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

Diff of /code/trunk/ChangeLog

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

revision 35 by nigel, Sat Feb 24 21:39:05 2007 UTC revision 565 by ph10, Sun Oct 31 18:18:48 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.11 10-Oct-2010
5    ------------------------
6    
7    1.  (*THEN) was not working properly if there were untried alternatives prior
8        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
9        backtracked to try for "b" instead of moving to the next alternative branch
10        at the same level (in this case, to look for "c"). The Perl documentation
11        is clear that when (*THEN) is backtracked onto, it goes to the "next
12        alternative in the innermost enclosing group".
13    
14    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
15        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
16        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
17        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
18        (*THEN).
19    
20    3.  If \s appeared in a character class, it removed the VT character from
21        the class, even if it had been included by some previous item, for example
22        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
23        of \s, but is part of the POSIX "space" class.)
24    
25    4.  A partial match never returns an empty string (because you can always
26        match an empty string at the end of the subject); however the checking for
27        an empty string was starting at the "start of match" point. This has been
28        changed to the "earliest inspected character" point, because the returned
29        data for a partial match starts at this character. This means that, for
30        example, /(?<=abc)def/ gives a partial match for the subject "abc"
31        (previously it gave "no match").
32    
33    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
34        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
35        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
36        has an implication that the given string is incomplete (because a partial
37        match is preferred over a full match). For this reason, these items now
38        give a partial match in this situation. [Aside: previously, the one case
39        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
40        match rather than a full match, which was wrong by the old rules, but is
41        now correct.]
42    
43    6.  There was a bug in the handling of #-introduced comments, recognized when
44        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
45        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
46        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
47        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
48        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
49        places in pcre_compile().
50    
51    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
52        comments when looking ahead for named forward references to subpatterns,
53        the only newline sequence it recognized was NL. It now handles newlines
54        according to the set newline convention.
55    
56    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
57        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
58        cater for a lack of strerror(). These oversights have been fixed.
59    
60    9.  Added --match-limit and --recursion-limit to pcregrep.
61    
62    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
63    
64    11. When the -o option was used, pcregrep was setting a return code of 1, even
65        when matches were found, and --line-buffered was not being honoured.
66    
67    12. Added an optional parentheses number to the -o and --only-matching options
68        of pcregrep.
69    
70    
71    Version 8.10 25-Jun-2010
72    ------------------------
73    
74    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
75        THEN.
76    
77    2.  (*ACCEPT) was not working when inside an atomic group.
78    
79    3.  Inside a character class, \B is treated as a literal by default, but
80        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
81        causes the error). The code is unchanged, but I tidied the documentation.
82    
83    4.  Inside a character class, PCRE always treated \R and \X as literals,
84        whereas Perl faults them if its -w option is set. I have changed PCRE so
85        that it faults them when PCRE_EXTRA is set.
86    
87    5.  Added support for \N, which always matches any character other than
88        newline. (It is the same as "." when PCRE_DOTALL is not set.)
89    
90    6.  When compiling pcregrep with newer versions of gcc which may have
91        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
92        declared with attribute warn_unused_result" were given. Just casting the
93        result to (void) does not stop the warnings; a more elaborate fudge is
94        needed. I've used a macro to implement this.
95    
96    7.  Minor change to pcretest.c to avoid a compiler warning.
97    
98    8.  Added four artifical Unicode properties to help with an option to make
99        \s etc use properties (see next item). The new properties are: Xan
100        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
101    
102    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
103        use Unicode properties. (*UCP) at the start of a pattern can be used to set
104        this option. Modified pcretest to add /W to test this facility. Added
105        REG_UCP to make it available via the POSIX interface.
106    
107    10. Added --line-buffered to pcregrep.
108    
109    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
110        studied, and the match started with a letter with a code point greater than
111        127 whose first byte was different to the first byte of the other case of
112        the letter, the other case of this starting letter was not recognized
113        (#976).
114    
115    12. If a pattern that was studied started with a repeated Unicode property
116        test, for example, \p{Nd}+, there was the theoretical possibility of
117        setting up an incorrect bitmap of starting bytes, but fortunately it could
118        not have actually happened in practice until change 8 above was made (it
119        added property types that matched character-matching opcodes).
120    
121    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
122        possible starting bytes for non-anchored patterns.
123    
124    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
125        \R, and also a number of cases that involve Unicode properties, both
126        explicit and implicit when PCRE_UCP is set.
127    
128    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
129        input, it could crash or give wrong results if characters with values
130        greater than 0xc0 were present in the subject string. (Detail: it assumed
131        UTF-8 input when processing these items.)
132    
133    16. Added a lot of (int) casts to avoid compiler warnings in systems where
134        size_t is 64-bit (#991).
135    
136    17. Added a check for running out of memory when PCRE is compiled with
137        --disable-stack-for-recursion (#990).
138    
139    18. If the last data line in a file for pcretest does not have a newline on
140        the end, a newline was missing in the output.
141    
142    19. The default pcre_chartables.c file recognizes only ASCII characters (values
143        less than 128) in its various bitmaps. However, there is a facility for
144        generating tables according to the current locale when PCRE is compiled. It
145        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
146        characters, are recognized by isspace() and therefore were getting set in
147        these tables, and indeed these tables seem to approximate to ISO 8859. This
148        caused a problem in UTF-8 mode when pcre_study() was used to create a list
149        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
150        which of course cannot start UTF-8 characters. I have changed the code so
151        that only real ASCII characters (less than 128) and the correct starting
152        bytes for UTF-8 encodings are set for characters greater than 127 when in
153        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
154        altogether.)
155    
156    20. Added the /T option to pcretest so as to be able to run tests with non-
157        standard character tables, thus making it possible to include the tests
158        used for 19 above in the standard set of tests.
159    
160    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
161        reference to a subpattern the other side of a comment that contains an
162        opening parenthesis caused either an internal compiling error, or a
163        reference to the wrong subpattern.
164    
165    
166    Version 8.02 19-Mar-2010
167    ------------------------
168    
169    1.  The Unicode data tables have been updated to Unicode 5.2.0.
170    
171    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
172        configured.
173    
174    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
175        original author of that file, following a query about its status.
176    
177    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
178        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
179    
180    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
181        quantifier applied to a forward-referencing subroutine call, could compile
182        incorrect code or give the error "internal error: previously-checked
183        referenced subpattern not found".
184    
185    6.  Both MS Visual Studio and Symbian OS have problems with initializing
186        variables to point to external functions. For these systems, therefore,
187        pcre_malloc etc. are now initialized to local functions that call the
188        relevant global functions.
189    
190    7.  There were two entries missing in the vectors called coptable and poptable
191        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
192        I've fixed the data, and added a kludgy way of testing at compile time that
193        the lengths are correct (equal to the number of opcodes).
194    
195    8.  Following on from 7, I added a similar kludge to check the length of the
196        eint vector in pcreposix.c.
197    
198    9.  Error texts for pcre_compile() are held as one long string to avoid too
199        much relocation at load time. To find a text, the string is searched,
200        counting zeros. There was no check for running off the end of the string,
201        which could happen if a new error number was added without updating the
202        string.
203    
204    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
205    
206    11. \K was not working if it appeared in an atomic group or in a group that
207        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
208        \K is "not well defined" if used in an assertion. PCRE now accepts it if
209        the assertion is positive, but not if it is negative.
210    
211    12. Change 11 fortuitously reduced the size of the stack frame used in the
212        "match()" function of pcre_exec.c by one pointer. Forthcoming
213        implementation of support for (*MARK) will need an extra pointer on the
214        stack; I have reserved it now, so that the stack frame size does not
215        decrease.
216    
217    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
218        item in branch that calls a recursion is a subroutine call - as in the
219        second branch in the above example - was incorrectly given the compile-
220        time error "recursive call could loop indefinitely" because pcre_compile()
221        was not correctly checking the subroutine for matching a non-empty string.
222    
223    14. The checks for overrunning compiling workspace could trigger after an
224        overrun had occurred. This is a "should never occur" error, but it can be
225        triggered by pathological patterns such as hundreds of nested parentheses.
226        The checks now trigger 100 bytes before the end of the workspace.
227    
228    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
229    
230    
231    Version 8.01 19-Jan-2010
232    ------------------------
233    
234    1.  If a pattern contained a conditional subpattern with only one branch (in
235        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
236        computed the wrong minimum data length (which is of course zero for such
237        subpatterns). This could cause incorrect "no match" results.
238    
239    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
240        the pattern is reset in the first branch, pcre_compile() failed with
241        "internal error: code overflow at offset...". This happened only when
242        the reset was to the original external option setting. (An optimization
243        abstracts leading options settings into an external setting, which was the
244        cause of this.)
245    
246    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
247        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
248        assertion pattern did not match (meaning that the assertion was true), it
249        was incorrectly treated as false if the SKIP had been reached during the
250        matching. This also applied to assertions used as conditions.
251    
252    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
253        assertion subpattern, including such a pattern used as a condition,
254        unpredictable results occurred, instead of the error return
255        PCRE_ERROR_DFA_UITEM.
256    
257    5.  The C++ GlobalReplace function was not working like Perl for the special
258        situation when an empty string is matched. It now does the fancy magic
259        stuff that is necessary.
260    
261    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
262        removed. (These were left over from very, very early versions of PCRE.)
263    
264    7.  Some cosmetic changes to the code to make life easier when compiling it
265        as part of something else:
266    
267        (a) Change DEBUG to PCRE_DEBUG.
268    
269        (b) In pcre_compile(), rename the member of the "branch_chain" structure
270            called "current" as "current_branch", to prevent a collision with the
271            Linux macro when compiled as a kernel module.
272    
273        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
274            prevent a collision with the Linux macro when compiled as a kernel
275            module.
276    
277    8.  In pcre_compile() there are some checks for integer overflows that used to
278        cast potentially large values to (double). This has been changed to that
279        when building, a check for int64_t is made, and if it is found, it is used
280        instead, thus avoiding the use of floating point arithmetic. (There is no
281        other use of FP in PCRE.) If int64_t is not found, the fallback is to
282        double.
283    
284    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
285        2005 (difference between two addresses compared to an unsigned value).
286    
287    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
288        custom one, because of the following reported problem in Windows:
289    
290          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
291              under Win32.
292          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
293              therefore missing the function definition.
294          - The compiler thus generates a "C" signature for the test function.
295          - The linker fails to find the "C" function.
296          - PCRE fails to configure if asked to do so against libbz2.
297    
298    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
299        messages were output:
300    
301          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
302          rerunning libtoolize, to keep the correct libtool macros in-tree.
303          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
304    
305        I have done both of these things.
306    
307    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
308        most of the time, it *can* run out if it is given a pattern that contains a
309        runaway infinite recursion. I updated the discussion in the pcrestack man
310        page.
311    
312    13. Now that we have gone to the x.xx style of version numbers, the minor
313        version may start with zero. Using 08 or 09 is a bad idea because users
314        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
315        interpreted as invalid octal numbers. I've updated the previous comment in
316        configure.ac, and also added a check that gives an error if 08 or 09 are
317        used.
318    
319    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
320        causing partial matching to fail when the end of the subject matched \W
321        in a UTF-8 pattern where \W was quantified with a minimum of 3.
322    
323    15. There were some discrepancies between the declarations in pcre_internal.h
324        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
325        their definitions. The declarations used "const uschar *" and the
326        definitions used USPTR. Even though USPTR is normally defined as "const
327        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
328        reported that: "This difference in casting confuses some C++ compilers, for
329        example, SunCC recognizes above declarations as different functions and
330        generates broken code for hbpcre." I have changed the declarations to use
331        USPTR.
332    
333    16. GNU libtool is named differently on some systems. The autogen.sh script now
334        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
335        (FreeBSD).
336    
337    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
338        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
339        comment: "Figure out how to create a longlong from a string: strtoll and
340        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
341        instance, but it only takes 2 args instead of 3!"
342    
343    18. A subtle bug concerned with back references has been fixed by a change of
344        specification, with a corresponding code fix. A pattern such as
345        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
346        refers, was giving matches when it shouldn't. For example, xa=xaaa would
347        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
348        same bug. Such groups have to be quantified to be useful, or contained
349        inside another quantified group. (If there's no repetition, the reference
350        can never match.) The problem arises because, having left the group and
351        moved on to the rest of the pattern, a later failure that backtracks into
352        the group uses the captured value from the final iteration of the group
353        rather than the correct earlier one. I have fixed this in PCRE by forcing
354        any group that contains a reference to itself to be an atomic group; that
355        is, there cannot be any backtracking into it once it has completed. This is
356        similar to recursive and subroutine calls.
357    
358    
359    Version 8.00 19-Oct-09
360    ----------------------
361    
362    1.  The table for translating pcre_compile() error codes into POSIX error codes
363        was out-of-date, and there was no check on the pcre_compile() error code
364        being within the table. This could lead to an OK return being given in
365        error.
366    
367    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
368        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
369        in a Windows environment.
370    
371    3.  The pcregrep --count option prints the count for each file even when it is
372        zero, as does GNU grep. However, pcregrep was also printing all files when
373        --files-with-matches was added. Now, when both options are given, it prints
374        counts only for those files that have at least one match. (GNU grep just
375        prints the file name in this circumstance, but including the count seems
376        more useful - otherwise, why use --count?) Also ensured that the
377        combination -clh just lists non-zero counts, with no names.
378    
379    4.  The long form of the pcregrep -F option was incorrectly implemented as
380        --fixed_strings instead of --fixed-strings. This is an incompatible change,
381        but it seems right to fix it, and I didn't think it was worth preserving
382        the old behaviour.
383    
384    5.  The command line items --regex=pattern and --regexp=pattern were not
385        recognized by pcregrep, which required --regex pattern or --regexp pattern
386        (with a space rather than an '='). The man page documented the '=' forms,
387        which are compatible with GNU grep; these now work.
388    
389    6.  No libpcreposix.pc file was created for pkg-config; there was just
390        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
391    
392    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
393        when UCP support is not needed, by modifying the Python script that
394        generates it from Unicode data files. This should not matter if the module
395        is correctly used as a library, but I received one complaint about 50K of
396        unwanted data. My guess is that the person linked everything into his
397        program rather than using a library. Anyway, it does no harm.
398    
399    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
400        was a minimum greater than 1 for a wide character in a possessive
401        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
402        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
403        character. Chaos in the form of incorrect output or a compiling loop could
404        result.
405    
406    9.  The restrictions on what a pattern can contain when partial matching is
407        requested for pcre_exec() have been removed. All patterns can now be
408        partially matched by this function. In addition, if there are at least two
409        slots in the offset vector, the offset of the earliest inspected character
410        for the match and the offset of the end of the subject are set in them when
411        PCRE_ERROR_PARTIAL is returned.
412    
413    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
414        synonymous with PCRE_PARTIAL, for backwards compatibility, and
415        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
416        and may be more useful for multi-segment matching.
417    
418    11. Partial matching with pcre_exec() is now more intuitive. A partial match
419        used to be given if ever the end of the subject was reached; now it is
420        given only if matching could not proceed because another character was
421        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
422        string "Z", which now yields "no match" instead of "partial match". In the
423        case of pcre_dfa_exec(), "no match" is given if every matching path for the
424        final character ended with (*FAIL).
425    
426    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
427        if the pattern had a "must contain" character that was already found in the
428        earlier partial match, unless partial matching was again requested. For
429        example, with the pattern /dog.(body)?/, the "must contain" character is
430        "g". If the first part-match was for the string "dog", restarting with
431        "sbody" failed. This bug has been fixed.
432    
433    13. The string returned by pcre_dfa_exec() after a partial match has been
434        changed so that it starts at the first inspected character rather than the
435        first character of the match. This makes a difference only if the pattern
436        starts with a lookbehind assertion or \b or \B (\K is not supported by
437        pcre_dfa_exec()). It's an incompatible change, but it makes the two
438        matching functions compatible, and I think it's the right thing to do.
439    
440    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
441        so that the demonstration program is easily available in environments where
442        PCRE has not been installed from source.
443    
444    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
445        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
446        library.
447    
448    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
449        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
450        is not the first non-POSIX option to be added. Clearly some people find
451        these options useful.
452    
453    17. If a caller to the POSIX matching function regexec() passes a non-zero
454        value for nmatch with a NULL value for pmatch, the value of
455        nmatch is forced to zero.
456    
457    18. RunGrepTest did not have a test for the availability of the -u option of
458        the diff command, as RunTest does. It now checks in the same way as
459        RunTest, and also checks for the -b option.
460    
461    19. If an odd number of negated classes containing just a single character
462        interposed, within parentheses, between a forward reference to a named
463        subpattern and the definition of the subpattern, compilation crashed with
464        an internal error, complaining that it could not find the referenced
465        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
466        [The bug was that it was starting one character too far in when skipping
467        over the character class, thus treating the ] as data rather than
468        terminating the class. This meant it could skip too much.]
469    
470    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
471        /g option in pcretest when the pattern contains \K, which makes it possible
472        to have an empty string match not at the start, even when the pattern is
473        anchored. Updated pcretest and pcredemo to use this option.
474    
475    21. If the maximum number of capturing subpatterns in a recursion was greater
476        than the maximum at the outer level, the higher number was returned, but
477        with unset values at the outer level. The correct (outer level) value is
478        now given.
479    
480    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
481        PCRE did not set those parentheses (unlike Perl). I have now found a way to
482        make it do so. The string so far is captured, making this feature
483        compatible with Perl.
484    
485    23. The tests have been re-organized, adding tests 11 and 12, to make it
486        possible to check the Perl 5.10 features against Perl 5.10.
487    
488    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
489        pattern matches a fixed length string. PCRE did not allow this; now it
490        does. Neither allows recursion.
491    
492    25. I finally figured out how to implement a request to provide the minimum
493        length of subject string that was needed in order to match a given pattern.
494        (It was back references and recursion that I had previously got hung up
495        on.) This code has now been added to pcre_study(); it finds a lower bound
496        to the length of subject needed. It is not necessarily the greatest lower
497        bound, but using it to avoid searching strings that are too short does give
498        some useful speed-ups. The value is available to calling programs via
499        pcre_fullinfo().
500    
501    26. While implementing 25, I discovered to my embarrassment that pcretest had
502        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
503        study optimizations had never been tested with that matching function.
504        Oops. What is worse, even when it was passed study data, there was a bug in
505        pcre_dfa_exec() that meant it never actually used it. Double oops. There
506        were also very few tests of studied patterns with pcre_dfa_exec().
507    
508    27. If (?| is used to create subpatterns with duplicate numbers, they are now
509        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
510        on the other side of the coin, they are no longer allowed to have different
511        names, because these cannot be distinguished in PCRE, and this has caused
512        confusion. (This is a difference from Perl.)
513    
514    28. When duplicate subpattern names are present (necessarily with different
515        numbers, as required by 27 above), and a test is made by name in a
516        conditional pattern, either for a subpattern having been matched, or for
517        recursion in such a pattern, all the associated numbered subpatterns are
518        tested, and the overall condition is true if the condition is true for any
519        one of them. This is the way Perl works, and is also more like the way
520        testing by number works.
521    
522    
523    Version 7.9 11-Apr-09
524    ---------------------
525    
526    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
527        (pcretest), all targets were linked against these libraries. This included
528        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
529        libraries. This caused unwanted dependencies to be created. This problem
530        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
531        pcretest is linked with readline.
532    
533    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
534        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
535        moved outside it again, because FALSE and TRUE are already defined in AIX,
536        but BOOL is not.
537    
538    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
539        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
540    
541    4.  The pcregrep documentation said spaces were inserted as well as colons (or
542        hyphens) following file names and line numbers when outputting matching
543        lines. This is not true; no spaces are inserted. I have also clarified the
544        wording for the --colour (or --color) option.
545    
546    5.  In pcregrep, when --colour was used with -o, the list of matching strings
547        was not coloured; this is different to GNU grep, so I have changed it to be
548        the same.
549    
550    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
551        each matching line was coloured. Now it goes on to look for further matches
552        of any of the test patterns, which is the same behaviour as GNU grep.
553    
554    7.  A pattern that could match an empty string could cause pcregrep to loop; it
555        doesn't make sense to accept an empty string match in pcregrep, so I have
556        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
557        seems to be how GNU grep behaves.
558    
559    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
560        start or after a newline", because the conditional assertion was not being
561        correctly handled. The rule now is that both the assertion and what follows
562        in the first alternative must satisfy the test.
563    
564    9.  If auto-callout was enabled in a pattern with a conditional group whose
565        condition was an assertion, PCRE could crash during matching, both with
566        pcre_exec() and pcre_dfa_exec().
567    
568    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
569        used for matching.
570    
571    11. Unicode property support in character classes was not working for
572        characters (bytes) greater than 127 when not in UTF-8 mode.
573    
574    12. Added the -M command line option to pcretest.
575    
576    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
577    
578    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
579    
580    16. Added comments and documentation about mis-use of no_arg in the C++
581        wrapper.
582    
583    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
584        from Martin Jerabek that uses macro names for all relevant character and
585        string constants.
586    
587    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
588        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
589        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
590        these, but not everybody uses configure.
591    
592    19. A conditional group that had only one branch was not being correctly
593        recognized as an item that could match an empty string. This meant that an
594        enclosing group might also not be so recognized, causing infinite looping
595        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
596        with the subject "ab", where knowledge that the repeated group can match
597        nothing is needed in order to break the loop.
598    
599    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
600        exec(), but without supplying a callout function, matching went wrong.
601    
602    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
603        leak if the size of the offset vector was greater than 30. When the vector
604        is smaller, the saved offsets during recursion go onto a local stack
605        vector, but for larger vectors malloc() is used. It was failing to free
606        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
607        error, in fact).
608    
609    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
610        heapframe that is used only when UTF-8 support is enabled. This caused no
611        problem, but was untidy.
612    
613    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
614        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
615        included within another project.
616    
617    24. Steven Van Ingelgem's patches to add more options to the CMake support,
618        slightly modified by me:
619    
620          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
621              not building pcregrep.
622    
623          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
624              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
625    
626    25. Forward references, both numeric and by name, in patterns that made use of
627        duplicate group numbers, could behave incorrectly or give incorrect errors,
628        because when scanning forward to find the reference group, PCRE was not
629        taking into account the duplicate group numbers. A pattern such as
630        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
631    
632    26. Changed a few more instances of "const unsigned char *" to USPTR, making
633        the feature of a custom pointer more persuasive (as requested by a user).
634    
635    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
636        pcretest.c, inside #ifndefs, because it seems they are sometimes already
637        pre-defined.
638    
639    28. Added support for (*UTF8) at the start of a pattern.
640    
641    29. Arrange for flags added by the "release type" setting in CMake to be shown
642        in the configuration summary.
643    
644    
645    Version 7.8 05-Sep-08
646    ---------------------
647    
648    1.  Replaced UCP searching code with optimized version as implemented for Ad
649        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
650        stage table and inline lookup instead of a function, giving speed ups of 2
651        to 5 times on some simple patterns that I tested. Permission was given to
652        distribute the MultiStage2.py script that generates the tables (it's not in
653        the tarball, but is in the Subversion repository).
654    
655    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
656        scripts.
657    
658    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
659        a group with a zero qualifier. The result of the study could be incorrect,
660        or the function might crash, depending on the pattern.
661    
662    4.  Caseless matching was not working for non-ASCII characters in back
663        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
664        It now works when Unicode Property Support is available.
665    
666    5.  In pcretest, an escape such as \x{de} in the data was always generating
667        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
668        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
669        truncation.
670    
671    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
672    
673    7.  Added two (int) casts to pcregrep when printing the difference of two
674        pointers, in case they are 64-bit values.
675    
676    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
677        test 2 if it fails.
678    
679    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
680        and a #define of that name to empty if it is not externally set. This is to
681        allow users of MSVC to set it if necessary.
682    
683    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
684        the convenience functions in the pcre_get.c source file.
685    
686    11. An option change at the start of a pattern that had top-level alternatives
687        could cause overwriting and/or a crash. This command provoked a crash in
688        some environments:
689    
690          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
691    
692        This potential security problem was recorded as CVE-2008-2371.
693    
694    12. For a pattern where the match had to start at the beginning or immediately
695        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
696        pcre_dfa_exec() could read past the end of the passed subject if there was
697        no match. To help with detecting such bugs (e.g. with valgrind), I modified
698        pcretest so that it places the subject at the end of its malloc-ed buffer.
699    
700    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
701        exec() might read past the end of the data buffer in UTF-8 mode.
702    
703    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
704        the data contained the byte 0x85 as part of a UTF-8 character within its
705        first line. This applied both to normal and DFA matching.
706    
707    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
708        /^[^d]*?$/8 failed to match "abc".
709    
710    16. Added a missing copyright notice to pcrecpp_internal.h.
711    
712    17. Make it more clear in the documentation that values returned from
713        pcre_exec() in ovector are byte offsets, not character counts.
714    
715    18. Tidied a few places to stop certain compilers from issuing warnings.
716    
717    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
718        supplied by Stefan Weber. I made a further small update for 7.8 because
719        there is a change of source arrangements: the pcre_searchfuncs.c module is
720        replaced by pcre_ucd.c.
721    
722    
723    Version 7.7 07-May-08
724    ---------------------
725    
726    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
727        a string to a long long, pretend we don't even have a long long." This is
728        done by checking for the strtoq, strtoll, and _strtoi64 functions.
729    
730    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
731        pre-7.6 versions, which defined a global no_arg variable instead of putting
732        it in the RE class. (See also #8 below.)
733    
734    3.  Remove a line of dead code, identified by coverity and reported by Nuno
735        Lopes.
736    
737    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
738    
739        (1) The include/exclude patterns were being applied to the whole pathnames
740            of files, instead of just to the final components.
741    
742        (2) If there was more than one level of directory, the subdirectories were
743            skipped unless they satisfied the include/exclude conditions. This is
744            inconsistent with GNU grep (and could even be seen as contrary to the
745            pcregrep specification - which I improved to make it absolutely clear).
746            The action now is always to scan all levels of directory, and just
747            apply the include/exclude patterns to regular files.
748    
749    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
750        --exclude_dir in the tests to avoid scanning .svn directories.
751    
752    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
753        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
754        doesn't support NULs in patterns.
755    
756    7.  Added some missing "const"s to declarations of static tables in
757        pcre_compile.c and pcre_dfa_exec.c.
758    
759    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
760        caused by fix #2  above. (Subsequently also a second patch to fix the
761        first patch. And a third patch - this was a messy problem.)
762    
763    9.  Applied Craig's patch to remove the use of push_back().
764    
765    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
766        matching function regexec().
767    
768    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
769        which, however, unlike Perl's \g{...}, are subroutine calls, not back
770        references. PCRE supports relative numbers with this syntax (I don't think
771        Oniguruma does).
772    
773    12. Previously, a group with a zero repeat such as (...){0} was completely
774        omitted from the compiled regex. However, this means that if the group
775        was called as a subroutine from elsewhere in the pattern, things went wrong
776        (an internal error was given). Such groups are now left in the compiled
777        pattern, with a new opcode that causes them to be skipped at execution
778        time.
779    
780    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
781        to the way PCRE behaves:
782    
783        (a) A lone ] character is dis-allowed (Perl treats it as data).
784    
785        (b) A back reference to an unmatched subpattern matches an empty string
786            (Perl fails the current match path).
787    
788        (c) A data ] in a character class must be notated as \] because if the
789            first data character in a class is ], it defines an empty class. (In
790            Perl it is not possible to have an empty class.) The empty class []
791            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
792            The negative empty class [^] matches any one character, independently
793            of the DOTALL setting.
794    
795    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
796        non-existent subpattern following a character class starting with ']' and
797        containing () gave an internal compiling error instead of "reference to
798        non-existent subpattern". Fortunately, when the pattern did exist, the
799        compiled code was correct. (When scanning forwards to check for the
800        existencd of the subpattern, it was treating the data ']' as terminating
801        the class, so got the count wrong. When actually compiling, the reference
802        was subsequently set up correctly.)
803    
804    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
805        it was being rejected as not supported by pcre_dfa_exec(), even though
806        other assertions are supported. I have made pcre_dfa_exec() support
807        (*FAIL).
808    
809    16. The implementation of 13c above involved the invention of a new opcode,
810        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
811        cannot be changed at match time, I realized I could make a small
812        improvement to matching performance by compiling OP_ALLANY instead of
813        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
814        on the OP_ANY path.
815    
816    17. Compiling pcretest on Windows with readline support failed without the
817        following two fixes: (1) Make the unistd.h include conditional on
818        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
819    
820    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
821        ncurses library to be included for pcretest when ReadLine support is
822        requested, but also to allow for it to be overridden. This patch came from
823        Daniel Bergström.
824    
825    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
826        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
827        any errors with the current Unicode tables. Thanks to Peter Kankowski for
828        spotting this.
829    
830    
831    Version 7.6 28-Jan-08
832    ---------------------
833    
834    1.  A character class containing a very large number of characters with
835        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
836        overflow.
837    
838    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
839        HAVE_LONG_LONG is not defined.
840    
841    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
842        bring them up to date and include new features. This patch includes:
843    
844        - Fixed PH's badly added libz and libbz2 support.
845        - Fixed a problem with static linking.
846        - Added pcredemo. [But later removed - see 7 below.]
847        - Fixed dftables problem and added an option.
848        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
849            HAVE_LONG_LONG.
850        - Added readline support for pcretest.
851        - Added an listing of the option settings after cmake has run.
852    
853    4.  A user submitted a patch to Makefile that makes it easy to create
854        "pcre.dll" under mingw when using Configure/Make. I added stuff to
855        Makefile.am that cause it to include this special target, without
856        affecting anything else. Note that the same mingw target plus all
857        the other distribution libraries and programs are now supported
858        when configuring with CMake (see 6 below) instead of with
859        Configure/Make.
860    
861    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
862        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
863        exported in the Windows port". It has not yet been confirmed that the patch
864        solves the problem, but it does no harm.
865    
866    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
867        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
868        with CMake, and also correct the comment about stack recursion.
869    
870    7.  Remove the automatic building of pcredemo from the ./configure system and
871        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
872        of a program that users should build themselves after PCRE is installed, so
873        building it automatically is not really right. What is more, it gave
874        trouble in some build environments.
875    
876    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
877    
878    
879    Version 7.5 10-Jan-08
880    ---------------------
881    
882    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
883        values in parens when parsing an RE using the C++ wrapper."
884    
885    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
886        Characters greater than 255 were excluded from the class instead of being
887        included.
888    
889    3.  The same bug as (2) above applied to negated POSIX classes such as
890        [:^space:].
891    
892    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
893        defined or documented. It seems to have been a typo for PCRE_STATIC, so
894        I have changed it.
895    
896    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
897        first named subpattern) and a construct such as (?&a) would reference the
898        first named subpattern whose name started with "a" (in other words, the
899        length check was missing). Both these problems are fixed. "Subpattern name
900        expected" is now given for (?&) (a zero-length name), and this patch also
901        makes it give the same error for \k'' (previously it complained that that
902        was a reference to a non-existent subpattern).
903    
904    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
905        this is right because (?- can be followed by option settings as well as by
906        digits. I have, however, made the messages clearer.
907    
908    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
909        than the number used in the conditional) now cause a compile-time error.
910        This is actually not compatible with Perl, which accepts such patterns, but
911        treats the conditional as always being FALSE (as PCRE used to), but it
912        seems to me that giving a diagnostic is better.
913    
914    8.  Change "alphameric" to the more common word "alphanumeric" in comments
915        and messages.
916    
917    9.  Fix two occurrences of "backslash" in comments that should have been
918        "backspace".
919    
920    10. Remove two redundant lines of code that can never be obeyed (their function
921        was moved elsewhere).
922    
923    11. The program that makes PCRE's Unicode character property table had a bug
924        which caused it to generate incorrect table entries for sequences of
925        characters that have the same character type, but are in different scripts.
926        It amalgamated them into a single range, with the script of the first of
927        them. In other words, some characters were in the wrong script. There were
928        thirteen such cases, affecting characters in the following ranges:
929    
930          U+002b0 - U+002c1
931          U+0060c - U+0060d
932          U+0061e - U+00612
933          U+0064b - U+0065e
934          U+0074d - U+0076d
935          U+01800 - U+01805
936          U+01d00 - U+01d77
937          U+01d9b - U+01dbf
938          U+0200b - U+0200f
939          U+030fc - U+030fe
940          U+03260 - U+0327f
941          U+0fb46 - U+0fbb1
942          U+10450 - U+1049d
943    
944    12. The -o option (show only the matching part of a line) for pcregrep was not
945        compatible with GNU grep in that, if there was more than one match in a
946        line, it showed only the first of them. It now behaves in the same way as
947        GNU grep.
948    
949    13. If the -o and -v options were combined for pcregrep, it printed a blank
950        line for every non-matching line. GNU grep prints nothing, and pcregrep now
951        does the same. The return code can be used to tell if there were any
952        non-matching lines.
953    
954    14. Added --file-offsets and --line-offsets to pcregrep.
955    
956    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
957        infinitely looping recursion. The bug was that positive lookaheads were not
958        being skipped when checking for a possible empty match (negative lookaheads
959        and both kinds of lookbehind were skipped).
960    
961    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
962        inclusion of <windows.h> to before rather than after the definition of
963        INVALID_FILE_ATTRIBUTES (patch from David Byron).
964    
965    17. Specifying a possessive quantifier with a specific limit for a Unicode
966        character property caused pcre_compile() to compile bad code, which led at
967        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
968        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
969        caused the error; without that there was no problem.
970    
971    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
972    
973    19. Added --enable-pcretest-libreadline.
974    
975    20. In pcrecpp.cc, the variable 'count' was incremented twice in
976        RE::GlobalReplace(). As a result, the number of replacements returned was
977        double what it should be. I removed one of the increments, but Craig sent a
978        later patch that removed the other one (the right fix) and added unit tests
979        that check the return values (which was not done before).
980    
981    21. Several CMake things:
982    
983        (1) Arranged that, when cmake is used on Unix, the libraries end up with
984            the names libpcre and libpcreposix, not just pcre and pcreposix.
985    
986        (2) The above change means that pcretest and pcregrep are now correctly
987            linked with the newly-built libraries, not previously installed ones.
988    
989        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
990    
991    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
992        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
993        UTF-8 newline character). The key issue is that the pattern starts .*;
994        this means that the match must be either at the beginning, or after a
995        newline. The bug was in the code for advancing after a failed match and
996        checking that the new position followed a newline. It was not taking
997        account of UTF-8 characters correctly.
998    
999    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1000        character classes. PCRE was not treating the sequence [:...:] as a
1001        character class unless the ... were all letters. Perl, however, seems to
1002        allow any characters between [: and :], though of course it rejects as
1003        unknown any "names" that contain non-letters, because all the known class
1004        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1005        for example, whereas PCRE did not - it did not recognize a POSIX character
1006        class. This seemed a bit dangerous, so the code has been changed to be
1007        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1008        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1009        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1010        Perl does, and where it didn't before.
1011    
1012    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1013        Windows environments %n is disabled by default.
1014    
1015    
1016    Version 7.4 21-Sep-07
1017    ---------------------
1018    
1019    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1020        means that a class such as [\s] counted as "explicit reference to CR or
1021        LF". That isn't really right - the whole point of the change was to try to
1022        help when there was an actual mention of one of the two characters. So now
1023        the change happens only if \r or \n (or a literal CR or LF) character is
1024        encountered.
1025    
1026    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1027        of both had grown to the point where there were only 3 bits left.
1028        Fortunately, there was spare space in the data structure, and so I have
1029        moved the internal flags into a new 16-bit field to free up more option
1030        bits.
1031    
1032    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1033        but did not set the internal JCHANGED flag - either of these is enough to
1034        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1035        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1036        start sets both bits.
1037    
1038    4.  Added options (at build time, compile time, exec time) to change \R from
1039        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1040    
1041    5.  doc/pcresyntax.html was missing from the distribution.
1042    
1043    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1044        compatibility, even though it is no longer used.
1045    
1046    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1047        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1048        windows.h file is present (where different names are used). [This was
1049        reversed later after testing - see 16 below.]
1050    
1051    8.  Changed all #include <config.h> to #include "config.h". There were also
1052        some further <pcre.h> cases that I changed to "pcre.h".
1053    
1054    9.  When pcregrep was used with the --colour option, it missed the line ending
1055        sequence off the lines that it output.
1056    
1057    10. It was pointed out to me that arrays of string pointers cause lots of
1058        relocations when a shared library is dynamically loaded. A technique of
1059        using a single long string with a table of offsets can drastically reduce
1060        these. I have refactored PCRE in four places to do this. The result is
1061        dramatic:
1062    
1063          Originally:                          290
1064          After changing UCP table:            187
1065          After changing error message table:   43
1066          After changing table of "verbs"       36
1067          After changing table of Posix names   22
1068    
1069        Thanks to the folks working on Gregex for glib for this insight.
1070    
1071    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1072        unicode-properties was also set.
1073    
1074    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1075    
1076    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1077        checked only for CRLF.
1078    
1079    14. Added casts to pcretest.c to avoid compiler warnings.
1080    
1081    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1082    
1083    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1084        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1085        entirely. This removes changes made in 7 above.
1086    
1087    17. The CMake files have been updated, and there is now more information about
1088        building with CMake in the NON-UNIX-USE document.
1089    
1090    
1091    Version 7.3 28-Aug-07
1092    ---------------------
1093    
1094     1. In the rejigging of the build system that eventually resulted in 7.1, the
1095        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1096        brackets there is not right, since it causes compilers to look for an
1097        installed pcre.h, not the version that is in the source that is being
1098        compiled (which of course may be different). I have changed it back to:
1099    
1100          #include "pcre.h"
1101    
1102        I have a vague recollection that the change was concerned with compiling in
1103        different directories, but in the new build system, that is taken care of
1104        by the VPATH setting the Makefile.
1105    
1106     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1107        when the subject happened to end in the byte 0x85 (e.g. if the last
1108        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1109        characters but of course it shouldn't be taken as a newline when it is part
1110        of another character. The bug was that, for an unlimited repeat of . in
1111        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1112        characters when looking for a newline.
1113    
1114     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1115    
1116     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1117        in debug output.
1118    
1119     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1120        long printing in the pcrecpp unittest when running under MinGW.
1121    
1122     6. ESC_K was left out of the EBCDIC table.
1123    
1124     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1125        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1126        limit also applies to "virtual nesting" when a pattern is recursive, and in
1127        this case 1000 isn't so big. I have been able to remove this limit at the
1128        expense of backing off one optimization in certain circumstances. Normally,
1129        when pcre_exec() would call its internal match() function recursively and
1130        immediately return the result unconditionally, it uses a "tail recursion"
1131        feature to save stack. However, when a subpattern that can match an empty
1132        string has an unlimited repetition quantifier, it no longer makes this
1133        optimization. That gives it a stack frame in which to save the data for
1134        checking that an empty string has been matched. Previously this was taken
1135        from the 1000-entry workspace that had been reserved. So now there is no
1136        explicit limit, but more stack is used.
1137    
1138     8. Applied Daniel's patches to solve problems with the import/export magic
1139        syntax that is required for Windows, and which was going wrong for the
1140        pcreposix and pcrecpp parts of the library. These were overlooked when this
1141        problem was solved for the main library.
1142    
1143     9. There were some crude static tests to avoid integer overflow when computing
1144        the size of patterns that contain repeated groups with explicit upper
1145        limits. As the maximum quantifier is 65535, the maximum group length was
1146        set at 30,000 so that the product of these two numbers did not overflow a
1147        32-bit integer. However, it turns out that people want to use groups that
1148        are longer than 30,000 bytes (though not repeat them that many times).
1149        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1150        made it possible to implement the integer overflow checks in a much more
1151        dynamic way, which I have now done. The artificial limitation on group
1152        length has been removed - we now have only the limit on the total length of
1153        the compiled pattern, which depends on the LINK_SIZE setting.
1154    
1155    10. Fixed a bug in the documentation for get/copy named substring when
1156        duplicate names are permitted. If none of the named substrings are set, the
1157        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1158        empty string.
1159    
1160    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1161        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1162        because the ] is interpreted as the first data character and the
1163        terminating ] is not found. PCRE has been made compatible with Perl in this
1164        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1165        cause memory overwriting.
1166    
1167    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1168        string has been matched (to stop an infinite loop). It was not recognizing
1169        a conditional subpattern that could match an empty string if that
1170        subpattern was within another subpattern. For example, it looped when
1171        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1172        condition was not nested. This bug has been fixed.
1173    
1174    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1175        past the start of the subject in the presence of bytes with the top bit
1176        set, for example "\x8aBCD".
1177    
1178    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1179        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1180    
1181    14. Optimized (?!) to (*FAIL).
1182    
1183    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1184        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1185        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1186        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1187        does: it's just the validity check that is more restrictive.
1188    
1189    16. Inserted checks for integer overflows during escape sequence (backslash)
1190        processing, and also fixed erroneous offset values for syntax errors during
1191        backslash processing.
1192    
1193    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1194        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1195    
1196    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1197        caused an overrun.
1198    
1199    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1200        something other than just ASCII characters) inside a group that had an
1201        unlimited repeat caused a loop at compile time (while checking to see
1202        whether the group could match an empty string).
1203    
1204    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1205        [\P{Any}] did so. (Error in the code for printing property names.)
1206    
1207    21. An orphan \E inside a character class could cause a crash.
1208    
1209    22. A repeated capturing bracket such as (A)? could cause a wild memory
1210        reference during compilation.
1211    
1212    23. There are several functions in pcre_compile() that scan along a compiled
1213        expression for various reasons (e.g. to see if it's fixed length for look
1214        behind). There were bugs in these functions when a repeated \p or \P was
1215        present in the pattern. These operators have additional parameters compared
1216        with \d, etc, and these were not being taken into account when moving along
1217        the compiled data. Specifically:
1218    
1219        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1220            length.
1221    
1222        (b) An item such as \pL+ within a repeated group could cause crashes or
1223            loops.
1224    
1225        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1226            "reference to non-existent subpattern" error.
1227    
1228        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1229    
1230    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1231        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1232    
1233    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1234    
1235    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1236        character were causing crashes (broken optimization).
1237    
1238    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1239        \p or \P) caused a compile-time loop.
1240    
1241    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1242        break. For example, the unstudied pattern [\r\n]A does not match the string
1243        "\r\nA" because change 7.0/46 below moves the current point on by two
1244        characters after failing to match at the start. However, the pattern \nA
1245        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1246        the same is true. There doesn't seem any very clean way out of this, but
1247        what I have chosen to do makes the common cases work: PCRE now takes note
1248        of whether there can be an explicit match for \r or \n anywhere in the
1249        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1250        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1251        pattern has explicit CR or LF references.
1252    
1253    29. Added (*CR) etc for changing newline setting at start of pattern.
1254    
1255    
1256    Version 7.2 19-Jun-07
1257    ---------------------
1258    
1259     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1260        which is apparently normally available under Windows.
1261    
1262     2. Re-jig the pcregrep tests with different newline settings in an attempt
1263        to make them independent of the local environment's newline setting.
1264    
1265     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1266    
1267     4. Some of the "internals" tests were previously cut out when the link size
1268        was not 2, because the output contained actual offsets. The recent new
1269        "Z" feature of pcretest means that these can be cut out, making the tests
1270        usable with all link sizes.
1271    
1272     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1273        stack recursion. This gives a massive performance boost under BSD, but just
1274        a small improvement under Linux. However, it saves one field in the frame
1275        in all cases.
1276    
1277     6. Added more features from the forthcoming Perl 5.10:
1278    
1279        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1280            recursion call. It refers to the nth most recently opened parentheses.
1281    
1282        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1283            to be opened parentheses.
1284    
1285        (c) Conditions that refer to capturing parentheses can be specified
1286            relatively, for example, (?(-2)... or (?(+3)...
1287    
1288        (d) \K resets the start of the current match so that everything before
1289            is not part of it.
1290    
1291        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1292    
1293        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1294            reference syntax.
1295    
1296        (g) (?| introduces a group in which the numbering of parentheses in each
1297            alternative starts with the same number.
1298    
1299        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1300    
1301     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1302        PCRE_INFO_JCHANGED.
1303    
1304     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1305        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1306        for detecting groups that can match an empty string.
1307    
1308     9. A pattern with a very large number of alternatives (more than several
1309        hundred) was running out of internal workspace during the pre-compile
1310        phase, where pcre_compile() figures out how much memory will be needed. A
1311        bit of new cunning has reduced the workspace needed for groups with
1312        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1313        workspace instead of running out of the 4096 that are available.
1314    
1315    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1316    
1317    11. Applied patch from Google to remove an optimization that didn't quite work.
1318        The report of the bug said:
1319    
1320          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1321          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1322          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1323    
1324    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1325        it matched the wrong number of bytes.
1326    
1327    
1328    Version 7.1 24-Apr-07
1329    ---------------------
1330    
1331     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1332        that is more "standard", making use of automake and other Autotools. There
1333        is some re-arrangement of the files and adjustment of comments consequent
1334        on this.
1335    
1336     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1337        for recursive directory scanning broke on some systems because the files
1338        are not scanned in any specific order and on different systems the order
1339        was different. A call to "sort" has been inserted into RunGrepTest for the
1340        approprate test as a short-term fix. In the longer term there may be an
1341        alternative.
1342    
1343     3. I had an email from Eric Raymond about problems translating some of PCRE's
1344        man pages to HTML (despite the fact that I distribute HTML pages, some
1345        people do their own conversions for various reasons). The problems
1346        concerned the use of low-level troff macros .br and .in. I have therefore
1347        removed all such uses from the man pages (some were redundant, some could
1348        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1349        HTML has been updated to handle .nf/.fi and to complain if it encounters
1350        .br or .in.
1351    
1352     4. Updated comments in configure.ac that get placed in config.h.in and also
1353        arranged for config.h to be included in the distribution, with the name
1354        config.h.generic, for the benefit of those who have to compile without
1355        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1356    
1357     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1358        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1359        makevp.bat for the current PCRE, using the additional files
1360        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1361    
1362     6. A Windows user reported a minor discrepancy with test 2, which turned out
1363        to be caused by a trailing space on an input line that had got lost in his
1364        copy. The trailing space was an accident, so I've just removed it.
1365    
1366     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1367        that is needed.
1368    
1369     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1370        as "const" (a) because they are and (b) because it helps the PHP
1371        maintainers who have recently made a script to detect big data structures
1372        in the php code that should be moved to the .rodata section. I remembered
1373        to update Builducptable as well, so it won't revert if ucptable.h is ever
1374        re-created.
1375    
1376     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1377        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1378        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1379        support is not required. This saves 1.5-2K of code, which is important in
1380        some applications.
1381    
1382        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1383        so as not to refer to the tables, even though these functions will never be
1384        called when UTF-8 support is disabled. Otherwise there are problems with a
1385        shared library.
1386    
1387    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1388    
1389        (a) It was defining its arguments as char * instead of void *.
1390    
1391        (b) It was assuming that all moves were upwards in memory; this was true
1392            a long time ago when I wrote it, but is no longer the case.
1393    
1394        The emulated memove() is provided for those environments that have neither
1395        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1396        is clearly not the case, as these two bugs were recently reported.
1397    
1398    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1399        and Detrail to create the HTML documentation, the .txt form of the man
1400        pages, and it removes trailing spaces from listed files. It also creates
1401        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1402        case, it wraps all the #defines with #ifndefs. This script should be run
1403        before "make dist".
1404    
1405    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1406        with Unicode property support.
1407    
1408        (a) For a maximizing quantifier, if the two different cases of the
1409            character were of different lengths in their UTF-8 codings (there are
1410            some cases like this - I found 11), and the matching function had to
1411            back up over a mixture of the two cases, it incorrectly assumed they
1412            were both the same length.
1413    
1414        (b) When PCRE was configured to use the heap rather than the stack for
1415            recursion during matching, it was not correctly preserving the data for
1416            the other case of a UTF-8 character when checking ahead for a match
1417            while processing a minimizing repeat. If the check also involved
1418            matching a wide character, but failed, corruption could cause an
1419            erroneous result when trying to check for a repeat of the original
1420            character.
1421    
1422    13. Some tidying changes to the testing mechanism:
1423    
1424        (a) The RunTest script now detects the internal link size and whether there
1425            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1426            values substituted by "configure". (The RunGrepTest script already did
1427            this for UTF-8.) The configure.ac script no longer substitutes the
1428            relevant variables.
1429    
1430        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1431            with length and offset values. This means that the output is different
1432            for different internal link sizes. Test 2 is skipped for link sizes
1433            other than 2 because of this, bypassing the problem. Unfortunately,
1434            there was also a test in test 3 (the locale tests) that used /B and
1435            failed for link sizes other than 2. Rather than cut the whole test out,
1436            I have added a new /Z option to pcretest that replaces the length and
1437            offset values with spaces. This is now used to make test 3 independent
1438            of link size. (Test 2 will be tidied up later.)
1439    
1440    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1441        segmentation fault instead of returning the appropriate error message.
1442    
1443    15. In multiline mode when the newline sequence was set to "any", the pattern
1444        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1445        This doesn't seem right; it now treats the CRLF combination as the line
1446        ending, and so does not match in that case. It's only a pattern such as ^$
1447        that would hit this one: something like ^ABC$ would have failed after \r
1448        and then tried again after \r\n.
1449    
1450    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1451        in an attempt to make files that differ only in their line terminators
1452        compare equal. This works on Linux.
1453    
1454    17. Under certain error circumstances pcregrep might try to free random memory
1455        as it exited. This is now fixed, thanks to valgrind.
1456    
1457    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1458        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1459        was because its rules for how to advance for /g after matching an empty
1460        string at the end of a line did not allow for this case. They now check for
1461        it specially.
1462    
1463    20. pcretest is supposed to handle patterns and data of any length, by
1464        extending its buffers when necessary. It was getting this wrong when the
1465        buffer for a data line had to be extended.
1466    
1467    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1468        CRLF as a newline sequence.
1469    
1470    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1471        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1472        I have nevertheless tidied it up.
1473    
1474    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1475    
1476    24. Added a man page for pcre-config.
1477    
1478    
1479    Version 7.0 19-Dec-06
1480    ---------------------
1481    
1482     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1483        moving to gcc 4.1.1.
1484    
1485     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1486        sys/time.h, which is documented as needed for this function. It doesn't
1487        seem to matter on Linux, but it showed up on some releases of OS X.
1488    
1489     3. It seems that there are systems where bytes whose values are greater than
1490        127 match isprint() in the "C" locale. The "C" locale should be the
1491        default when a C program starts up. In most systems, only ASCII printing
1492        characters match isprint(). This difference caused the output from pcretest
1493        to vary, making some of the tests fail. I have changed pcretest so that:
1494    
1495        (a) When it is outputting text in the compiled version of a pattern, bytes
1496            other than 32-126 are always shown as hex escapes.
1497    
1498        (b) When it is outputting text that is a matched part of a subject string,
1499            it does the same, unless a different locale has been set for the match
1500            (using the /L modifier). In this case, it uses isprint() to decide.
1501    
1502     4. Fixed a major bug that caused incorrect computation of the amount of memory
1503        required for a compiled pattern when options that changed within the
1504        pattern affected the logic of the preliminary scan that determines the
1505        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1506        that the computed length was too small. The symptoms of this bug were
1507        either the PCRE error "internal error: code overflow" from pcre_compile(),
1508        or a glibc crash with a message such as "pcretest: free(): invalid next
1509        size (fast)". Examples of patterns that provoked this bug (shown in
1510        pcretest format) are:
1511    
1512          /(?-x: )/x
1513          /(?x)(?-x: \s*#\s*)/
1514          /((?i)[\x{c0}])/8
1515          /(?i:[\x{c0}])/8
1516    
1517        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1518        is now done differently.
1519    
1520     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1521        wrapper classes; (b) implement a new function in the C++ scanner that is
1522        more efficient than the old way of doing things because it avoids levels of
1523        recursion in the regex matching; (c) add a paragraph to the documentation
1524        for the FullMatch() function.
1525    
1526     6. The escape sequence \n was being treated as whatever was defined as
1527        "newline". Not only was this contrary to the documentation, which states
1528        that \n is character 10 (hex 0A), but it also went horribly wrong when
1529        "newline" was defined as CRLF. This has been fixed.
1530    
1531     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1532        was being set to -1 for the "end of line" case (supposedly a value that no
1533        character can have). Though this value is never used (the check for end of
1534        line is "zero bytes in current character"), it caused compiler complaints.
1535        I've changed it to 0xffffffff.
1536    
1537     8. In pcre_version.c, the version string was being built by a sequence of
1538        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1539        string (as it is for production releases) called a macro with an empty
1540        argument. The C standard says the result of this is undefined. The gcc
1541        compiler treats it as an empty string (which was what was wanted) but it is
1542        reported that Visual C gives an error. The source has been hacked around to
1543        avoid this problem.
1544    
1545     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1546        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1547        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1548        of them did).
1549    
1550    10. Originally, pcretest opened its input and output without "b"; then I was
1551        told that "b" was needed in some environments, so it was added for release
1552        5.0 to both the input and output. (It makes no difference on Unix-like
1553        systems.) Later I was told that it is wrong for the input on Windows. I've
1554        now abstracted the modes into two macros, to make it easier to fiddle with
1555        them, and removed "b" from the input mode under Windows.
1556    
1557    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1558    
1559    12. Added -help and --help to pcretest as an official way of being reminded
1560        of the options.
1561    
1562    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1563        and pcrecpp.cc because they annoy compilers at high warning levels.
1564    
1565    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1566    
1567    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1568        scripts are not C programs :-) and which was not noticed because it works
1569        on Linux.
1570    
1571    16. pcretest is supposed to handle any length of pattern and data line (as one
1572        line or as a continued sequence of lines) by extending its input buffer if
1573        necessary. This feature was broken for very long pattern lines, leading to
1574        a string of junk being passed to pcre_compile() if the pattern was longer
1575        than about 50K.
1576    
1577    17. I have done a major re-factoring of the way pcre_compile() computes the
1578        amount of memory needed for a compiled pattern. Previously, there was code
1579        that made a preliminary scan of the pattern in order to do this. That was
1580        OK when PCRE was new, but as the facilities have expanded, it has become
1581        harder and harder to keep it in step with the real compile phase, and there
1582        have been a number of bugs (see for example, 4 above). I have now found a
1583        cunning way of running the real compile function in a "fake" mode that
1584        enables it to compute how much memory it would need, while actually only
1585        ever using a few hundred bytes of working memory and without too many
1586        tests of the mode. This should make future maintenance and development
1587        easier. A side effect of this work is that the limit of 200 on the nesting
1588        depth of parentheses has been removed (though this was never a serious
1589        limitation, I suspect). However, there is a downside: pcre_compile() now
1590        runs more slowly than before (30% or more, depending on the pattern). I
1591        hope this isn't a big issue. There is no effect on runtime performance.
1592    
1593    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1594        newline (only possible for the last line of a file) and it was a
1595        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1596    
1597    19. Added additional timing features to pcretest. (1) The -tm option now times
1598        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1599        separate command line item, by a number that specifies the number of
1600        repeats to use when timing. The default is 50000; this gives better
1601        precision, but takes uncomfortably long for very large patterns.
1602    
1603    20. Extended pcre_study() to be more clever in cases where a branch of a
1604        subpattern has no definite first character. For example, (a*|b*)[cd] would
1605        previously give no result from pcre_study(). Now it recognizes that the
1606        first character must be a, b, c, or d.
1607    
1608    21. There was an incorrect error "recursive call could loop indefinitely" if
1609        a subpattern (or the entire pattern) that was being tested for matching an
1610        empty string contained only one non-empty item after a nested subpattern.
1611        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1612        incorrectly, because the \d was being skipped in the check.
1613    
1614    22. The pcretest program now has a new pattern option /B and a command line
1615        option -b, which is equivalent to adding /B to every pattern. This causes
1616        it to show the compiled bytecode, without the additional information that
1617        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1618        is the same as /B/I).
1619    
1620    23. A new optimization is now able automatically to treat some sequences such
1621        as a*b as a*+b. More specifically, if something simple (such as a character
1622        or a simple class like \d) has an unlimited quantifier, and is followed by
1623        something that cannot possibly match the quantified thing, the quantifier
1624        is automatically "possessified".
1625    
1626    24. A recursive reference to a subpattern whose number was greater than 39
1627        went wrong under certain circumstances in UTF-8 mode. This bug could also
1628        have affected the operation of pcre_study().
1629    
1630    25. Realized that a little bit of performance could be had by replacing
1631        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1632    
1633    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1634    
1635    27. Possessive quantifiers such as a++ were previously implemented by turning
1636        them into atomic groups such as ($>a+). Now they have their own opcodes,
1637        which improves performance. This includes the automatically created ones
1638        from 23 above.
1639    
1640    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1641        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1642        the first matched character to be a colon. This applied both to named and
1643        numbered groups.
1644    
1645    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1646    
1647    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1648        building PCRE on Windows easier, so I have included it in the distribution.
1649    
1650    31. There is now a check in pcretest against a ridiculously large number being
1651        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1652        loop, the loop is abandoned.
1653    
1654    32. Forward references to subpatterns in conditions such as (?(2)...) where
1655        subpattern 2 is defined later cause pcre_compile() to search forwards in
1656        the pattern for the relevant set of parentheses. This search went wrong
1657        when there were unescaped parentheses in a character class, parentheses
1658        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1659    
1660    33. "Subroutine" calls and backreferences were previously restricted to
1661        referencing subpatterns earlier in the regex. This restriction has now
1662        been removed.
1663    
1664    34. Added a number of extra features that are going to be in Perl 5.10. On the
1665        whole, these are just syntactic alternatives for features that PCRE had
1666        previously implemented using the Python syntax or my own invention. The
1667        other formats are all retained for compatibility.
1668    
1669        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1670            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1671            also .NET compatible.
1672    
1673        (b) A recursion or subroutine call to a named group can now be defined as
1674            (?&name) as well as (?P>name).
1675    
1676        (c) A backreference to a named group can now be defined as \k<name> or
1677            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1678            5.10, are also .NET compatible.
1679    
1680        (d) A conditional reference to a named group can now use the syntax
1681            (?(<name>) or (?('name') as well as (?(name).
1682    
1683        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1684            groups (named and numbered) that are never evaluated inline, but can be
1685            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1686            is always false. There may be only one alternative in such a group.
1687    
1688        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1689            as the simple (?(R). The condition is true only if the most recent
1690            recursion is that of the given number or name. It does not search out
1691            through the entire recursion stack.
1692    
1693        (g) The escape \gN or \g{N} has been added, where N is a positive or
1694            negative number, specifying an absolute or relative reference.
1695    
1696    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1697        some "unreachable code" warnings.
1698    
1699    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1700        things, this adds five new scripts.
1701    
1702    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1703        There were also incompatibilities regarding the handling of \Q..\E inside
1704        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1705        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1706    
1707    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1708        matches an empty string, and forcibly breaks the loop. There were bugs in
1709        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1710        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1711        separate and independent bugs (that affected different cases) have been
1712        fixed.
1713    
1714    39. Refactored the code to abolish the use of different opcodes for small
1715        capturing bracket numbers. This is a tidy that I avoided doing when I
1716        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1717        The new approach is not only tidier, it makes it possible to reduce the
1718        memory needed to fix the previous bug (38).
1719    
1720    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1721        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1722        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1723        mode.
1724    
1725    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1726        report.
1727    
1728    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1729        copy construction and assignment in the C++ wrapper.
1730    
1731    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1732        couple of bugs that could have given wrong results in the "--newline=crlf"
1733        case.
1734    
1735    44. Added a number of casts and did some reorganization of signed/unsigned int
1736        variables following suggestions from Dair Grant. Also renamed the variable
1737        "this" as "item" because it is a C++ keyword.
1738    
1739    45. Arranged for dftables to add
1740    
1741          #include "pcre_internal.h"
1742    
1743        to pcre_chartables.c because without it, gcc 4.x may remove the array
1744        definition from the final binary if PCRE is built into a static library and
1745        dead code stripping is activated.
1746    
1747    46. For an unanchored pattern, if a match attempt fails at the start of a
1748        newline sequence, and the newline setting is CRLF or ANY, and the next two
1749        characters are CRLF, advance by two characters instead of one.
1750    
1751    
1752    Version 6.7 04-Jul-06
1753    ---------------------
1754    
1755     1. In order to handle tests when input lines are enormously long, pcretest has
1756        been re-factored so that it automatically extends its buffers when
1757        necessary. The code is crude, but this _is_ just a test program. The
1758        default size has been increased from 32K to 50K.
1759    
1760     2. The code in pcre_study() was using the value of the re argument before
1761        testing it for NULL. (Of course, in any sensible call of the function, it
1762        won't be NULL.)
1763    
1764     3. The memmove() emulation function in pcre_internal.h, which is used on
1765        systems that lack both memmove() and bcopy() - that is, hardly ever -
1766        was missing a "static" storage class specifier.
1767    
1768     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1769        containing an extended class (one that cannot be represented by a bitmap
1770        because it contains high-valued characters or Unicode property items, e.g.
1771        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1772        pattern, but PCRE should not loop if you do not (it no longer does).
1773        [Detail: two cases were found: (a) a repeated subpattern containing an
1774        extended class; (b) a recursive reference to a subpattern that followed a
1775        previous extended class. It wasn't skipping over the extended class
1776        correctly when UTF-8 mode was not set.]
1777    
1778     5. A negated single-character class was not being recognized as fixed-length
1779        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1780        compile error "lookbehind assertion is not fixed length".
1781    
1782     6. The RunPerlTest auxiliary script was showing an unexpected difference
1783        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1784        write a Perl script that can interpret lines of an input file either as
1785        byte characters or as UTF-8, which is what "perltest" was being required to
1786        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1787        can't do is switch easily at run time between having the "use utf8;" pragma
1788        or not. In the end, I fudged it by using the RunPerlTest script to insert
1789        "use utf8;" explicitly for the UTF-8 tests.
1790    
1791     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1792        the end of the subject string, contrary to the documentation and to what
1793        Perl does. This was true of both matching functions. Now it matches only at
1794        the start of the subject and immediately after *internal* newlines.
1795    
1796     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1797        a pointer to an int instead of a pointer to an unsigned long int. This
1798        caused problems on 64-bit systems.
1799    
1800     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1801        instance of the 'standard' template library not being so standard".
1802    
1803    10. There was no check on the number of named subpatterns nor the maximum
1804        length of a subpattern name. The product of these values is used to compute
1805        the size of the memory block for a compiled pattern. By supplying a very
1806        long subpattern name and a large number of named subpatterns, the size
1807        computation could be caused to overflow. This is now prevented by limiting
1808        the length of names to 32 characters, and the number of named subpatterns
1809        to 10,000.
1810    
1811    11. Subpatterns that are repeated with specific counts have to be replicated in
1812        the compiled pattern. The size of memory for this was computed from the
1813        length of the subpattern and the repeat count. The latter is limited to
1814        65535, but there was no limit on the former, meaning that integer overflow
1815        could in principle occur. The compiled length of a repeated subpattern is
1816        now limited to 30,000 bytes in order to prevent this.
1817    
1818    12. Added the optional facility to have named substrings with the same name.
1819    
1820    13. Added the ability to use a named substring as a condition, using the
1821        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1822        are numbers (not recommended). Forward references are permitted.
1823    
1824    14. Added forward references in named backreferences (if you see what I mean).
1825    
1826    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1827        pattern could run off the end of the subject. For example, the pattern
1828        "(?s)(.{1,5})"8 did this with the subject "ab".
1829    
1830    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1831        PCRE_CASELESS was set when matching characters that were quantified with ?
1832        or *.
1833    
1834    17. A character class other than a single negated character that had a minimum
1835        but no maximum quantifier - for example [ab]{6,} - was not handled
1836        correctly by pce_dfa_exec(). It would match only one character.
1837    
1838    18. A valid (though odd) pattern that looked like a POSIX character
1839        class but used an invalid character after [ (for example [[,abc,]]) caused
1840        pcre_compile() to give the error "Failed: internal error: code overflow" or
1841        in some cases to crash with a glibc free() error. This could even happen if
1842        the pattern terminated after [[ but there just happened to be a sequence of
1843        letters, a binary zero, and a closing ] in the memory that followed.
1844    
1845    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1846        over the years. Originally (before any Unicode support), just the bottom 8
1847        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1848        output from "man perlunicode" includes this:
1849    
1850          The regular expression compiler produces polymorphic opcodes.  That
1851          is, the pattern adapts to the data and automatically switches to
1852          the Unicode character scheme when presented with Unicode data--or
1853          instead uses a traditional byte scheme when presented with byte
1854          data.
1855    
1856        Sadly, a wide octal escape does not cause a switch, and in a string with
1857        no other multibyte characters, these octal escapes are treated as before.
1858        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1859        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1860        Unicode string.
1861    
1862        I have not perpetrated such confusion in PCRE. Up till now, it took just
1863        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1864        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1865        translate to the appropriate multibyte character.
1866    
1867    29. Applied some refactoring to reduce the number of warnings from Microsoft
1868        and Borland compilers. This has included removing the fudge introduced
1869        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1870        a warning about an unused variable.
1871    
1872    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1873        characters since release 4.0, because Perl (from release 5.004) does not.
1874        [Or at least, is documented not to: some releases seem to be in conflict
1875        with the documentation.] However, when a pattern was studied with
1876        pcre_study() and all its branches started with \s, PCRE still included VT
1877        as a possible starting character. Of course, this did no harm; it just
1878        caused an unnecessary match attempt.
1879    
1880    22. Removed a now-redundant internal flag bit that recorded the fact that case
1881        dependency changed within the pattern. This was once needed for "required
1882        byte" processing, but is no longer used. This recovers a now-scarce options
1883        bit. Also moved the least significant internal flag bit to the most-
1884        significant bit of the word, which was not previously used (hangover from
1885        the days when it was an int rather than a uint) to free up another bit for
1886        the future.
1887    
1888    23. Added support for CRLF line endings as well as CR and LF. As well as the
1889        default being selectable at build time, it can now be changed at runtime
1890        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1891        specify that it is scanning data with non-default line endings.
1892    
1893    24. Changed the definition of CXXLINK to make it agree with the definition of
1894        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1895    
1896    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1897        recursions. This makes a big different to stack usage for some patterns.
1898    
1899    26. If a subpattern containing a named recursion or subroutine reference such
1900        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1901        the space required for the compiled pattern went wrong and gave too small a
1902        value. Depending on the environment, this could lead to "Failed: internal
1903        error: code overflow at offset 49" or "glibc detected double free or
1904        corruption" errors.
1905    
1906    27. Applied patches from Google (a) to support the new newline modes and (b) to
1907        advance over multibyte UTF-8 characters in GlobalReplace.
1908    
1909    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1910        difference for some implementation of PCRE in some Windows version.
1911    
1912    29. Added some extra testing facilities to pcretest:
1913    
1914        \q<number>   in a data line sets the "match limit" value
1915        \Q<number>   in a data line sets the "match recursion limt" value
1916        -S <number>  sets the stack size, where <number> is in megabytes
1917    
1918        The -S option isn't available for Windows.
1919    
1920    
1921    Version 6.6 06-Feb-06
1922    ---------------------
1923    
1924     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1925        in pcreposix.h. I have copied the definition from pcre.h.
1926    
1927     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1928        because pcre.h is no longer a built file.
1929    
1930     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1931        not normally included in the compiled code.
1932    
1933    
1934    Version 6.5 01-Feb-06
1935    ---------------------
1936    
1937     1. When using the partial match feature with pcre_dfa_exec(), it was not
1938        anchoring the second and subsequent partial matches at the new starting
1939        point. This could lead to incorrect results. For example, with the pattern
1940        /1234/, partially matching against "123" and then "a4" gave a match.
1941    
1942     2. Changes to pcregrep:
1943    
1944        (a) All non-match returns from pcre_exec() were being treated as failures
1945            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1946            error message is output. Some extra information is given for the
1947            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1948            probably the only errors that are likely to be caused by users (by
1949            specifying a regex that has nested indefinite repeats, for instance).
1950            If there are more than 20 of these errors, pcregrep is abandoned.
1951    
1952        (b) A binary zero was treated as data while matching, but terminated the
1953            output line if it was written out. This has been fixed: binary zeroes
1954            are now no different to any other data bytes.
1955    
1956        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1957            used to set a locale for matching. The --locale=xxxx long option has
1958            been added (no short equivalent) to specify a locale explicitly on the
1959            pcregrep command, overriding the environment variables.
1960    
1961        (d) When -B was used with -n, some line numbers in the output were one less
1962            than they should have been.
1963    
1964        (e) Added the -o (--only-matching) option.
1965    
1966        (f) If -A or -C was used with -c (count only), some lines of context were
1967            accidentally printed for the final match.
1968    
1969        (g) Added the -H (--with-filename) option.
1970    
1971        (h) The combination of options -rh failed to suppress file names for files
1972            that were found from directory arguments.
1973    
1974        (i) Added the -D (--devices) and -d (--directories) options.
1975    
1976        (j) Added the -F (--fixed-strings) option.
1977    
1978        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1979    
1980        (l) Added the --colo(u)r option.
1981    
1982        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1983            is not present by default.
1984    
1985     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1986        items such as (?R) or (?1), when the recursion could match a number of
1987        alternatives. If it matched one of the alternatives, but subsequently,
1988        outside the recursion, there was a failure, the code tried to back up into
1989        the recursion. However, because of the way PCRE is implemented, this is not
1990        possible, and the result was an incorrect result from the match.
1991    
1992        In order to prevent this happening, the specification of recursion has
1993        been changed so that all such subpatterns are automatically treated as
1994        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1995    
1996     4. I had overlooked the fact that, in some locales, there are characters for
1997        which isalpha() is true but neither isupper() nor islower() are true. In
1998        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1999        and ordfeminine) are like this. This affected the treatment of \w and \W
2000        when they appeared in character classes, but not when they appeared outside
2001        a character class. The bit map for "word" characters is now created
2002        separately from the results of isalnum() instead of just taking it from the
2003        upper, lower, and digit maps. (Plus the underscore character, of course.)
2004    
2005     5. The above bug also affected the handling of POSIX character classes such as
2006        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2007        permanent tables. Instead, the bit maps for such a class were previously
2008        created as the appropriate unions of the upper, lower, and digit bitmaps.
2009        Now they are created by subtraction from the [[:word:]] class, which has
2010        its own bitmap.
2011    
2012     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2013        It is created by subtracting the vertical space characters (\x09, \x0a,
2014        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2015        subtraction was done in the overall bitmap for a character class, meaning
2016        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2017        be recognized. This bug has been fixed.
2018    
2019     7. Patches from the folks at Google:
2020    
2021          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2022          real life, but is still worth protecting against".
2023    
2024          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2025          regular expressions".
2026    
2027          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2028          have it.
2029    
2030          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2031          "configure" and the latter not, in order to fix a problem somebody had
2032          with compiling the Arg class on HP-UX.
2033    
2034          (e) Improve the error-handling of the C++ wrapper a little bit.
2035    
2036          (f) New tests for checking recursion limiting.
2037    
2038     8. The pcre_memmove() function, which is used only if the environment does not
2039        have a standard memmove() function (and is therefore rarely compiled),
2040        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2041        returning a result (though PCRE never actually uses the result).
2042    
2043     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2044        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2045        returned instead of calling malloc() with an overflowing number that would
2046        most likely cause subsequent chaos.
2047    
2048    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2049    
2050    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2051        with this option is matched, the nmatch and pmatch options of regexec() are
2052        ignored.
2053    
2054    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2055        provided in case anyone wants to the the POSIX interface with UTF-8
2056        strings.
2057    
2058    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2059        C++ linking (needed for some HP-UX environments).
2060    
2061    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2062        (unused parameter) and in the pcre_printint() function (omitted "default"
2063        switch label when the default is to do nothing).
2064    
2065    15. Added some code to make it possible, when PCRE is compiled as a C++
2066        library, to replace subject pointers for pcre_exec() with a smart pointer
2067        class, thus making it possible to process discontinuous strings.
2068    
2069    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2070        much the same function. They were added by different people who were trying
2071        to make PCRE easy to compile on non-Unix systems. It has been suggested
2072        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2073        for compiling on Windows systems. I have therefore replaced it with
2074        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2075        defaults to "extern" for C or "extern C" for C++, which works fine on
2076        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2077        SCOPE with something explicit in config.h. In addition:
2078    
2079        (a) pcreposix.h still had just "extern" instead of either of these macros;
2080            I have replaced it with PCRE_DATA_SCOPE.
2081    
2082        (b) Functions such as _pcre_xclass(), which are internal to the library,
2083            but external in the C sense, all had PCRE_EXPORT in their definitions.
2084            This is apparently wrong for the Windows case, so I have removed it.
2085            (It makes no difference on Unix-like systems.)
2086    
2087    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2088        of recursive calls to match(). This is different to MATCH_LIMIT because
2089        that limits the total number of calls to match(), not all of which increase
2090        the depth of recursion. Limiting the recursion depth limits the amount of
2091        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2092        when PCRE is compiled, and changed at run time. A patch from Google adds
2093        this functionality to the C++ interface.
2094    
2095    18. Changes to the handling of Unicode character properties:
2096    
2097        (a) Updated the table to Unicode 4.1.0.
2098    
2099        (b) Recognize characters that are not in the table as "Cn" (undefined).
2100    
2101        (c) I revised the way the table is implemented to a much improved format
2102            which includes recognition of ranges. It now supports the ranges that
2103            are defined in UnicodeData.txt, and it also amalgamates other
2104            characters into ranges. This has reduced the number of entries in the
2105            table from around 16,000 to around 3,000, thus reducing its size
2106            considerably. I realized I did not need to use a tree structure after
2107            all - a binary chop search is just as efficient. Having reduced the
2108            number of entries, I extended their size from 6 bytes to 8 bytes to
2109            allow for more data.
2110    
2111        (d) Added support for Unicode script names via properties such as \p{Han}.
2112    
2113    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2114        matching that character.
2115    
2116    20. When matching a repeated Unicode property with a minimum greater than zero,
2117        (for example \pL{2,}), PCRE could look past the end of the subject if it
2118        reached it while seeking the minimum number of characters. This could
2119        happen only if some of the characters were more than one byte long, because
2120        there is a check for at least the minimum number of bytes.
2121    
2122    21. Refactored the implementation of \p and \P so as to be more general, to
2123        allow for more different types of property in future. This has changed the
2124        compiled form incompatibly. Anybody with saved compiled patterns that use
2125        \p or \P will have to recompile them.
2126    
2127    22. Added "Any" and "L&" to the supported property types.
2128    
2129    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2130        but give a compile time error if the value is greater than 0xff.
2131    
2132    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2133        accidentally not being installed or uninstalled.
2134    
2135    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2136        made were to insert the current release number. This seemed silly, because
2137        it made things harder for people building PCRE on systems that don't run
2138        "configure". I have turned pcre.h into a distributed file, no longer built
2139        by "configure", with the version identification directly included. There is
2140        no longer a pcre.h.in file.
2141    
2142        However, this change necessitated a change to the pcre-config script as
2143        well. It is built from pcre-config.in, and one of the substitutions was the
2144        release number. I have updated configure.ac so that ./configure now finds
2145        the release number by grepping pcre.h.
2146    
2147    26. Added the ability to run the tests under valgrind.
2148    
2149    
2150    Version 6.4 05-Sep-05
2151    ---------------------
2152    
2153     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2154        "--" to be printed when multiple files were scanned, even when none of the
2155        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2156        consider it to be a bug, and have restored the previous behaviour.
2157    
2158     2. A couple of code tidies to get rid of compiler warnings.
2159    
2160     3. The pcretest program used to cheat by referring to symbols in the library
2161        whose names begin with _pcre_. These are internal symbols that are not
2162        really supposed to be visible externally, and in some environments it is
2163        possible to suppress them. The cheating is now confined to including
2164        certain files from the library's source, which is a bit cleaner.
2165    
2166     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2167        file's purpose clearer.
2168    
2169     5. Reorganized pcre_ucp_findchar().
2170    
2171    
2172    Version 6.3 15-Aug-05
2173    ---------------------
2174    
2175     1. The file libpcre.pc.in did not have general read permission in the tarball.
2176    
2177     2. There were some problems when building without C++ support:
2178    
2179        (a) If C++ support was not built, "make install" and "make test" still
2180            tried to test it.
2181    
2182        (b) There were problems when the value of CXX was explicitly set. Some
2183            changes have been made to try to fix these, and ...
2184    
2185        (c) --disable-cpp can now be used to explicitly disable C++ support.
2186    
2187        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2188            backslash in a target when C++ was disabled. This confuses some
2189            versions of "make", apparently. Using an intermediate variable solves
2190            this. (Same for CPP_LOBJ.)
2191    
2192     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2193        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2194        necessary on certain architectures.
2195    
2196     4. Added a setting of -export-symbols-regex to the link command to remove
2197        those symbols that are exported in the C sense, but actually are local
2198        within the library, and not documented. Their names all begin with
2199        "_pcre_". This is not a perfect job, because (a) we have to except some
2200        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2201        available (and never for static libraries). I have made a note to try to
2202        find a way round (a) in the future.
2203    
2204    
2205    Version 6.2 01-Aug-05
2206    ---------------------
2207    
2208     1. There was no test for integer overflow of quantifier values. A construction
2209        such as {1111111111111111} would give undefined results. What is worse, if
2210        a minimum quantifier for a parenthesized subpattern overflowed and became
2211        negative, the calculation of the memory size went wrong. This could have
2212        led to memory overwriting.
2213    
2214     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2215    
2216     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2217        operating environments where this matters.
2218    
2219     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2220        PCRE options from within the C++ wrapper.
2221    
2222     5. Named capturing subpatterns were not being correctly counted when a pattern
2223        was compiled. This caused two problems: (a) If there were more than 100
2224        such subpatterns, the calculation of the memory needed for the whole
2225        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2226        back references of the form \12, where the number was greater than 9, were
2227        not recognized as back references, even though there were sufficient
2228        previous subpatterns.
2229    
2230     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2231        versions of gcc, e.g. 2.95.4.
2232    
2233    
2234    Version 6.1 21-Jun-05
2235    ---------------------
2236    
2237     1. There was one reference to the variable "posix" in pcretest.c that was not
2238        surrounded by "#if !defined NOPOSIX".
2239    
2240     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2241        the cross-check on the old pcre_info() function, for the benefit of the
2242        cut-down version of PCRE that is currently imported into Exim.
2243    
2244     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2245        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2246        patterns that start (?i) but I don't think that matters. The use of (?i) is
2247        just an example; this all applies to the other options as well.
2248    
2249     4. Since libtool seems to echo the compile commands it is issuing, the output
2250        from "make" can be reduced a bit by putting "@" in front of each libtool
2251        compile command.
2252    
2253     5. Patch from the folks at Google for configure.in to be a bit more thorough
2254        in checking for a suitable C++ installation before trying to compile the
2255        C++ stuff. This should fix a reported problem when a compiler was present,
2256        but no suitable headers.
2257    
2258     6. The man pages all had just "PCRE" as their title. I have changed them to
2259        be the relevant file name. I have also arranged that these names are
2260        retained in the file doc/pcre.txt, which is a concatenation in text format
2261        of all the man pages except the little individual ones for each function.
2262    
2263     7. The NON-UNIX-USE file had not been updated for the different set of source
2264        files that come with release 6. I also added a few comments about the C++
2265        wrapper.
2266    
2267    
2268    Version 6.0 07-Jun-05
2269    ---------------------
2270    
2271     1. Some minor internal re-organization to help with my DFA experiments.
2272    
2273     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2274        didn't matter for the library itself when fully configured, but did matter
2275        when compiling without UCP support, or within Exim, where the ucp files are
2276        not imported.
2277    
2278     3. Refactoring of the library code to split up the various functions into
2279        different source modules. The addition of the new DFA matching code (see
2280        below) to a single monolithic source would have made it really too
2281        unwieldy, quite apart from causing all the code to be include in a
2282        statically linked application, when only some functions are used. This is
2283        relevant even without the DFA addition now that patterns can be compiled in
2284        one application and matched in another.
2285    
2286        The downside of splitting up is that there have to be some external
2287        functions and data tables that are used internally in different modules of
2288        the library but which are not part of the API. These have all had their
2289        names changed to start with "_pcre_" so that they are unlikely to clash
2290        with other external names.
2291    
2292     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2293        a different (DFA) algorithm. Although it is slower than the original
2294        function, it does have some advantages for certain types of matching
2295        problem.
2296    
2297     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2298        including restarting after a partial match.
2299    
2300     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2301        defined when compiling for Windows was sent to me. I have put it into the
2302        code, though I have no means of testing or verifying it.
2303    
2304     7. Added the pcre_refcount() auxiliary function.
2305    
2306     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2307        match before or at the first newline in the subject string. In pcretest,
2308        the /f option on a pattern can be used to set this.
2309    
2310     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2311        would behave wrongly. This has been present in PCRE since release 4.0.
2312    
2313    10. A number of changes to the pcregrep command:
2314    
2315        (a) Refactored how -x works; insert ^(...)$ instead of setting
2316            PCRE_ANCHORED and checking the length, in preparation for adding
2317            something similar for -w.
2318    
2319        (b) Added the -w (match as a word) option.
2320    
2321        (c) Refactored the way lines are read and buffered so as to have more
2322            than one at a time available.
2323    
2324        (d) Implemented a pcregrep test script.
2325    
2326        (e) Added the -M (multiline match) option. This allows patterns to match
2327            over several lines of the subject. The buffering ensures that at least
2328            8K, or the rest of the document (whichever is the shorter) is available
2329            for matching (and similarly the previous 8K for lookbehind assertions).
2330    
2331        (f) Changed the --help output so that it now says
2332    
2333              -w, --word-regex(p)
2334    
2335            instead of two lines, one with "regex" and the other with "regexp"
2336            because that confused at least one person since the short forms are the
2337            same. (This required a bit of code, as the output is generated
2338            automatically from a table. It wasn't just a text change.)
2339    
2340        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2341            option but starts with a hyphen. Could be a pattern or a path name
2342            starting with a hyphen, for instance.
2343    
2344        (h) "-" can be given as a file name to represent stdin.
2345    
2346        (i) When file names are being printed, "(standard input)" is used for
2347            the standard input, for compatibility with GNU grep. Previously
2348            "<stdin>" was used.
2349    
2350        (j) The option --label=xxx can be used to supply a name to be used for
2351            stdin when file names are being printed. There is no short form.
2352    
2353        (k) Re-factored the options decoding logic because we are going to add
2354            two more options that take data. Such options can now be given in four
2355            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2356    
2357        (l) Added the -A, -B, and -C options for requesting that lines of context
2358            around matches be printed.
2359    
2360        (m) Added the -L option to print the names of files that do not contain
2361            any matching lines, that is, the complement of -l.
2362    
2363        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2364            continue to scan other files.
2365    
2366        (o) The -s option was incorrectly implemented. For compatibility with other
2367            greps, it now suppresses the error message for a non-existent or non-
2368            accessible file (but not the return code). There is a new option called
2369            -q that suppresses the output of matching lines, which was what -s was
2370            previously doing.
2371    
2372        (p) Added --include and --exclude options to specify files for inclusion
2373            and exclusion when recursing.
2374    
2375    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2376        Hopefully, it now does.
2377    
2378    12. Missing cast in pcre_study().
2379    
2380    13. Added an "uninstall" target to the makefile.
2381    
2382    14. Replaced "extern" in the function prototypes in Makefile.in with
2383        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2384        world, but is set differently for Windows.
2385    
2386    15. Added a second compiling function called pcre_compile2(). The only
2387        difference is that it has an extra argument, which is a pointer to an
2388        integer error code. When there is a compile-time failure, this is set
2389        non-zero, in addition to the error test pointer being set to point to an
2390        error message. The new argument may be NULL if no error number is required
2391        (but then you may as well call pcre_compile(), which is now just a
2392        wrapper). This facility is provided because some applications need a
2393        numeric error indication, but it has also enabled me to tidy up the way
2394        compile-time errors are handled in the POSIX wrapper.
2395    
2396    16. Added VPATH=.libs to the makefile; this should help when building with one
2397        prefix path and installing with another. (Or so I'm told by someone who
2398        knows more about this stuff than I do.)
2399    
2400    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2401        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2402        match everything, including newlines. This is not POSIX-compatible, but
2403        somebody wanted the feature. From pcretest it can be activated by using
2404        both the P and the s flags.
2405    
2406    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2407    
2408    19. libpcre.pc was being incorrectly installed as executable.
2409    
2410    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2411        it now also looks for '\r' so that it will work unmodified on Windows.
2412    
2413    21. Added Google's contributed C++ wrapper to the distribution.
2414    
2415    22. Added some untidy missing memory free() calls in pcretest, to keep
2416        Electric Fence happy when testing.
2417    
2418    
2419    
2420    Version 5.0 13-Sep-04
2421    ---------------------
2422    
2423     1. Internal change: literal characters are no longer packed up into items
2424        containing multiple characters in a single byte-string. Each character
2425        is now matched using a separate opcode. However, there may be more than one
2426        byte in the character in UTF-8 mode.
2427    
2428     2. The pcre_callout_block structure has two new fields: pattern_position and
2429        next_item_length. These contain the offset in the pattern to the next match
2430        item, and its length, respectively.
2431    
2432     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2433        insertion of callouts before each pattern item. Added the /C option to
2434        pcretest to make use of this.
2435    
2436     4. On the advice of a Windows user, the lines
2437    
2438          #if defined(_WIN32) || defined(WIN32)
2439          _setmode( _fileno( stdout ), 0x8000 );
2440          #endif  /* defined(_WIN32) || defined(WIN32) */
2441    
2442        have been added to the source of pcretest. This apparently does useful
2443        magic in relation to line terminators.
2444    
2445     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2446        for the benefit of those environments where the "b" makes a difference.
2447    
2448     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2449        to know about it. I have put a hack into configure.in that adds in code
2450        to set GCC=yes if CC=icc. This seems to end up at a point in the
2451        generated configure script that is early enough to affect the setting of
2452        compiler options, which is what is needed, but I have no means of testing
2453        whether it really works. (The user who reported this had patched the
2454        generated configure script, which of course I cannot do.)
2455    
2456        LATER: After change 22 below (new libtool files), the configure script
2457        seems to know about icc (and also ecc). Therefore, I have commented out
2458        this hack in configure.in.
2459    
2460     7. Added support for pkg-config (2 patches were sent in).
2461    
2462     8. Negated POSIX character classes that used a combination of internal tables
2463        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2464        [[:^ascii]]. Typically, they would match almost any characters. The other
2465        POSIX classes were not broken in this way.
2466    
2467     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2468        to find the match, as PCRE was deluded into thinking that the match had to
2469        start at the start point or following a newline. The same bug applied to
2470        patterns with negative forward assertions or any backward assertions
2471        preceding ".*" at the start, unless the pattern required a fixed first
2472        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2473    
2474    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2475        starting at the last subject character, bytes beyond the end of the subject
2476        string were read.
2477    
2478    11. Renamed the variable "class" as "classbits" to make life easier for C++
2479        users. (Previously there was a macro definition, but it apparently wasn't
2480        enough.)
2481    
2482    12. Added the new field "tables" to the extra data so that tables can be passed
2483        in at exec time, or the internal tables can be re-selected. This allows
2484        a compiled regex to be saved and re-used at a later time by a different
2485        program that might have everything at different addresses.
2486    
2487    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2488        -R library as well as a -L library.
2489    
2490    14. The debugging options of pcretest (-d on the command line or D on a
2491        pattern) showed incorrect output for anything following an extended class
2492        that contained multibyte characters and which was followed by a quantifier.
2493    
2494    15. Added optional support for general category Unicode character properties
2495        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2496        support. It adds about 90K to the size of the library. The meanings of the
2497        inbuilt class escapes such as \d and \s have NOT been changed.
2498    
2499    16. Updated pcredemo.c to include calls to free() to release the memory for the
2500        compiled pattern.
2501    
2502    17. The generated file chartables.c was being created in the source directory
2503        instead of in the building directory. This caused the build to fail if the
2504        source directory was different from the building directory, and was
2505        read-only.
2506    
2507    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2508        file. No doubt somebody will tell me if they don't make sense... Also added
2509        Dan Mooney's comments about building on OpenVMS.
2510    
2511    19. Added support for partial matching via the PCRE_PARTIAL option for
2512        pcre_exec() and the \P data escape in pcretest.
2513    
2514    20. Extended pcretest with 3 new pattern features:
2515    
2516        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2517              write the compiled pattern to the file whose name is "rest-of-line".
2518              This is a straight binary dump of the data, with the saved pointer to
2519              the character tables forced to be NULL. The study data, if any, is
2520              written too. After writing, pcretest reads a new pattern.
2521    
2522        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2523              compiled pattern from the given file. There must not be any
2524              occurrences of "<" in the file name (pretty unlikely); if there are,
2525              pcretest will instead treat the initial "<" as a pattern delimiter.
2526              After reading in the pattern, pcretest goes on to read data lines as
2527              usual.
2528    
2529        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2530              and 16-bit fields in a compiled pattern, to simulate a pattern that
2531              was compiled on a host of opposite endianness.
2532    
2533    21. The pcre-exec() function can now cope with patterns that were compiled on
2534        hosts of opposite endianness, with this restriction:
2535    
2536          As for any compiled expression that is saved and used later, the tables
2537          pointer field cannot be preserved; the extra_data field in the arguments
2538          to pcre_exec() should be used to pass in a tables address if a value
2539          other than the default internal tables were used at compile time.
2540    
2541    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2542        now diagnosed as an error. Previously, most of the time, a negative number
2543        would have been treated as zero, but if in addition "ovector" was passed as
2544        NULL, a crash could occur.
2545    
2546    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2547        new versions from the libtool 1.5 distribution (the last one is a copy of
2548        a file called libtool.m4). This seems to have fixed the need to patch
2549        "configure" to support Darwin 1.3 (which I used to do). However, I still
2550        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2551        workstation).
2552    
2553    24. Changed the PCRE licence to be the more standard "BSD" licence.
2554    
2555    
2556    Version 4.5 01-Dec-03
2557    ---------------------
2558    
2559     1. There has been some re-arrangement of the code for the match() function so
2560        that it can be compiled in a version that does not call itself recursively.
2561        Instead, it keeps those local variables that need separate instances for
2562        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2563        needs to "recurse". Keeping track of where control must go is done by means
2564        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2565        hide most of the details from the main code, and operates only if
2566        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2567        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2568        operating.
2569    
2570        To make it easier for callers to provide specially tailored get/free
2571        functions for this usage, two new functions, pcre_stack_malloc, and
2572        pcre_stack_free, are used. They are always called in strict stacking order,
2573        and the size of block requested is always the same.
2574    
2575        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2576        PCRE has been compiled to use the stack or the heap for recursion. The
2577        -C option of pcretest uses this to show which version is compiled.
2578    
2579        A new data escape \S, is added to pcretest; it causes the amounts of store
2580        obtained and freed by both kinds of malloc/free at match time to be added
2581        to the output.
2582    
2583     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2584        what's available on my current Linux desktop machine.
2585    
2586     3. When matching a UTF-8 string, the test for a valid string at the start has
2587        been extended. If start_offset is not zero, PCRE now checks that it points
2588        to a byte that is the start of a UTF-8 character. If not, it returns
2589        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2590        this is necessary because there may be backward assertions in the pattern.
2591        When matching the same subject several times, it may save resources to use
2592        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2593    
2594     4. The code for checking the validity of UTF-8 strings has been tightened so
2595        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2596        containing "overlong sequences".
2597    
2598     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2599        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2600        should have been "&", but it just so happened that all the cases this let
2601        through by mistake were picked up later in the function.
2602    
2603     6. I had used a variable called "isblank" - this is a C99 function, causing
2604        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2605    
2606     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2607        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2608        so the version is shown; (c) stop "make test" echoing "./RunTest".
2609    
2610     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2611    
2612     9. The prototype for memmove() for systems that don't have it was using
2613        size_t, but the inclusion of the header that defines size_t was later. I've
2614        moved the #includes for the C headers earlier to avoid this.
2615    
2616    10. Added some adjustments to the code to make it easier to compiler on certain
2617        special systems:
2618    
2619          (a) Some "const" qualifiers were missing.
2620          (b) Added the macro EXPORT before all exported functions; by default this
2621              is defined to be empty.
2622          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2623              that it reads its output file name as an argument instead of writing
2624              to the standard output and assuming this can be redirected.
2625    
2626    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2627        class containing characters with values greater than 255, PCRE compilation
2628        went into a loop.
2629    
2630    12. A recursive reference to a subpattern that was within another subpattern
2631        that had a minimum quantifier of zero caused PCRE to crash. For example,
2632        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2633        recursion. If the recursively-called subpattern itself had a zero repeat,
2634        that was OK.
2635    
2636    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2637        buffer into which it was copied (for escape processing) was still set at
2638        1024, so long lines caused crashes.
2639    
2640    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2641        "internal error: code overflow...". This applied to any character class
2642        that was followed by a possessive quantifier.
2643    
2644    15. Modified the Makefile to add libpcre.la as a prerequisite for
2645        libpcreposix.la because I was told this is needed for a parallel build to
2646        work.
2647    
2648    16. If a pattern that contained .* following optional items at the start was
2649        studied, the wrong optimizing data was generated, leading to matching
2650        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2651        matching string must start with a or b or c. The correct conclusion for
2652        this pattern is that a match can start with any character.
2653    
2654    
2655    Version 4.4 13-Aug-03
2656    ---------------------
2657    
2658     1. In UTF-8 mode, a character class containing characters with values between
2659        127 and 255 was not handled correctly if the compiled pattern was studied.
2660        In fixing this, I have also improved the studying algorithm for such
2661        classes (slightly).
2662    
2663     2. Three internal functions had redundant arguments passed to them. Removal
2664        might give a very teeny performance improvement.
2665    
2666     3. Documentation bug: the value of the capture_top field in a callout is *one
2667        more than* the number of the hightest numbered captured substring.
2668    
2669     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2670        in incorrectly linking with a previously installed version. They now link
2671        explicitly with libpcre.la.
2672    
2673     5. configure.in no longer needs to recognize Cygwin specially.
2674    
2675     6. A problem in pcre.in for Windows platforms is fixed.
2676    
2677     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2678        pcretest, it used to include the size of the study block as part of its
2679        output. Unfortunately, the structure contains a field that has a different
2680        size on different hardware architectures. This meant that the tests that
2681        showed this size failed. As the block is currently always of a fixed size,
2682        this information isn't actually particularly useful in pcretest output, so
2683        I have just removed it.
2684    
2685     8. Three pre-processor statements accidentally did not start in column 1.
2686        Sadly, there are *still* compilers around that complain, even though
2687        standard C has not required this for well over a decade. Sigh.
2688    
2689     9. In pcretest, the code for checking callouts passed small integers in the
2690        callout_data field, which is a void * field. However, some picky compilers
2691        complained about the casts involved for this on 64-bit systems. Now
2692        pcretest passes the address of the small integer instead, which should get
2693        rid of the warnings.
2694    
2695    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2696        both compile and run time, and gives an error if an invalid UTF-8 sequence
2697        is found. There is a option for disabling this check in cases where the
2698        string is known to be correct and/or the maximum performance is wanted.
2699    
2700    11. In response to a bug report, I changed one line in Makefile.in from
2701    
2702            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2703        to
2704            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2705    
2706        to look similar to other lines, but I have no way of telling whether this
2707        is the right thing to do, as I do not use Windows. No doubt I'll get told
2708        if it's wrong...
2709    
2710    
2711    Version 4.3 21-May-03
2712    ---------------------
2713    
2714    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2715       Makefile.
2716    
2717    2. Some refactoring to improve the quality of the code:
2718    
2719       (i)   The utf8_table... variables are now declared "const".
2720    
2721       (ii)  The code for \cx, which used the "case flipping" table to upper case
2722             lower case letters, now just substracts 32. This is ASCII-specific,
2723             but the whole concept of \cx is ASCII-specific, so it seems
2724             reasonable.
2725    
2726       (iii) PCRE was using its character types table to recognize decimal and
2727             hexadecimal digits in the pattern. This is silly, because it handles
2728             only 0-9, a-f, and A-F, but the character types table is locale-
2729             specific, which means strange things might happen. A private
2730             table is now used for this - though it costs 256 bytes, a table is
2731             much faster than multiple explicit tests. Of course, the standard
2732             character types table is still used for matching digits in subject
2733             strings against \d.
2734    
2735       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2736             ending in _t are). So I've renamed it as ESC_tee.
2737    
2738    3. The first argument for regexec() in the POSIX wrapper should have been
2739       defined as "const".
2740    
2741    4. Changed pcretest to use malloc() for its buffers so that they can be
2742       Electric Fenced for debugging.
2743    
2744    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2745       to read one or more bytes before the start of the subject string. Often this
2746       had no effect on PCRE's behaviour, but in some circumstances it could
2747       provoke a segmentation fault.
2748    
2749    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2750       to try to read one or more bytes before the start of the subject string.
2751    
2752    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2753       UTF-8 support could misbehave in various ways if the subject string
2754       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2755       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2756       back over UTF-8 characters.)
2757    
2758    
2759    Version 4.2 14-Apr-03
2760    ---------------------
2761    
2762    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2763    
2764    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2765         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2766         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2767         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2768         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2769           and BUILD_EXEEXT
2770         Note: automatic setting of the BUILD variables is not yet working
2771         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2772           compile-time but not at link-time
2773         [LINK]: use for linking executables only
2774         make different versions for Windows and non-Windows
2775         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2776           libraries
2777         [LINK_FOR_BUILD]: new variable
2778         [OBJEXT]: use throughout
2779         [EXEEXT]: use throughout
2780         <winshared>: new target
2781         <wininstall>: new target
2782         <dftables.o>: use native compiler
2783         <dftables>: use native linker
2784         <install>: handle Windows platform correctly
2785         <clean>: ditto
2786         <check>: ditto
2787         copy DLL to top builddir before testing
2788    
2789       As part of these changes, -no-undefined was removed again. This was reported
2790       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2791       in any case.
2792    
2793    3. Some tidies to get rid of compiler warnings:
2794    
2795       . In the match_data structure, match_limit was an unsigned long int, whereas
2796         match_call_count was an int. I've made them both unsigned long ints.
2797    
2798       . In pcretest the fact that a const uschar * doesn't automatically cast to
2799         a void * provoked a warning.
2800    
2801       . Turning on some more compiler warnings threw up some "shadow" variables
2802         and a few more missing casts.
2803    
2804    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2805       option, a class that contained a single character with a value between 128
2806       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2807    
2808    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2809       option, a class that contained several characters, but with at least one
2810       whose value was between 128 and 255 caused PCRE to crash.
2811    
2812    
2813    Version 4.1 12-Mar-03
2814    ---------------------
2815    
2816    1. Compiling with gcc -pedantic found a couple of places where casts were
2817    needed, and a string in dftables.c that was longer than standard compilers are
2818    required to support.
2819    
2820    2. Compiling with Sun's compiler found a few more places where the code could
2821    be tidied up in order to avoid warnings.
2822    
2823    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2824    first of these names is deprecated in the latest Autoconf in favour of the name
2825    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2826    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2827    analogy I have changed it to CFLAGS_FOR_BUILD.
2828    
2829    4. Added -no-undefined to the linking command in the Makefile, because this is
2830    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2831    linking step for the pcreposix library.
2832    
2833    5. PCRE was failing to diagnose the case of two named groups with the same
2834    name.
2835    
2836    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2837    literal character that is needed in the subject for a match, and scans along to
2838    ensure that it is present before embarking on the full matching process. This
2839    saves time in cases of nested unlimited repeats that are never going to match.
2840    Problem: the scan can take a lot of time if the subject is very long (e.g.
2841    megabytes), thus penalizing straightforward matches. It is now done only if the
2842    amount of subject to be scanned is less than 1000 bytes.
2843    
2844    7. A lesser problem with the same optimization is that it was recording the
2845    first character of an anchored pattern as "needed", thus provoking a search
2846    right along the subject, even when the first match of the pattern was going to
2847    fail. The "needed" character is now not set for anchored patterns, unless it
2848    follows something in the pattern that is of non-fixed length. Thus, it still
2849    fulfils its original purpose of finding quick non-matches in cases of nested
2850    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2851    
2852    
2853    Version 4.0 17-Feb-03
2854    ---------------------
2855    
2856    1. If a comment in an extended regex that started immediately after a meta-item
2857    extended to the end of string, PCRE compiled incorrect data. This could lead to
2858    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2859    
2860    2. Moved to autoconf 2.53 and libtool 1.4.2.
2861    
2862    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2863    the special perltest8 script is no longer needed - all the tests can be run
2864    from a single perltest script.
2865    
2866    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2867    by \s. It has now been removed in PCRE. This means it isn't recognized as
2868    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2869    class [:space:] *does* include VT, thereby creating a mess.
2870    
2871    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2872    space and tab.
2873    
2874    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2875    its new features into the main test script, reducing the number of scripts.
2876    
2877    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2878    were backward compatible, and made the (?i) apply to the whole pattern, as if
2879    /i were given. Now it behaves more logically, and applies the option setting
2880    only to what follows. PCRE has been changed to follow suit. However, if it
2881    finds options settings right at the start of the pattern, it extracts them into
2882    the global options, as before. Thus, they show up in the info data.
2883    
2884    8. Added support for the \Q...\E escape sequence. Characters in between are
2885    treated as literals. This is slightly different from Perl in that $ and @ are
2886    also handled as literals inside the quotes. In Perl, they will cause variable
2887    interpolation. Note the following examples:
2888    
2889        Pattern            PCRE matches      Perl matches
2890    
2891        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2892        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2893        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2894    
2895    For compatibility with Perl, \Q...\E sequences are recognized inside character
2896    classes as well as outside them.
2897    
2898    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2899    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2900    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2901    signed/unsigned warnings.
2902    
2903    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2904    option for pcretest, so I've replaced it by a simple function that does just
2905    that job.
2906    
2907    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2908    "pcregrep -".
2909    
2910    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2911    Java package. This provides some syntactic sugar for simple cases of what my
2912    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2913    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2914    item, you can use this simplified notation. Note that only makes sense with
2915    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2916    greediness, whatever the setting of the PCRE_UNGREEDY option.
2917    
2918    13. A change of greediness default within a pattern was not taking effect at
2919    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2920    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2921    was abstracted outside.
2922    
2923    14. PCRE now supports the \G assertion. It is true when the current matching
2924    position is at the start point of the match. This differs from \A when the
2925    starting offset is non-zero. Used with the /g option of pcretest (or similar
2926    code), it works in the same way as it does for Perl's /g option. If all
2927    alternatives of a regex begin with \G, the expression is anchored to the start
2928    match position, and the "anchored" flag is set in the compiled expression.
2929    
2930    15. Some bugs concerning the handling of certain option changes within patterns
2931    have been fixed. These applied to options other than (?ims). For example,
2932    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2933    been the other way round. Some of this was related to change 7 above.
2934    
2935    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2936    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2937    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2938    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2939    
2940    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2941    mode. Unlike ".", it always matches newline, whatever the setting of
2942    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2943    assertions. Perl allows it, but it doesn't (in general) work because it can't
2944    calculate the length of the lookbehind. At least, that's the case for Perl
2945    5.8.0 - I've been told they are going to document that it doesn't work in
2946    future.
2947    
2948    18. Added an error diagnosis for escapes that PCRE does not support: these are
2949    \L, \l, \N, \P, \p, \U, \u, and \X.
2950    
2951    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2952    reading past the end of the pattern in cases such as /[abcd/.
2953    
2954    20. PCRE was getting more memory than necessary for patterns with classes that
2955    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2956    
2957    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2958    compiling PCRE for use with Virtual Pascal.
2959    
2960    22. Small fix to the Makefile to make it work properly if the build is done
2961    outside the source tree.
2962    
2963    23. Added a new extension: a condition to go with recursion. If a conditional
2964    subpattern starts with (?(R) the "true" branch is used if recursion has
2965    happened, whereas the "false" branch is used only at the top level.
2966    
2967    24. When there was a very long string of literal characters (over 255 bytes
2968    without UTF support, over 250 bytes with UTF support), the computation of how
2969    much memory was required could be incorrect, leading to segfaults or other
2970    strange effects.
2971    
2972    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2973    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2974    there was a subsequent back reference to those brackets. This meant that, for
2975    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2976    possible to check for precisely this case. All we can do is abandon the
2977    optimization if .* occurs inside capturing brackets when there are any back
2978    references whatsoever. (See below for a better fix that came later.)
2979    
2980    26. The handling of the optimization for finding the first character of a
2981    non-anchored pattern, and for finding a character that is required later in the
2982    match were failing in some cases. This didn't break the matching; it just
2983    failed to optimize when it could. The way this is done has been re-implemented.
2984    
2985    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2986    
2987    28. Added a new feature that provides some of the functionality that Perl
2988    provides with (?{...}). The facility is termed a "callout". The way it is done
2989    in PCRE is for the caller to provide an optional function, by setting
2990    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2991    global variable. By default it is unset, which disables all calling out. To get
2992    the function called, the regex must include (?C) at appropriate points. This
2993    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2994    This provides a means of identifying different callout points. When PCRE
2995    reaches such a point in the regex, if pcre_callout has been set, the external
2996    function is called. It is provided with data in a structure called
2997    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2998    matching continues; if it returns a non-zero value, the match at the current
2999    point fails. However, backtracking will occur if possible. [This was changed
3000    later and other features added - see item 49 below.]
3001    
3002    29. pcretest is upgraded to test the callout functionality. It provides a
3003    callout function that displays information. By default, it shows the start of
3004    the match and the current position in the text. There are some new data escapes
3005    to vary what happens:
3006    
3007        \C+         in addition, show current contents of captured substrings
3008        \C-         do not supply a callout function
3009        \C!n        return 1 when callout number n is reached
3010        \C!n!m      return 1 when callout number n is reached for the mth time
3011    
3012    30. If pcregrep was called with the -l option and just a single file name, it
3013    output "<stdin>" if a match was found, instead of the file name.
3014    
3015    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3016    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3017    pcre_exec(). This saves a malloc/free per call. The default value of
3018    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3019    when configuring.
3020    
3021    32. The default maximum size of a compiled pattern is 64K. There have been a
3022    few cases of people hitting this limit. The code now uses macros to handle the
3023    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3024    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3025    configuring. Tests 2 and 5 work only with 2-byte links because they output
3026    debugging information about compiled patterns.
3027    
3028    33. Internal code re-arrangements:
3029    
3030    (a) Moved the debugging function for printing out a compiled regex into
3031        its own source file (printint.c) and used #include to pull it into
3032        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3033        separate copies.
3034    
3035    (b) Defined the list of op-code names for debugging as a macro in
3036        internal.h so that it is next to the definition of the opcodes.
3037    
3038    (c) Defined a table of op-code lengths for simpler skipping along compiled
3039        code. This is again a macro in internal.h so that it is next to the
3040        definition of the opcodes.
3041    
3042    34. Added support for recursive calls to individual subpatterns, along the
3043    lines of Robin Houston's patch (but implemented somewhat differently).
3044    
3045    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3046    allow it to read and process whole directories in Win32. This code was
3047    contributed by Lionel Fourquaux; it has not been tested by me.
3048    
3049    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3050    used to name a group. Names consist of alphanumerics and underscores, and must
3051    be unique. Back references use the syntax (?P=name) and recursive calls use
3052    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3053    numbers. The function pcre_fullinfo() can be used after compilation to extract
3054    a name/number map. There are three relevant calls:
3055    
3056      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3057      PCRE_INFO_NAMECOUNT            yields the number of entries
3058      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3059    
3060    The map is a vector of fixed-size entries. The size of each entry depends on
3061    the length of the longest name used. The first two bytes of each entry are the
3062    group number, most significant byte first. There follows the corresponding
3063    name, zero terminated. The names are in alphabetical order.
3064    
3065    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3066    case instead of 255. Making it the same both with and without UTF-8 support
3067    means that the same test output works with both.
3068    
3069    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3070    calling malloc() with a zero argument.
3071    
3072    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3073    optimization. I've improved things by keeping a bitmap of backreferences with
3074    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3075    fact referenced, the optimization can be applied. It is unlikely that a
3076    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3077    the match to follow \n) will appear inside brackets with a number greater than
3078    31, but if it does, any back reference > 31 suppresses the optimization.
3079    
3080    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3081    of disabling numbered capturing parentheses. Any opening parenthesis that is
3082    not followed by ? behaves as if it were followed by ?: but named parentheses
3083    can still be used for capturing (and they will acquire numbers in the usual
3084    way).
3085    
3086    41. Redesigned the return codes from the match() function into yes/no/error so
3087    that errors can be passed back from deep inside the nested calls. A malloc
3088    failure while inside a recursive subpattern call now causes the
3089    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3090    
3091    42. It is now possible to set a limit on the number of times the match()
3092    function is called in a call to pcre_exec(). This facility makes it possible to
3093    limit the amount of recursion and backtracking, though not in a directly
3094    obvious way, because the match() function is used in a number of different
3095    circumstances. The count starts from zero for each position in the subject
3096    string (for non-anchored patterns). The default limit is, for compatibility, a
3097    large number, namely 10 000 000. You can change this in two ways:
3098    
3099    (a) When configuring PCRE before making, you can use --with-match-limit=n
3100        to set a default value for the compiled library.
3101    
3102    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3103        a different value is set. See 45 below.
3104    
3105    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3106    
3107    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3108    of things that can be changed at compile time. The first argument specifies
3109    what is wanted and the second points to where the information is to be placed.
3110    The current list of available information is:
3111    
3112      PCRE_CONFIG_UTF8
3113    
3114    The output is an integer that is set to one if UTF-8 support is available;
3115    otherwise it is set to zero.
3116    
3117      PCRE_CONFIG_NEWLINE
3118    
3119    The output is an integer that it set to the value of the code that is used for
3120    newline. It is either LF (10) or CR (13).
3121    
3122      PCRE_CONFIG_LINK_SIZE
3123    
3124    The output is an integer that contains the number of bytes used for internal
3125    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3126    
3127      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3128    
3129    The output is an integer that contains the threshold above which the POSIX
3130    interface uses malloc() for output vectors. See item 31 above.
3131    
3132      PCRE_CONFIG_MATCH_LIMIT
3133    
3134    The output is an unsigned integer that contains the default limit of the number
3135    of match() calls in a pcre_exec() execution. See 42 above.
3136    
3137    44. pcretest has been upgraded by the addition of the -C option. This causes it
3138    to extract all the available output from the new pcre_config() function, and to
3139    output it. The program then exits immediately.
3140    
3141    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3142    order to support additional features. One way would have been to define
3143    pcre_exec2() (for example) with extra arguments, but this would not have been
3144    extensible, and would also have required all calls to the original function to
3145    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3146    is used for passing in "extra" data from pcre_study().
3147    
3148    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3149    contains the following fields:
3150    
3151      flags         a bitmap indicating which of the following fields are set
3152      study_data    opaque data from pcre_study()
3153      match_limit   a way of specifying a limit on match() calls for a specific
3154                      call to pcre_exec()
3155      callout_data  data for callouts (see 49 below)
3156    
3157    The flag bits are also defined in pcre.h, and are
3158    
3159      PCRE_EXTRA_STUDY_DATA
3160      PCRE_EXTRA_MATCH_LIMIT
3161      PCRE_EXTRA_CALLOUT_DATA
3162    
3163    The pcre_study() function now returns one of these new pcre_extra blocks, with
3164    the actual study data pointed to by the study_data field, and the
3165    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3166    before. That is, this change is entirely upwards-compatible and requires no
3167    change to existing code.
3168    
3169    If you want to pass in additional data to pcre_exec(), you can either place it
3170    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3171    block.
3172    
3173    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3174    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3175    times with different match limits, until it finds the minimum value needed for
3176    pcre_exec() to complete. The value is then output. This can be instructive; for
3177    most simple matches the number is quite small, but for pathological cases it
3178    gets very large very quickly.
3179    
3180    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3181    returns the size of the data block pointed to by the study_data field in a
3182    pcre_extra block, that is, the value that was passed as the argument to
3183    pcre_malloc() when PCRE was getting memory in which to place the information
3184    created by pcre_study(). The fourth argument should point to a size_t variable.
3185    pcretest has been extended so that this information is shown after a successful
3186    pcre_study() call when information about the compiled regex is being displayed.
3187    
3188    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3189    because what follows is always an absolute path. (Later: it turns out that this
3190    is more than cosmetic for MinGW, because it doesn't like empty path
3191    components.)
3192    
3193    49. Some changes have been made to the callout feature (see 28 above):
3194    
3195    (i)  A callout function now has three choices for what it returns:
3196    
3197           0  =>  success, carry on matching
3198         > 0  =>  failure at this point, but backtrack if possible
3199         < 0  =>  serious error, return this value from pcre_exec()
3200    
3201         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3202         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3203         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3204         use by callout functions. It will never be used by PCRE itself.
3205    
3206    (ii) The pcre_extra structure (see 45 above) has a void * field called
3207         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3208         pcre_callout_block structure has a field of the same name. The contents of
3209         the field passed in the pcre_extra structure are passed to the callout
3210         function in the corresponding field in the callout block. This makes it
3211         easier to use the same callout-containing regex from multiple threads. For
3212         testing, the pcretest program has a new data escape
3213    
3214           \C*n        pass the number n (may be negative) as callout_data
3215    
3216         If the callout function in pcretest receives a non-zero value as
3217         callout_data, it returns that value.
3218    
3219    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3220    there were some redundant $(CFLAGS) in commands that are now specified as
3221    $(LINK), which already includes $(CFLAGS).
3222    
3223    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3224    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3225    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3226    one-byte characters throughout. Note that case-insensitive matching applies
3227    only to characters whose values are less than 256. PCRE doesn't support the
3228    notion of cases for higher-valued characters.
3229    
3230    (i)   A character class whose characters are all within 0-255 is handled as
3231          a bit map, and the map is inverted for negative classes. Previously, a
3232          character > 255 always failed to match such a class; however it should
3233          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3234    
3235    (ii)  A negated character class with a single character < 255 is coded as
3236          "not this character" (OP_NOT). This wasn't working properly when the test
3237          character was multibyte, either singly or repeated.
3238    
3239    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3240          mode, for example: \x{100}{2,3}.
3241    
3242    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3243          singly or repeated) now correctly test multibyte characters. However,
3244          PCRE doesn't recognize any characters with values greater than 255 as
3245          digits, spaces, or word characters. Such characters always match \D, \S,
3246          and \W, and never match \d, \s, or \w.
3247    
3248    (v)   Classes may now contain characters and character ranges with values
3249          greater than 255. For example: [ab\x{100}-\x{400}].
3250    
3251    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3252          PCRE in UTF-8 mode.
3253    
3254    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3255    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3256    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3257    value.)
3258    
3259    53. The single man page has become too large. I have therefore split it up into
3260    a number of separate man pages. These also give rise to individual HTML pages;
3261    these are now put in a separate directory, and there is an index.html page that
3262    lists them all. Some hyperlinking between the pages has been installed.
3263    
3264    54. Added convenience functions for handling named capturing parentheses.
3265    
3266    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3267    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3268    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3269    are faulted.
3270    
3271    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3272    calling configure. These values are used when compiling the dftables.c program
3273    which is run to generate the source of the default character tables. They
3274    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3275    you will need to set these values.
3276    
3277    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3278    
3279    
3280    Version 3.9 02-Jan-02
3281    ---------------------
3282    
3283    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3284    
3285    2. If --disable-static was given, the building process failed when trying to
3286    build pcretest and pcregrep. (For some reason it was using libtool to compile
3287    them, which is not right, as they aren't part of the library.)
3288    
3289    
3290    Version 3.8 18-Dec-01
3291    ---------------------
3292    
3293    1. The experimental UTF-8 code was completely screwed up. It was packing the
3294    bytes in the wrong order. How dumb can you get?
3295    
3296    
3297    Version 3.7 29-Oct-01
3298    ---------------------
3299    
3300    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3301    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3302    this didn't happen under Solaris 8, where I normally test things.
3303    
3304    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3305    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3306    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3307    this fix an hour or so after the initial 3.7 release.)
3308    
3309    
3310    Version 3.6 23-Oct-01
3311    ---------------------
3312    
3313    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3314    offsets passed as NULL with zero offset count.
3315    
3316    2. The config.guess and config.sub files had not been updated when I moved to
3317    the latest autoconf.
3318    
3319    
3320    Version 3.5 15-Aug-01
3321    ---------------------
3322    
3323    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3324    had been forgotten.
3325    
3326    2. By using declared but undefined structures, we can avoid using "void"
3327    definitions in pcre.h while keeping the internal definitions of the structures
3328    private.
3329    
3330    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3331    user point of view, this means that both static and shared libraries are built
3332    by default, but this can be individually controlled. More of the work of
3333    handling this static/shared cases is now inside libtool instead of PCRE's make
3334    file.
3335    
3336    4. The pcretest utility is now installed along with pcregrep because it is
3337    useful for users (to test regexs) and by doing this, it automatically gets
3338    relinked by libtool. The documentation has been turned into a man page, so
3339    there are now .1, .txt, and .html versions in /doc.
3340    
3341    5. Upgrades to pcregrep:
3342       (i)   Added long-form option names like gnu grep.
3343       (ii)  Added --help to list all options with an explanatory phrase.
3344       (iii) Added -r, --recursive to recurse into sub-directories.
3345       (iv)  Added -f, --file to read patterns from a file.
3346    
3347    6. pcre_exec() was referring to its "code" argument before testing that
3348    argument for NULL (and giving an error if it was NULL).
3349    
3350    7. Upgraded Makefile.in to allow for compiling in a different directory from
3351    the source directory.
3352    
3353    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3354    options bits, the pointer it was passed was to an int instead of to an unsigned
3355    long int. This mattered only on 64-bit systems.
3356    
3357    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3358    generated) instead of pcre.in, which it its source. Also made the same change
3359    in several of the .c files.
3360    
3361    10. A new release of gcc defines printf() as a macro, which broke pcretest
3362    because it had an ifdef in the middle of a string argument for printf(). Fixed
3363    by using separate calls to printf().
3364    
3365    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3366    script, to force use of CR or LF instead of \n in the source. On non-Unix
3367    systems, the value can be set in config.h.
3368    
3369    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3370    absolute limit. Changed the text of the error message to make this clear, and
3371    likewise updated the man page.
3372    
3373    13. The limit of 99 on the number of capturing subpatterns has been removed.
3374    The new limit is 65535, which I hope will not be a "real" limit.
3375    
3376    
3377    Version 3.4 22-Aug-00
3378    ---------------------
3379    
3380    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3381    
3382    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3383    
3384    
3385    Version 3.3 01-Aug-00
3386    ---------------------
3387    
3388    1. If an octal character was given, but the value was greater than \377, it
3389    was not getting masked to the least significant bits, as documented. This could
3390    lead to crashes in some systems.
3391    
3392    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3393    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3394    
3395    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3396    These just pass their arguments on to (pcre_free)(), but they are provided
3397    because some uses of PCRE bind it to non-C systems that can call its functions,
3398    but cannot call free() or pcre_free() directly.
3399    
3400    4. Add "make test" as a synonym for "make check". Corrected some comments in
3401    the Makefile.
3402    
3403    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3404    Makefile.
3405    
3406    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3407    command called pgrep for grepping around the active processes.
3408    
3409    7. Added the beginnings of support for UTF-8 character strings.
3410    
3411    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3412    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3413    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3414    out for the ar command.)
3415    
3416    
3417    Version 3.2 12-May-00
3418    ---------------------
3419    
3420    This is purely a bug fixing release.
3421    
3422    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3423    of ZA. This was just one example of several cases that could provoke this bug,
3424    which was introduced by change 9 of version 2.00. The code for breaking
3425    infinite loops after an iteration that matches an empty string was't working
3426    correctly.
3427    
3428    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3429    when matched against abbab (for example). After matching an empty string, it
3430    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3431    caused it to match further down the string than it should.
3432    
3433    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3434    was there because it doesn't seem to be needed, and it causes trouble on some
3435    systems, as it is not a Standard C header. It has been removed.
3436    
3437    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3438    were reported on the Macintosh. The changes were from
3439    
3440      while ((c = *(++ptr)) != 0 && c != '\n');
3441    to
3442      while ((c = *(++ptr)) != 0 && c != '\n') ;
3443    
3444    Totally extraordinary, but if that's what it takes...
3445    
3446    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3447    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3448    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3449    assumes the way PCRE uses memmove() (always moving upwards).
3450    
3451    6. PCRE is being used in one environment where strchr() is not available. There
3452    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3453    faster code anyway.
3454    
3455    
3456    Version 3.1 09-Feb-00
3457    ---------------------
3458    
3459    The only change in this release is the fixing of some bugs in Makefile.in for
3460    the "install" target:
3461    
3462    (1) It was failing to install pcreposix.h.
3463    
3464    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3465    
3466    
3467    Version 3.0 01-Feb-00
3468    ---------------------
3469    
3470    1. Add support for the /+ modifier to perltest (to output $` like it does in
3471    pcretest).
3472    
3473    2. Add support for the /g modifier to perltest.
3474    
3475    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3476    matches null strings.
3477    
3478    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3479    pattern. Perl treats empty patterns specially - it reuses the most recent
3480    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3481    effect.
3482    
3483    5. The POSIX interface was broken in that it was just handing over the POSIX
3484    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3485    required a bigger vector, with some working space on the end. This means that
3486    the POSIX wrapper now has to get and free some memory, and copy the results.
3487    
3488    6. Added some simple autoconf support, placing the test data and the
3489    documentation in separate directories, re-organizing some of the
3490    information files, and making it build pcre-config (a GNU standard). Also added
3491    libtool support for building PCRE as a shared library, which is now the
3492    default.
3493    
3494    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3495    09 are not valid octal constants. Single digits will be used for minor values
3496    less than 10.
3497    
3498    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3499    existing programs that set these in the POSIX interface can use PCRE without
3500    modification.
3501    
3502    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3503    return all that pcre_info() returns, plus additional data. The pcre_info()
3504    function is retained for compatibility, but is considered to be obsolete.
3505    
3506    10. Added experimental recursion feature (?R) to handle one common case that
3507    Perl 5.6 will be able to do with (?p{...}).
3508    
3509    11. Added support for POSIX character classes like [:alpha:], which Perl is
3510    adopting.
3511    
3512    
3513    Version 2.08 31-Aug-99
3514    ----------------------
3515    
3516    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3517    trying to match at the startoffset position, but instead was moving forward to
3518    the next newline as if a previous match had failed.
3519    
3520    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3521    and could get into a loop if a null string was matched other than at the start
3522    of the subject.
3523    
3524    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3525    be distinguished at compile time, and for completeness also added PCRE_DATE.
3526    
3527    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3528    in GnuWin32 environments.
3529    
3530    
3531    Version 2.07 29-Jul-99
3532    ----------------------
3533    
3534    1. The documentation is now supplied in plain text form and HTML as well as in
3535    the form of man page sources.
3536    
3537    2. C++ compilers don't like assigning (void *) values to other pointer types.
3538    In particular this affects malloc(). Although there is no problem in Standard
3539    C, I've put in casts to keep C++ compilers happy.
3540    
3541    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3542    should be (const char *).
3543    
3544    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3545    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3546    However, I haven't made this a standard facility. The documentation doesn't
3547    mention it, and the Makefile doesn't support it.
3548    
3549    5. The Makefile now contains an "install" target, with editable destinations at
3550    the top of the file. The pcretest program is not installed.
3551    
3552    6. pgrep -V now gives the PCRE version number and date.
3553    
3554    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3555    causing the entire string to be ignored, instead of just the last character.
3556    
3557    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3558    non-matching string, it can take a very, very long time, even for strings of
3559    quite modest length, because of the nested recursion. PCRE now does better in
3560    some of these cases. It does this by remembering the last required literal
3561    character in the pattern, and pre-searching the subject to ensure it is present
3562    before running the real match. In other words, it applies a heuristic to detect
3563    some types of certain failure quickly, and in the above example, if presented
3564    with a string that has no trailing " it gives "no match" very quickly.
3565    
3566    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3567    other alternatives are tried instead.
3568    
3569    
3570  Version 2.06 09-Jun-99  Version 2.06 09-Jun-99
3571  ----------------------  ----------------------

Legend:
Removed from v.35  
changed lines
  Added in v.565

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12