a single byte-string. Each character
2347        is now matched using a separate opcode. However, there may be more than one
2348        byte in the character in UTF-8 mode.
2349    
2350     2. The pcre_callout_block structure has two new fields: pattern_position and
2351        next_item_length. These contain the offset in the pattern to the next match
2352        item, and its length, respectively.
2353    
2354     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2355        insertion of callouts before each pattern item. Added the /C option to
2356        pcretest to make use of this.
2357    
2358     4. On the advice of a Windows user, the lines
2359    
2360          #if defined(_WIN32) || defined(WIN32)
2361          _setmode( _fileno( stdout ), 0x8000 );
2362          #endif  /* defined(_WIN32) || defined(WIN32) */
2363    
2364        have been added to the source of pcretest. This apparently does useful
2365        magic in relation to line terminators.
2366    
2367     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2368        for the benefit of those environments where the "b" makes a difference.
2369    
2370     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2371        to know about it. I have put a hack into configure.in that adds in code
2372        to set GCC=yes if CC=icc. This seems to end up at a point in the
2373        generated configure script that is early enough to affect the setting of
2374        compiler options, which is what is needed, but I have no means of testing
2375        whether it really works. (The user who reported this had patched the
2376        generated configure script, which of course I cannot do.)
2377    
2378        LATER: After change 22 below (new libtool files), the configure script
2379        seems to know about icc (and also ecc). Therefore, I have commented out
2380        this hack in configure.in.
2381    
2382     7. Added support for pkg-config (2 patches were sent in).
2383    
2384     8. Negated POSIX character classes that used a combination of internal tables
2385        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2386        [[:^ascii]]. Typically, they would match almost any characters. The other
2387        POSIX classes were not broken in this way.
2388    
2389     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2390        to find the match, as PCRE was deluded into thinking that the match had to
2391        start at the start point or following a newline. The same bug applied to
2392        patterns with negative forward assertions or any backward assertions
2393        preceding ".*" at the start, unless the pattern required a fixed first
2394        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2395    
2396    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2397        starting at the last subject character, bytes beyond the end of the subject
2398        string were read.
2399    
2400    11. Renamed the variable "class" as "classbits" to make life easier for C++
2401        users. (Previously there was a macro definition, but it apparently wasn't
2402        enough.)
2403    
2404    12. Added the new field "tables" to the extra data so that tables can be passed
2405        in at exec time, or the internal tables can be re-selected. This allows
2406        a compiled regex to be saved and re-used at a later time by a different
2407        program that might have everything at different addresses.
2408    
2409    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2410        -R library as well as a -L library.
2411    
2412    14. The debugging options of pcretest (-d on the command line or D on a
2413        pattern) showed incorrect output for anything following an extended class
2414        that contained multibyte characters and which was followed by a quantifier.
2415    
2416    15. Added optional support for general category Unicode character properties
2417        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2418        support. It adds about 90K to the size of the library. The meanings of the
2419        inbuilt class escapes such as \d and \s have NOT been changed.
2420    
2421    16. Updated pcredemo.c to include calls to free() to release the memory for the
2422        compiled pattern.
2423    
2424    17. The generated file chartables.c was being created in the source directory
2425        instead of in the building directory. This caused the build to fail if the
2426        source directory was different from the building directory, and was
2427        read-only.
2428    
2429    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2430        file. No doubt somebody will tell me if they don't make sense... Also added
2431        Dan Mooney's comments about building on OpenVMS.
2432    
2433    19. Added support for partial matching via the PCRE_PARTIAL option for
2434        pcre_exec() and the \P data escape in pcretest.
2435    
2436    20. Extended pcretest with 3 new pattern features:
2437    
2438        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2439              write the compiled pattern to the file whose name is "rest-of-line".
2440              This is a straight binary dump of the data, with the saved pointer to
2441              the character tables forced to be NULL. The study data, if any, is
2442              written too. After writing, pcretest reads a new pattern.
2443    
2444        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2445              compiled pattern from the given file. There must not be any
2446              occurrences of "<" in the file name (pretty unlikely); if there are,
2447              pcretest will instead treat the initial "<" as a pattern delimiter.
2448              After reading in the pattern, pcretest goes on to read data lines as
2449              usual.
2450    
2451        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2452              and 16-bit fields in a compiled pattern, to simulate a pattern that
2453              was compiled on a host of opposite endianness.
2454    
2455    21. The pcre-exec() function can now cope with patterns that were compiled on
2456        hosts of opposite endianness, with this restriction:
2457    
2458          As for any compiled expression that is saved and used later, the tables
2459          pointer field cannot be preserved; the extra_data field in the arguments
2460          to pcre_exec() should be used to pass in a tables address if a value
2461          other than the default internal tables were used at compile time.
2462    
2463    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2464        now diagnosed as an error. Previously, most of the time, a negative number
2465        would have been treated as zero, but if in addition "ovector" was passed as
2466        NULL, a crash could occur.
2467    
2468    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2469        new versions from the libtool 1.5 distribution (the last one is a copy of
2470        a file called libtool.m4). This seems to have fixed the need to patch
2471        "configure" to support Darwin 1.3 (which I used to do). However, I still
2472        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2473        workstation).
2474    
2475    24. Changed the PCRE licence to be the more standard "BSD" licence.
2476    
2477    
2478    Version 4.5 01-Dec-03
2479    ---------------------
2480    
2481     1. There has been some re-arrangement of the code for the match() function so
2482        that it can be compiled in a version that does not call itself recursively.
2483        Instead, it keeps those local variables that need separate instances for
2484        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2485        needs to "recurse". Keeping track of where control must go is done by means
2486        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2487        hide most of the details from the main code, and operates only if
2488        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2489        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2490        operating.
2491    
2492        To make it easier for callers to provide specially tailored get/free
2493        functions for this usage, two new functions, pcre_stack_malloc, and
2494        pcre_stack_free, are used. They are always called in strict stacking order,
2495        and the size of block requested is always the same.
2496    
2497        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2498        PCRE has been compiled to use the stack or the heap for recursion. The
2499        -C option of pcretest uses this to show which version is compiled.
2500    
2501        A new data escape \S, is added to pcretest; it causes the amounts of store
2502        obtained and freed by both kinds of malloc/free at match time to be added
2503        to the output.
2504    
2505     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2506        what's available on my current Linux desktop machine.
2507    
2508     3. When matching a UTF-8 string, the test for a valid string at the start has
2509        been extended. If start_offset is not zero, PCRE now checks that it points
2510        to a byte that is the start of a UTF-8 character. If not, it returns
2511        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2512        this is necessary because there may be backward assertions in the pattern.
2513        When matching the same subject several times, it may save resources to use
2514        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2515    
2516     4. The code for checking the validity of UTF-8 strings has been tightened so
2517        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2518        containing "overlong sequences".
2519    
2520     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2521        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2522        should have been "&", but it just so happened that all the cases this let
2523        through by mistake were picked up later in the function.
2524    
2525     6. I had used a variable called "isblank" - this is a C99 function, causing
2526        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2527    
2528     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2529        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2530        so the version is shown; (c) stop "make test" echoing "./RunTest".
2531    
2532     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2533    
2534     9. The prototype for memmove() for systems that don't have it was using
2535        size_t, but the inclusion of the header that defines size_t was later. I've
2536        moved the #includes for the C headers earlier to avoid this.
2537    
2538    10. Added some adjustments to the code to make it easier to compiler on certain
2539        special systems:
2540    
2541          (a) Some "const" qualifiers were missing.
2542          (b) Added the macro EXPORT before all exported functions; by default this
2543              is defined to be empty.
2544          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2545              that it reads its output file name as an argument instead of writing
2546              to the standard output and assuming this can be redirected.
2547    
2548    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2549        class containing characters with values greater than 255, PCRE compilation
2550        went into a loop.
2551    
2552    12. A recursive reference to a subpattern that was within another subpattern
2553        that had a minimum quantifier of zero caused PCRE to crash. For example,
2554        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2555        recursion. If the recursively-called subpattern itself had a zero repeat,
2556        that was OK.
2557    
2558    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2559        buffer into which it was copied (for escape processing) was still set at
2560        1024, so long lines caused crashes.
2561    
2562    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2563        "internal error: code overflow...". This applied to any character class
2564        that was followed by a possessive quantifier.
2565    
2566    15. Modified the Makefile to add libpcre.la as a prerequisite for
2567        libpcreposix.la because I was told this is needed for a parallel build to
2568        work.
2569    
2570    16. If a pattern that contained .* following optional items at the start was
2571        studied, the wrong optimizing data was generated, leading to matching
2572        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2573        matching string must start with a or b or c. The correct conclusion for
2574        this pattern is that a match can start with any character.
2575    
2576    
2577    Version 4.4 13-Aug-03
2578    ---------------------
2579    
2580     1. In UTF-8 mode, a character class containing characters with values between
2581        127 and 255 was not handled correctly if the compiled pattern was studied.
2582        In fixing this, I have also improved the studying algorithm for such
2583        classes (slightly).
2584    
2585     2. Three internal functions had redundant arguments passed to them. Removal
2586        might give a very teeny performance improvement.
2587    
2588     3. Documentation bug: the value of the capture_top field in a callout is *one
2589        more than* the number of the hightest numbered captured substring.
2590    
2591     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2592        in incorrectly linking with a previously installed version. They now link
2593        explicitly with libpcre.la.
2594    
2595     5. configure.in no longer needs to recognize Cygwin specially.
2596    
2597     6. A problem in pcre.in for Windows platforms is fixed.
2598    
2599     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2600        pcretest, it used to include the size of the study block as part of its
2601        output. Unfortunately, the structure contains a field that has a different
2602        size on different hardware architectures. This meant that the tests that
2603        showed this size failed. As the block is currently always of a fixed size,
2604        this information isn't actually particularly useful in pcretest output, so
2605        I have just removed it.
2606    
2607     8. Three pre-processor statements accidentally did not start in column 1.
2608        Sadly, there are *still* compilers around that complain, even though
2609        standard C has not required this for well over a decade. Sigh.
2610    
2611     9. In pcretest, the code for checking callouts passed small integers in the
2612        callout_data field, which is a void * field. However, some picky compilers
2613        complained about the casts involved for this on 64-bit systems. Now
2614        pcretest passes the address of the small integer instead, which should get
2615        rid of the warnings.
2616    
2617    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2618        both compile and run time, and gives an error if an invalid UTF-8 sequence
2619        is found. There is a option for disabling this check in cases where the
2620        string is known to be correct and/or the maximum performance is wanted.
2621    
2622    11. In response to a bug report, I changed one line in Makefile.in from
2623    
2624            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2625        to
2626            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2627    
2628        to look similar to other lines, but I have no way of telling whether this
2629        is the right thing to do, as I do not use Windows. No doubt I'll get told
2630        if it's wrong...
2631    
2632    
2633    Version 4.3 21-May-03
2634    ---------------------
2635    
2636    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2637       Makefile.
2638    
2639    2. Some refactoring to improve the quality of the code:
2640    
2641       (i)   The utf8_table... variables are now declared "const".
2642    
2643       (ii)  The code for \cx, which used the "case flipping" table to upper case
2644             lower case letters, now just substracts 32. This is ASCII-specific,
2645             but the whole concept of \cx is ASCII-specific, so it seems
2646             reasonable.
2647    
2648       (iii) PCRE was using its character types table to recognize decimal and
2649             hexadecimal digits in the pattern. This is silly, because it handles
2650             only 0-9, a-f, and A-F, but the character types table is locale-
2651             specific, which means strange things might happen. A private
2652             table is now used for this - though it costs 256 bytes, a table is
2653             much faster than multiple explicit tests. Of course, the standard
2654             character types table is still used for matching digits in subject
2655             strings against \d.
2656    
2657       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2658             ending in _t are). So I've renamed it as ESC_tee.
2659    
2660    3. The first argument for regexec() in the POSIX wrapper should have been
2661       defined as "const".
2662    
2663    4. Changed pcretest to use malloc() for its buffers so that they can be
2664       Electric Fenced for debugging.
2665    
2666    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2667       to read one or more bytes before the start of the subject string. Often this
2668       had no effect on PCRE's behaviour, but in some circumstances it could
2669       provoke a segmentation fault.
2670    
2671    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2672       to try to read one or more bytes before the start of the subject string.
2673    
2674    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2675       UTF-8 support could misbehave in various ways if the subject string
2676       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2677       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2678       back over UTF-8 characters.)
2679    
2680    
2681    Version 4.2 14-Apr-03
2682    ---------------------
2683    
2684    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2685    
2686    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2687         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2688         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2689         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2690         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2691           and BUILD_EXEEXT
2692         Note: automatic setting of the BUILD variables is not yet working
2693         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2694           compile-time but not at link-time
2695         [LINK]: use for linking executables only
2696         make different versions for Windows and non-Windows
2697         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2698           libraries
2699         [LINK_FOR_BUILD]: new variable
2700         [OBJEXT]: use throughout
2701         [EXEEXT]: use throughout
2702         <winshared>: new target
2703         <wininstall>: new target
2704         <dftables.o>: use native compiler
2705         <dftables>: use native linker
2706         <install>: handle Windows platform correctly
2707         <clean>: ditto
2708         <check>: ditto
2709         copy DLL to top builddir before testing
2710    
2711       As part of these changes, -no-undefined was removed again. This was reported
2712       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2713       in any case.
2714    
2715    3. Some tidies to get rid of compiler warnings:
2716    
2717       . In the match_data structure, match_limit was an unsigned long int, whereas
2718         match_call_count was an int. I've made them both unsigned long ints.
2719    
2720       . In pcretest the fact that a const uschar * doesn't automatically cast to
2721         a void * provoked a warning.
2722    
2723       . Turning on some more compiler warnings threw up some "shadow" variables
2724         and a few more missing casts.
2725    
2726    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2727       option, a class that contained a single character with a value between 128
2728       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2729    
2730    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2731       option, a class that contained several characters, but with at least one
2732       whose value was between 128 and 255 caused PCRE to crash.
2733    
2734    
2735    Version 4.1 12-Mar-03
2736    ---------------------
2737    
2738    1. Compiling with gcc -pedantic found a couple of places where casts were
2739    needed, and a string in dftables.c that was longer than standard compilers are
2740    required to support.
2741    
2742    2. Compiling with Sun's compiler found a few more places where the code could
2743    be tidied up in order to avoid warnings.
2744    
2745    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2746    first of these names is deprecated in the latest Autoconf in favour of the name
2747    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2748    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2749    analogy I have changed it to CFLAGS_FOR_BUILD.
2750    
2751    4. Added -no-undefined to the linking command in the Makefile, because this is
2752    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2753    linking step for the pcreposix library.
2754    
2755    5. PCRE was failing to diagnose the case of two named groups with the same
2756    name.
2757    
2758    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2759    literal character that is needed in the subject for a match, and scans along to
2760    ensure that it is present before embarking on the full matching process. This
2761    saves time in cases of nested unlimited repeats that are never going to match.
2762    Problem: the scan can take a lot of time if the subject is very long (e.g.
2763    megabytes), thus penalizing straightforward matches. It is now done only if the
2764    amount of subject to be scanned is less than 1000 bytes.
2765    
2766    7. A lesser problem with the same optimization is that it was recording the
2767    first character of an anchored pattern as "needed", thus provoking a search
2768    right along the subject, even when the first match of the pattern was going to
2769    fail. The "needed" character is now not set for anchored patterns, unless it
2770    follows something in the pattern that is of non-fixed length. Thus, it still
2771    fulfils its original purpose of finding quick non-matches in cases of nested
2772    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2773    
2774    
2775    Version 4.0 17-Feb-03
2776    ---------------------
2777    
2778    1. If a comment in an extended regex that started immediately after a meta-item
2779    extended to the end of string, PCRE compiled incorrect data. This could lead to
2780    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2781    
2782    2. Moved to autoconf 2.53 and libtool 1.4.2.
2783    
2784    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2785    the special perltest8 script is no longer needed - all the tests can be run
2786    from a single perltest script.
2787    
2788    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2789    by \s. It has now been removed in PCRE. This means it isn't recognized as
2790    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2791    class [:space:] *does* include VT, thereby creating a mess.
2792    
2793    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2794    space and tab.
2795    
2796    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2797    its new features into the main test script, reducing the number of scripts.
2798    
2799    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2800    were backward compatible, and made the (?i) apply to the whole pattern, as if
2801    /i were given. Now it behaves more logically, and applies the option setting
2802    only to what follows. PCRE has been changed to follow suit. However, if it
2803    finds options settings right at the start of the pattern, it extracts them into
2804    the global options, as before. Thus, they show up in the info data.
2805    
2806    8. Added support for the \Q...\E escape sequence. Characters in between are
2807    treated as literals. This is slightly different from Perl in that $ and @ are
2808    also handled as literals inside the quotes. In Perl, they will cause variable
2809    interpolation. Note the following examples:
2810    
2811        Pattern            PCRE matches      Perl matches
2812    
2813        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2814        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2815        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2816    
2817    For compatibility with Perl, \Q...\E sequences are recognized inside character
2818    classes as well as outside them.
2819    
2820    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2821    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2822    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2823    signed/unsigned warnings.
2824    
2825    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2826    option for pcretest, so I've replaced it by a simple function that does just
2827    that job.
2828    
2829    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2830    "pcregrep -".
2831    
2832    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2833    Java package. This provides some syntactic sugar for simple cases of what my
2834    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2835    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2836    item, you can use this simplified notation. Note that only makes sense with
2837    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2838    greediness, whatever the setting of the PCRE_UNGREEDY option.
2839    
2840    13. A change of greediness default within a pattern was not taking effect at
2841    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2842    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2843    was abstracted outside.
2844    
2845    14. PCRE now supports the \G assertion. It is true when the current matching
2846    position is at the start point of the match. This differs from \A when the
2847    starting offset is non-zero. Used with the /g option of pcretest (or similar
2848    code), it works in the same way as it does for Perl's /g option. If all
2849    alternatives of a regex begin with \G, the expression is anchored to the start
2850    match position, and the "anchored" flag is set in the compiled expression.
2851    
2852    15. Some bugs concerning the handling of certain option changes within patterns
2853    have been fixed. These applied to options other than (?ims). For example,
2854    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2855    been the other way round. Some of this was related to change 7 above.
2856    
2857    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2858    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2859    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2860    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2861    
2862    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2863    mode. Unlike ".", it always matches newline, whatever the setting of
2864    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2865    assertions. Perl allows it, but it doesn't (in general) work because it can't
2866    calculate the length of the lookbehind. At least, that's the case for Perl
2867    5.8.0 - I've been told they are going to document that it doesn't work in
2868    future.
2869    
2870    18. Added an error diagnosis for escapes that PCRE does not support: these are
2871    \L, \l, \N, \P, \p, \U, \u, and \X.
2872    
2873    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2874    reading past the end of the pattern in cases such as /[abcd/.
2875    
2876    20. PCRE was getting more memory than necessary for patterns with classes that
2877    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2878    
2879    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2880    compiling PCRE for use with Virtual Pascal.
2881    
2882    22. Small fix to the Makefile to make it work properly if the build is done
2883    outside the source tree.
2884    
2885    23. Added a new extension: a condition to go with recursion. If a conditional
2886    subpattern starts with (?(R) the "true" branch is used if recursion has
2887    happened, whereas the "false" branch is used only at the top level.
2888    
2889    24. When there was a very long string of literal characters (over 255 bytes
2890    without UTF support, over 250 bytes with UTF support), the computation of how
2891    much memory was required could be incorrect, leading to segfaults or other
2892    strange effects.
2893    
2894    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2895    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2896    there was a subsequent back reference to those brackets. This meant that, for
2897    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2898    possible to check for precisely this case. All we can do is abandon the
2899    optimization if .* occurs inside capturing brackets when there are any back
2900    references whatsoever. (See below for a better fix that came later.)
2901    
2902    26. The handling of the optimization for finding the first character of a
2903    non-anchored pattern, and for finding a character that is required later in the
2904    match were failing in some cases. This didn't break the matching; it just
2905    failed to optimize when it could. The way this is done has been re-implemented.
2906    
2907    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2908    
2909    28. Added a new feature that provides some of the functionality that Perl
2910    provides with (?{...}). The facility is termed a "callout". The way it is done
2911    in PCRE is for the caller to provide an optional function, by setting
2912    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2913    global variable. By default it is unset, which disables all calling out. To get
2914    the function called, the regex must include (?C) at appropriate points. This
2915    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2916    This provides a means of identifying different callout points. When PCRE
2917    reaches such a point in the regex, if pcre_callout has been set, the external
2918    function is called. It is provided with data in a structure called
2919    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2920    matching continues; if it returns a non-zero value, the match at the current
2921    point fails. However, backtracking will occur if possible. [This was changed
2922    later and other features added - see item 49 below.]
2923    
2924    29. pcretest is upgraded to test the callout functionality. It provides a
2925    callout function that displays information. By default, it shows the start of
2926    the match and the current position in the text. There are some new data escapes
2927    to vary what happens:
2928    
2929        \C+         in addition, show current contents of captured substrings
2930        \C-         do not supply a callout function
2931        \C!n        return 1 when callout number n is reached
2932        \C!n!m      return 1 when callout number n is reached for the mth time
2933    
2934    30. If pcregrep was called with the -l option and just a single file name, it
2935    output "<stdin>" if a match was found, instead of the file name.
2936    
2937    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2938    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2939    pcre_exec(). This saves a malloc/free per call. The default value of
2940    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2941    when configuring.
2942    
2943    32. The default maximum size of a compiled pattern is 64K. There have been a
2944    few cases of people hitting this limit. The code now uses macros to handle the
2945    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2946    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2947    configuring. Tests 2 and 5 work only with 2-byte links because they output
2948    debugging information about compiled patterns.
2949    
2950    33. Internal code re-arrangements:
2951    
2952    (a) Moved the debugging function for printing out a compiled regex into
2953        its own source file (printint.c) and used #include to pull it into
2954        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2955        separate copies.
2956    
2957    (b) Defined the list of op-code names for debugging as a macro in
2958        internal.h so that it is next to the definition of the opcodes.
2959    
2960    (c) Defined a table of op-code lengths for simpler skipping along compiled
2961        code. This is again a macro in internal.h so that it is next to the
2962        definition of the opcodes.
2963    
2964    34. Added support for recursive calls to individual subpatterns, along the
2965    lines of Robin Houston's patch (but implemented somewhat differently).
2966    
2967    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2968    allow it to read and process whole directories in Win32. This code was
2969    contributed by Lionel Fourquaux; it has not been tested by me.
2970    
2971    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2972    used to name a group. Names consist of alphanumerics and underscores, and must
2973    be unique. Back references use the syntax (?P=name) and recursive calls use
2974    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2975    numbers. The function pcre_fullinfo() can be used after compilation to extract
2976    a name/number map. There are three relevant calls:
2977    
2978      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2979      PCRE_INFO_NAMECOUNT            yields the number of entries
2980      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2981    
2982    The map is a vector of fixed-size entries. The size of each entry depends on
2983    the length of the longest name used. The first two bytes of each entry are the
2984    group number, most significant byte first. There follows the corresponding
2985    name, zero terminated. The names are in alphabetical order.
2986    
2987    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2988    case instead of 255. Making it the same both with and without UTF-8 support
2989    means that the same test output works with both.
2990    
2991    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2992    calling malloc() with a zero argument.
2993    
2994    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2995    optimization. I've improved things by keeping a bitmap of backreferences with
2996    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2997    fact referenced, the optimization can be applied. It is unlikely that a
2998    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2999    the match to follow \n) will appear inside brackets with a number greater than
3000    31, but if it does, any back reference > 31 suppresses the optimization.
3001    
3002    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3003    of disabling numbered capturing parentheses. Any opening parenthesis that is
3004    not followed by ? behaves as if it were followed by ?: but named parentheses
3005    can still be used for capturing (and they will acquire numbers in the usual
3006    way).
3007    
3008    41. Redesigned the return codes from the match() function into yes/no/error so
3009    that errors can be passed back from deep inside the nested calls. A malloc
3010    failure while inside a recursive subpattern call now causes the
3011    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3012    
3013    42. It is now possible to set a limit on the number of times the match()
3014    function is called in a call to pcre_exec(). This facility makes it possible to
3015    limit the amount of recursion and backtracking, though not in a directly
3016    obvious way, because the match() function is used in a number of different
3017    circumstances. The count starts from zero for each position in the subject
3018    string (for non-anchored patterns). The default limit is, for compatibility, a
3019    large number, namely 10 000 000. You can change this in two ways:
3020    
3021    (a) When configuring PCRE before making, you can use --with-match-limit=n
3022        to set a default value for the compiled library.
3023    
3024    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3025        a different value is set. See 45 below.
3026    
3027    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3028    
3029    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3030    of things that can be changed at compile time. The first argument specifies
3031    what is wanted and the second points to where the information is to be placed.
3032    The current list of available information is:
3033    
3034      PCRE_CONFIG_UTF8
3035    
3036    The output is an integer that is set to one if UTF-8 support is available;
3037    otherwise it is set to zero.
3038    
3039      PCRE_CONFIG_NEWLINE
3040    
3041    The output is an integer that it set to the value of the code that is used for
3042    newline. It is either LF (10) or CR (13).
3043    
3044      PCRE_CONFIG_LINK_SIZE
3045    
3046    The output is an integer that contains the number of bytes used for internal
3047    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3048    
3049      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3050    
3051    The output is an integer that contains the threshold above which the POSIX
3052    interface uses malloc() for output vectors. See item 31 above.
3053    
3054      PCRE_CONFIG_MATCH_LIMIT
3055    
3056    The output is an unsigned integer that contains the default limit of the number
3057    of match() calls in a pcre_exec() execution. See 42 above.
3058    
3059    44. pcretest has been upgraded by the addition of the -C option. This causes it
3060    to extract all the available output from the new pcre_config() function, and to
3061    output it. The program then exits immediately.
3062    
3063    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3064    order to support additional features. One way would have been to define
3065    pcre_exec2() (for example) with extra arguments, but this would not have been
3066    extensible, and would also have required all calls to the original function to
3067    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3068    is used for passing in "extra" data from pcre_study().
3069    
3070    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3071    contains the following fields:
3072    
3073      flags         a bitmap indicating which of the following fields are set
3074      study_data    opaque data from pcre_study()
3075      match_limit   a way of specifying a limit on match() calls for a specific
3076                      call to pcre_exec()
3077      callout_data  data for callouts (see 49 below)
3078    
3079    The flag bits are also defined in pcre.h, and are
3080    
3081      PCRE_EXTRA_STUDY_DATA
3082      PCRE_EXTRA_MATCH_LIMIT
3083      PCRE_EXTRA_CALLOUT_DATA
3084    
3085    The pcre_study() function now returns one of these new pcre_extra blocks, with
3086    the actual study data pointed to by the study_data field, and the
3087    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3088    before. That is, this change is entirely upwards-compatible and requires no
3089    change to existing code.
3090    
3091    If you want to pass in additional data to pcre_exec(), you can either place it
3092    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3093    block.
3094    
3095    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3096    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3097    times with different match limits, until it finds the minimum value needed for
3098    pcre_exec() to complete. The value is then output. This can be instructive; for
3099    most simple matches the number is quite small, but for pathological cases it
3100    gets very large very quickly.
3101    
3102    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3103    returns the size of the data block pointed to by the study_data field in a
3104    pcre_extra block, that is, the value that was passed as the argument to
3105    pcre_malloc() when PCRE was getting memory in which to place the information
3106    created by pcre_study(). The fourth argument should point to a size_t variable.
3107    pcretest has been extended so that this information is shown after a successful
3108    pcre_study() call when information about the compiled regex is being displayed.
3109    
3110    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3111    because what follows is always an absolute path. (Later: it turns out that this
3112    is more than cosmetic for MinGW, because it doesn't like empty path
3113    components.)
3114    
3115    49. Some changes have been made to the callout feature (see 28 above):
3116    
3117    (i)  A callout function now has three choices for what it returns:
3118    
3119           0  =>  success, carry on matching
3120         > 0  =>  failure at this point, but backtrack if possible
3121         < 0  =>  serious error, return this value from pcre_exec()
3122    
3123         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3124         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3125         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3126         use by callout functions. It will never be used by PCRE itself.
3127    
3128    (ii) The pcre_extra structure (see 45 above) has a void * field called
3129         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3130         pcre_callout_block structure has a field of the same name. The contents of
3131         the field passed in the pcre_extra structure are passed to the callout
3132         function in the corresponding field in the callout block. This makes it
3133         easier to use the same callout-containing regex from multiple threads. For
3134         testing, the pcretest program has a new data escape
3135    
3136           \C*n        pass the number n (may be negative) as callout_data
3137    
3138         If the callout function in pcretest receives a non-zero value as
3139         callout_data, it returns that value.
3140    
3141    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3142    there were some redundant $(CFLAGS) in commands that are now specified as
3143    $(LINK), which already includes $(CFLAGS).
3144    
3145    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3146    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3147    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3148    one-byte characters throughout. Note that case-insensitive matching applies
3149    only to characters whose values are less than 256. PCRE doesn't support the
3150    notion of cases for higher-valued characters.
3151    
3152    (i)   A character class whose characters are all within 0-255 is handled as
3153          a bit map, and the map is inverted for negative classes. Previously, a
3154          character > 255 always failed to match such a class; however it should
3155          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3156    
3157    (ii)  A negated character class with a single character < 255 is coded as
3158          "not this character" (OP_NOT). This wasn't working properly when the test
3159          character was multibyte, either singly or repeated.
3160    
3161    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3162          mode, for example: \x{100}{2,3}.
3163    
3164    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3165          singly or repeated) now correctly test multibyte characters. However,
3166          PCRE doesn't recognize any characters with values greater than 255 as
3167          digits, spaces, or word characters. Such characters always match \D, \S,
3168          and \W, and never match \d, \s, or \w.
3169    
3170    (v)   Classes may now contain characters and character ranges with values
3171          greater than 255. For example: [ab\x{100}-\x{400}].
3172    
3173    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3174          PCRE in UTF-8 mode.
3175    
3176    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3177    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3178    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3179    value.)
3180    
3181    53. The single man page has become too large. I have therefore split it up into
3182    a number of separate man pages. These also give rise to individual HTML pages;
3183    these are now put in a separate directory, and there is an index.html page that
3184    lists them all. Some hyperlinking between the pages has been installed.
3185    
3186    54. Added convenience functions for handling named capturing parentheses.
3187    
3188    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3189    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3190    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3191    are faulted.
3192    
3193    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3194    calling configure. These values are used when compiling the dftables.c program
3195    which is run to generate the source of the default character tables. They
3196    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3197    you will need to set these values.
3198    
3199    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3200    
3201    
3202    Version 3.9 02-Jan-02
3203    ---------------------
3204    
3205    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3206    
3207    2. If --disable-static was given, the building process failed when trying to
3208    build pcretest and pcregrep. (For some reason it was using libtool to compile
3209    them, which is not right, as they aren't part of the library.)
3210    
3211    
3212    Version 3.8 18-Dec-01
3213    ---------------------
3214    
3215    1. The experimental UTF-8 code was completely screwed up. It was packing the
3216    bytes in the wrong order. How dumb can you get?
3217    
3218    
3219    Version 3.7 29-Oct-01
3220    ---------------------
3221    
3222    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3223    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3224    this didn't happen under Solaris 8, where I normally test things.
3225    
3226    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3227    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3228    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3229    this fix an hour or so after the initial 3.7 release.)
3230    
3231    
3232    Version 3.6 23-Oct-01
3233    ---------------------
3234    
3235    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3236    offsets passed as NULL with zero offset count.
3237    
3238    2. The config.guess and config.sub files had not been updated when I moved to
3239    the latest autoconf.
3240    
3241    
3242    Version 3.5 15-Aug-01
3243    ---------------------
3244    
3245    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3246    had been forgotten.
3247    
3248    2. By using declared but undefined structures, we can avoid using "void"
3249    definitions in pcre.h while keeping the internal definitions of the structures
3250    private.
3251    
3252    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3253    user point of view, this means that both static and shared libraries are built
3254    by default, but this can be individually controlled. More of the work of
3255    handling this static/shared cases is now inside libtool instead of PCRE's make
3256    file.
3257    
3258    4. The pcretest utility is now installed along with pcregrep because it is
3259    useful for users (to test regexs) and by doing this, it automatically gets
3260    relinked by libtool. The documentation has been turned into a man page, so
3261    there are now .1, .txt, and .html versions in /doc.
3262    
3263    5. Upgrades to pcregrep:
3264       (i)   Added long-form option names like gnu grep.
3265       (ii)  Added --help to list all options with an explanatory phrase.
3266       (iii) Added -r, --recursive to recurse into sub-directories.
3267       (iv)  Added -f, --file to read patterns from a file.
3268    
3269    6. pcre_exec() was referring to its "code" argument before testing that
3270    argument for NULL (and giving an error if it was NULL).
3271    
3272    7. Upgraded Makefile.in to allow for compiling in a different directory from
3273    the source directory.
3274    
3275    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3276    options bits, the pointer it was passed was to an int instead of to an unsigned
3277    long int. This mattered only on 64-bit systems.
3278    
3279    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3280    generated) instead of pcre.in, which it its source. Also made the same change
3281    in several of the .c files.
3282    
3283    10. A new release of gcc defines printf() as a macro, which broke pcretest
3284    because it had an ifdef in the middle of a string argument for printf(). Fixed
3285    by using separate calls to printf().
3286    
3287    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3288    script, to force use of CR or LF instead of \n in the source. On non-Unix
3289    systems, the value can be set in config.h.
3290    
3291    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3292    absolute limit. Changed the text of the error message to make this clear, and
3293    likewise updated the man page.
3294    
3295    13. The limit of 99 on the number of capturing subpatterns has been removed.
3296    The new limit is 65535, which I hope will not be a "real" limit.
3297    
3298    
3299    Version 3.4 22-Aug-00
3300    ---------------------
3301    
3302    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3303    
3304    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3305    
3306    
3307    Version 3.3 01-Aug-00
3308    ---------------------
3309    
3310    1. If an octal character was given, but the value was greater than \377, it
3311    was not getting masked to the least significant bits, as documented. This could
3312    lead to crashes in some systems.
3313    
3314    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3315    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3316    
3317    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3318    These just pass their arguments on to (pcre_free)(), but they are provided
3319    because some uses of PCRE bind it to non-C systems that can call its functions,
3320    but cannot call free() or pcre_free() directly.
3321    
3322    4. Add "make test" as a synonym for "make check". Corrected some comments in
3323    the Makefile.
3324    
3325    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3326    Makefile.
3327    
3328    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3329    command called pgrep for grepping around the active processes.
3330    
3331    7. Added the beginnings of support for UTF-8 character strings.
3332    
3333    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3334    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3335    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3336    out for the ar command.)
3337    
3338    
3339    Version 3.2 12-May-00
3340    ---------------------
3341    
3342    This is purely a bug fixing release.
3343    
3344    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3345    of ZA. This was just one example of several cases that could provoke this bug,
3346    which was introduced by change 9 of version 2.00. The code for breaking
3347    infinite loops after an iteration that matches an empty string was't working
3348    correctly.
3349    
3350    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3351    when matched against abbab (for example). After matching an empty string, it
3352    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3353    caused it to match further down the string than it should.
3354    
3355    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3356    was there because it doesn't seem to be needed, and it causes trouble on some
3357    systems, as it is not a Standard C header. It has been removed.
3358    
3359    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3360    were reported on the Macintosh. The changes were from
3361    
3362      while ((c = *(++ptr)) != 0 && c != '\n');
3363    to
3364      while ((c = *(++ptr)) != 0 && c != '\n') ;
3365    
3366    Totally extraordinary, but if that's what it takes...
3367    
3368    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3369    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3370    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3371    assumes the way PCRE uses memmove() (always moving upwards).
3372    
3373    6. PCRE is being used in one environment where strchr() is not available. There
3374    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3375    faster code anyway.
3376    
3377    
3378    Version 3.1 09-Feb-00
3379    ---------------------
3380    
3381    The only change in this release is the fixing of some bugs in Makefile.in for
3382    the "install" target:
3383    
3384    (1) It was failing to install pcreposix.h.
3385    
3386    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3387    
3388    
3389    Version 3.0 01-Feb-00
3390    ---------------------
3391    
3392    1. Add support for the /+ modifier to perltest (to output $` like it does in
3393    pcretest).
3394    
3395    2. Add support for the /g modifier to perltest.
3396    
3397    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3398    matches null strings.
3399    
3400    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3401    pattern. Perl treats empty patterns specially - it reuses the most recent
3402    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3403    effect.
3404    
3405    5. The POSIX interface was broken in that it was just handing over the POSIX
3406    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3407    required a bigger vector, with some working space on the end. This means that
3408    the POSIX wrapper now has to get and free some memory, and copy the results.
3409    
3410    6. Added some simple autoconf support, placing the test data and the
3411    documentation in separate directories, re-organizing some of the
3412    information files, and making it build pcre-config (a GNU standard). Also added
3413    libtool support for building PCRE as a shared library, which is now the
3414    default.
3415    
3416    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3417    09 are not valid octal constants. Single digits will be used for minor values
3418    less than 10.
3419    
3420    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3421    existing programs that set these in the POSIX interface can use PCRE without
3422    modification.
3423    
3424    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3425    return all that pcre_info() returns, plus additional data. The pcre_info()
3426    function is retained for compatibility, but is considered to be obsolete.
3427    
3428    10. Added experimental recursion feature (?R) to handle one common case that
3429    Perl 5.6 will be able to do with (?p{...}).
3430    
3431    11. Added support for POSIX character classes like [:alpha:], which Perl is
3432    adopting.
3433    
3434    
3435    Version 2.08 31-Aug-99
3436    ----------------------
3437    
3438    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3439    trying to match at the startoffset position, but instead was moving forward to
3440    the next newline as if a previous match had failed.
3441    
3442    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3443    and could get into a loop if a null string was matched other than at the start
3444    of the subject.
3445    
3446    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3447    be distinguished at compile time, and for completeness also added PCRE_DATE.
3448    
3449    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3450    in GnuWin32 environments.
3451    
3452    
3453    Version 2.07 29-Jul-99
3454    ----------------------
3455    
3456    1. The documentation is now supplied in plain text form and HTML as well as in
3457    the form of man page sources.
3458    
3459    2. C++ compilers don't like assigning (void *) values to other pointer types.
3460    In particular this affects malloc(). Although there is no problem in Standard
3461    C, I've put in casts to keep C++ compilers happy.
3462    
3463    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3464    should be (const char *).
3465    
3466    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3467    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3468    However, I haven't made this a standard facility. The documentation doesn't
3469    mention it, and the Makefile doesn't support it.
3470    
3471    5. The Makefile now contains an "install" target, with editable destinations at
3472    the top of the file. The pcretest program is not installed.
3473    
3474    6. pgrep -V now gives the PCRE version number and date.
3475    
3476    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3477    causing the entire string to be ignored, instead of just the last character.
3478    
3479    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3480    non-matching string, it can take a very, very long time, even for strings of
3481    quite modest length, because of the nested recursion. PCRE now does better in
3482    some of these cases. It does this by remembering the last required literal
3483    character in the pattern, and pre-searching the subject to ensure it is present
3484    before running the real match. In other words, it applies a heuristic to detect
3485    some types of certain failure quickly, and in the above example, if presented
3486    with a string that has no trailing " it gives "no match" very quickly.
3487    
3488    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3489    other alternatives are tried instead.
3490    
3491    
3492    Version 2.06 09-Jun-99
3493    ----------------------
3494    
3495    1. Change pcretest's output for amount of store used to show just the code
3496    space, because the remainder (the data block) varies in size between 32-bit and
3497    64-bit systems.
3498    
3499    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3500    start matching at. This allows lookbehinds to work when searching for multiple
3501    occurrences in a string.
3502    
3503    3. Added additional options to pcretest for testing multiple occurrences:
3504    
3505       /+   outputs the rest of the string that follows a match
3506       /g   loops for multiple occurrences, using the new startoffset argument
3507       /G   loops for multiple occurrences by passing an incremented pointer
3508    
3509    4. PCRE wasn't doing the "first character" optimization for patterns starting
3510    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3511    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3512    the letter 'x'. On long subject strings, this gives a significant speed-up.
3513    
3514    
3515    Version 2.05 21-Apr-99
3516    ----------------------
3517    
3518    1. Changed the type of magic_number from int to long int so that it works
3519    properly on 16-bit systems.
3520    
3521    2. Fixed a bug which caused patterns starting with .* not to work correctly
3522    when the subject string contained newline characters. PCRE was assuming
3523    anchoring for such patterns in all cases, which is not correct because .* will
3524    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3525    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3526    must be retried after every newline in the subject.
3527    
3528    
3529    Version 2.04 18-Feb-99
3530    ----------------------
3531    
3532    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3533    computation of the store needed to hold the pattern was incorrect (too large).
3534    If such patterns were nested a few deep, this could multiply and become a real
3535    problem.
3536    
3537    2. Added /M option to pcretest to show the memory requirement of a specific
3538    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3539    
3540    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3541    compiled in such a way that the backtracking after subsequent failure was
3542    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3543    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3544    
3545    
3546    Version 2.03 02-Feb-99
3547    ----------------------
3548    
3549    1. Fixed typo and small mistake in man page.
3550    
3551    2. Added 4th condition (GPL supersedes if conflict) and created separate
3552    LICENCE file containing the conditions.
3553    
3554    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3555    Perl, that is the internal \ allows the delimiter to be included in the
3556    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
3557    the final delimiter, add \ to the end of the pattern (to test the error).
3558    
3559    4. Added the convenience functions for extracting substrings after a successful
3560    match. Updated pcretest to make it able to test these functions.
3561    
3562    
3563    Version 2.02 14-Jan-99
3564    ----------------------
3565    
3566    1. Initialized the working variables associated with each extraction so that
3567    their saving and restoring doesn't refer to uninitialized store.
3568    
3569    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
3570    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
3571    fix the problem.
3572    
3573    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
3574    calls, and wasn't printing the correct value for compiling calls. Increased the
3575    default value of LOOPREPEAT, and the number of significant figures in the
3576    times.
3577    
3578    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
3579    
3580    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
3581    a building problem on Windows NT with a FAT file system.
3582    
3583    
3584    Version 2.01 21-Oct-98
3585    ----------------------
3586    
3587    1. Changed the API for pcre_compile() to allow for the provision of a pointer
3588    to character tables built by pcre_maketables() in the current locale. If NULL
3589    is passed, the default tables are used.
3590    
3591    
3592  Version 2.00 24-Sep-98  Version 2.00 24-Sep-98
3593  ----------------------  ----------------------

Legend:
Removed from v.23  
changed lines
  Added in v.539

  ViewVC Help
Powered by ViewVC 1.1.5