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

Diff of /code/trunk/ChangeLog

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

revision 85 by nigel, Sat Feb 24 21:41:13 2007 UTC revision 518 by ph10, Tue May 18 15:47:01 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.10 03 May-2010
5    ------------------------
6    
7    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8        THEN.
9    
10    2.  (*ACCEPT) was not working when inside an atomic group.
11    
12    3.  Inside a character class, \B is treated as a literal by default, but
13        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14        causes the error). The code is unchanged, but I tidied the documentation.
15    
16    4.  Inside a character class, PCRE always treated \R and \X as literals,
17        whereas Perl faults them if its -w option is set. I have changed PCRE so
18        that it faults them when PCRE_EXTRA is set.
19    
20    5.  Added support for \N, which always matches any character other than
21        newline. (It is the same as "." when PCRE_DOTALL is not set.)
22    
23    6.  When compiling pcregrep with newer versions of gcc which may have
24        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
25        declared with attribute warn_unused_result" were given. Just casting the
26        result to (void) does not stop the warnings; a more elaborate fudge is
27        needed. I've used a macro to implement this.
28    
29    7.  Minor change to pcretest.c to avoid a compiler warning.
30    
31    8.  Added four artifical Unicode properties to help with an option to make
32        \s etc use properties (see next item). The new properties are: Xan
33        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
34    
35    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
36        use Unicode properties. (*UCP) at the start of a pattern can be used to set
37        this option. Modified pcretest to add /W to test this facility. Added
38        REG_UCP to make it available via the POSIX interface.
39    
40    
41    Version 8.02 19-Mar-2010
42    ------------------------
43    
44    1.  The Unicode data tables have been updated to Unicode 5.2.0.
45    
46    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
47        configured.
48    
49    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
50        original author of that file, following a query about its status.
51    
52    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
53        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
54    
55    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
56        quantifier applied to a forward-referencing subroutine call, could compile
57        incorrect code or give the error "internal error: previously-checked
58        referenced subpattern not found".
59    
60    6.  Both MS Visual Studio and Symbian OS have problems with initializing
61        variables to point to external functions. For these systems, therefore,
62        pcre_malloc etc. are now initialized to local functions that call the
63        relevant global functions.
64    
65    7.  There were two entries missing in the vectors called coptable and poptable
66        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
67        I've fixed the data, and added a kludgy way of testing at compile time that
68        the lengths are correct (equal to the number of opcodes).
69    
70    8.  Following on from 7, I added a similar kludge to check the length of the
71        eint vector in pcreposix.c.
72    
73    9.  Error texts for pcre_compile() are held as one long string to avoid too
74        much relocation at load time. To find a text, the string is searched,
75        counting zeros. There was no check for running off the end of the string,
76        which could happen if a new error number was added without updating the
77        string.
78    
79    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
80    
81    11. \K was not working if it appeared in an atomic group or in a group that
82        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
83        \K is "not well defined" if used in an assertion. PCRE now accepts it if
84        the assertion is positive, but not if it is negative.
85    
86    12. Change 11 fortuitously reduced the size of the stack frame used in the
87        "match()" function of pcre_exec.c by one pointer. Forthcoming
88        implementation of support for (*MARK) will need an extra pointer on the
89        stack; I have reserved it now, so that the stack frame size does not
90        decrease.
91    
92    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
93        item in branch that calls a recursion is a subroutine call - as in the
94        second branch in the above example - was incorrectly given the compile-
95        time error "recursive call could loop indefinitely" because pcre_compile()
96        was not correctly checking the subroutine for matching a non-empty string.
97    
98    14. The checks for overrunning compiling workspace could trigger after an
99        overrun had occurred. This is a "should never occur" error, but it can be
100        triggered by pathological patterns such as hundreds of nested parentheses.
101        The checks now trigger 100 bytes before the end of the workspace.
102    
103    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
104    
105    
106    Version 8.01 19-Jan-2010
107    ------------------------
108    
109    1.  If a pattern contained a conditional subpattern with only one branch (in
110        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
111        computed the wrong minimum data length (which is of course zero for such
112        subpatterns). This could cause incorrect "no match" results.
113    
114    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
115        the pattern is reset in the first branch, pcre_compile() failed with
116        "internal error: code overflow at offset...". This happened only when
117        the reset was to the original external option setting. (An optimization
118        abstracts leading options settings into an external setting, which was the
119        cause of this.)
120    
121    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
122        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
123        assertion pattern did not match (meaning that the assertion was true), it
124        was incorrectly treated as false if the SKIP had been reached during the
125        matching. This also applied to assertions used as conditions.
126    
127    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
128        assertion subpattern, including such a pattern used as a condition,
129        unpredictable results occurred, instead of the error return
130        PCRE_ERROR_DFA_UITEM.
131    
132    5.  The C++ GlobalReplace function was not working like Perl for the special
133        situation when an empty string is matched. It now does the fancy magic
134        stuff that is necessary.
135    
136    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
137        removed. (These were left over from very, very early versions of PCRE.)
138    
139    7.  Some cosmetic changes to the code to make life easier when compiling it
140        as part of something else:
141    
142        (a) Change DEBUG to PCRE_DEBUG.
143    
144        (b) In pcre_compile(), rename the member of the "branch_chain" structure
145            called "current" as "current_branch", to prevent a collision with the
146            Linux macro when compiled as a kernel module.
147    
148        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
149            prevent a collision with the Linux macro when compiled as a kernel
150            module.
151    
152    8.  In pcre_compile() there are some checks for integer overflows that used to
153        cast potentially large values to (double). This has been changed to that
154        when building, a check for int64_t is made, and if it is found, it is used
155        instead, thus avoiding the use of floating point arithmetic. (There is no
156        other use of FP in PCRE.) If int64_t is not found, the fallback is to
157        double.
158    
159    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
160        2005 (difference between two addresses compared to an unsigned value).
161    
162    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
163        custom one, because of the following reported problem in Windows:
164    
165          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
166              under Win32.
167          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
168              therefore missing the function definition.
169          - The compiler thus generates a "C" signature for the test function.
170          - The linker fails to find the "C" function.
171          - PCRE fails to configure if asked to do so against libbz2.
172    
173    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
174        messages were output:
175    
176          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
177          rerunning libtoolize, to keep the correct libtool macros in-tree.
178          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
179    
180        I have done both of these things.
181    
182    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
183        most of the time, it *can* run out if it is given a pattern that contains a
184        runaway infinite recursion. I updated the discussion in the pcrestack man
185        page.
186    
187    13. Now that we have gone to the x.xx style of version numbers, the minor
188        version may start with zero. Using 08 or 09 is a bad idea because users
189        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
190        interpreted as invalid octal numbers. I've updated the previous comment in
191        configure.ac, and also added a check that gives an error if 08 or 09 are
192        used.
193    
194    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
195        causing partial matching to fail when the end of the subject matched \W
196        in a UTF-8 pattern where \W was quantified with a minimum of 3.
197    
198    15. There were some discrepancies between the declarations in pcre_internal.h
199        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
200        their definitions. The declarations used "const uschar *" and the
201        definitions used USPTR. Even though USPTR is normally defined as "const
202        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
203        reported that: "This difference in casting confuses some C++ compilers, for
204        example, SunCC recognizes above declarations as different functions and
205        generates broken code for hbpcre." I have changed the declarations to use
206        USPTR.
207    
208    16. GNU libtool is named differently on some systems. The autogen.sh script now
209        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
210        (FreeBSD).
211    
212    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
213        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
214        comment: "Figure out how to create a longlong from a string: strtoll and
215        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
216        instance, but it only takes 2 args instead of 3!"
217    
218    18. A subtle bug concerned with back references has been fixed by a change of
219        specification, with a corresponding code fix. A pattern such as
220        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
221        refers, was giving matches when it shouldn't. For example, xa=xaaa would
222        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
223        same bug. Such groups have to be quantified to be useful, or contained
224        inside another quantified group. (If there's no repetition, the reference
225        can never match.) The problem arises because, having left the group and
226        moved on to the rest of the pattern, a later failure that backtracks into
227        the group uses the captured value from the final iteration of the group
228        rather than the correct earlier one. I have fixed this in PCRE by forcing
229        any group that contains a reference to itself to be an atomic group; that
230        is, there cannot be any backtracking into it once it has completed. This is
231        similar to recursive and subroutine calls.
232    
233    
234    Version 8.00 19-Oct-09
235    ----------------------
236    
237    1.  The table for translating pcre_compile() error codes into POSIX error codes
238        was out-of-date, and there was no check on the pcre_compile() error code
239        being within the table. This could lead to an OK return being given in
240        error.
241    
242    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
243        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
244        in a Windows environment.
245    
246    3.  The pcregrep --count option prints the count for each file even when it is
247        zero, as does GNU grep. However, pcregrep was also printing all files when
248        --files-with-matches was added. Now, when both options are given, it prints
249        counts only for those files that have at least one match. (GNU grep just
250        prints the file name in this circumstance, but including the count seems
251        more useful - otherwise, why use --count?) Also ensured that the
252        combination -clh just lists non-zero counts, with no names.
253    
254    4.  The long form of the pcregrep -F option was incorrectly implemented as
255        --fixed_strings instead of --fixed-strings. This is an incompatible change,
256        but it seems right to fix it, and I didn't think it was worth preserving
257        the old behaviour.
258    
259    5.  The command line items --regex=pattern and --regexp=pattern were not
260        recognized by pcregrep, which required --regex pattern or --regexp pattern
261        (with a space rather than an '='). The man page documented the '=' forms,
262        which are compatible with GNU grep; these now work.
263    
264    6.  No libpcreposix.pc file was created for pkg-config; there was just
265        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
266    
267    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
268        when UCP support is not needed, by modifying the Python script that
269        generates it from Unicode data files. This should not matter if the module
270        is correctly used as a library, but I received one complaint about 50K of
271        unwanted data. My guess is that the person linked everything into his
272        program rather than using a library. Anyway, it does no harm.
273    
274    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
275        was a minimum greater than 1 for a wide character in a possessive
276        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
277        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
278        character. Chaos in the form of incorrect output or a compiling loop could
279        result.
280    
281    9.  The restrictions on what a pattern can contain when partial matching is
282        requested for pcre_exec() have been removed. All patterns can now be
283        partially matched by this function. In addition, if there are at least two
284        slots in the offset vector, the offset of the earliest inspected character
285        for the match and the offset of the end of the subject are set in them when
286        PCRE_ERROR_PARTIAL is returned.
287    
288    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
289        synonymous with PCRE_PARTIAL, for backwards compatibility, and
290        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
291        and may be more useful for multi-segment matching.
292    
293    11. Partial matching with pcre_exec() is now more intuitive. A partial match
294        used to be given if ever the end of the subject was reached; now it is
295        given only if matching could not proceed because another character was
296        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
297        string "Z", which now yields "no match" instead of "partial match". In the
298        case of pcre_dfa_exec(), "no match" is given if every matching path for the
299        final character ended with (*FAIL).
300    
301    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
302        if the pattern had a "must contain" character that was already found in the
303        earlier partial match, unless partial matching was again requested. For
304        example, with the pattern /dog.(body)?/, the "must contain" character is
305        "g". If the first part-match was for the string "dog", restarting with
306        "sbody" failed. This bug has been fixed.
307    
308    13. The string returned by pcre_dfa_exec() after a partial match has been
309        changed so that it starts at the first inspected character rather than the
310        first character of the match. This makes a difference only if the pattern
311        starts with a lookbehind assertion or \b or \B (\K is not supported by
312        pcre_dfa_exec()). It's an incompatible change, but it makes the two
313        matching functions compatible, and I think it's the right thing to do.
314    
315    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
316        so that the demonstration program is easily available in environments where
317        PCRE has not been installed from source.
318    
319    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
320        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
321        library.
322    
323    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
324        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
325        is not the first non-POSIX option to be added. Clearly some people find
326        these options useful.
327    
328    17. If a caller to the POSIX matching function regexec() passes a non-zero
329        value for nmatch with a NULL value for pmatch, the value of
330        nmatch is forced to zero.
331    
332    18. RunGrepTest did not have a test for the availability of the -u option of
333        the diff command, as RunTest does. It now checks in the same way as
334        RunTest, and also checks for the -b option.
335    
336    19. If an odd number of negated classes containing just a single character
337        interposed, within parentheses, between a forward reference to a named
338        subpattern and the definition of the subpattern, compilation crashed with
339        an internal error, complaining that it could not find the referenced
340        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
341        [The bug was that it was starting one character too far in when skipping
342        over the character class, thus treating the ] as data rather than
343        terminating the class. This meant it could skip too much.]
344    
345    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
346        /g option in pcretest when the pattern contains \K, which makes it possible
347        to have an empty string match not at the start, even when the pattern is
348        anchored. Updated pcretest and pcredemo to use this option.
349    
350    21. If the maximum number of capturing subpatterns in a recursion was greater
351        than the maximum at the outer level, the higher number was returned, but
352        with unset values at the outer level. The correct (outer level) value is
353        now given.
354    
355    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
356        PCRE did not set those parentheses (unlike Perl). I have now found a way to
357        make it do so. The string so far is captured, making this feature
358        compatible with Perl.
359    
360    23. The tests have been re-organized, adding tests 11 and 12, to make it
361        possible to check the Perl 5.10 features against Perl 5.10.
362    
363    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
364        pattern matches a fixed length string. PCRE did not allow this; now it
365        does. Neither allows recursion.
366    
367    25. I finally figured out how to implement a request to provide the minimum
368        length of subject string that was needed in order to match a given pattern.
369        (It was back references and recursion that I had previously got hung up
370        on.) This code has now been added to pcre_study(); it finds a lower bound
371        to the length of subject needed. It is not necessarily the greatest lower
372        bound, but using it to avoid searching strings that are too short does give
373        some useful speed-ups. The value is available to calling programs via
374        pcre_fullinfo().
375    
376    26. While implementing 25, I discovered to my embarrassment that pcretest had
377        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
378        study optimizations had never been tested with that matching function.
379        Oops. What is worse, even when it was passed study data, there was a bug in
380        pcre_dfa_exec() that meant it never actually used it. Double oops. There
381        were also very few tests of studied patterns with pcre_dfa_exec().
382    
383    27. If (?| is used to create subpatterns with duplicate numbers, they are now
384        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
385        on the other side of the coin, they are no longer allowed to have different
386        names, because these cannot be distinguished in PCRE, and this has caused
387        confusion. (This is a difference from Perl.)
388    
389    28. When duplicate subpattern names are present (necessarily with different
390        numbers, as required by 27 above), and a test is made by name in a
391        conditional pattern, either for a subpattern having been matched, or for
392        recursion in such a pattern, all the associated numbered subpatterns are
393        tested, and the overall condition is true if the condition is true for any
394        one of them. This is the way Perl works, and is also more like the way
395        testing by number works.
396    
397    
398    Version 7.9 11-Apr-09
399    ---------------------
400    
401    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
402        (pcretest), all targets were linked against these libraries. This included
403        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
404        libraries. This caused unwanted dependencies to be created. This problem
405        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
406        pcretest is linked with readline.
407    
408    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
409        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
410        moved outside it again, because FALSE and TRUE are already defined in AIX,
411        but BOOL is not.
412    
413    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
414        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
415    
416    4.  The pcregrep documentation said spaces were inserted as well as colons (or
417        hyphens) following file names and line numbers when outputting matching
418        lines. This is not true; no spaces are inserted. I have also clarified the
419        wording for the --colour (or --color) option.
420    
421    5.  In pcregrep, when --colour was used with -o, the list of matching strings
422        was not coloured; this is different to GNU grep, so I have changed it to be
423        the same.
424    
425    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
426        each matching line was coloured. Now it goes on to look for further matches
427        of any of the test patterns, which is the same behaviour as GNU grep.
428    
429    7.  A pattern that could match an empty string could cause pcregrep to loop; it
430        doesn't make sense to accept an empty string match in pcregrep, so I have
431        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
432        seems to be how GNU grep behaves.
433    
434    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
435        start or after a newline", because the conditional assertion was not being
436        correctly handled. The rule now is that both the assertion and what follows
437        in the first alternative must satisfy the test.
438    
439    9.  If auto-callout was enabled in a pattern with a conditional group whose
440        condition was an assertion, PCRE could crash during matching, both with
441        pcre_exec() and pcre_dfa_exec().
442    
443    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
444        used for matching.
445    
446    11. Unicode property support in character classes was not working for
447        characters (bytes) greater than 127 when not in UTF-8 mode.
448    
449    12. Added the -M command line option to pcretest.
450    
451    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
452    
453    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
454    
455    16. Added comments and documentation about mis-use of no_arg in the C++
456        wrapper.
457    
458    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
459        from Martin Jerabek that uses macro names for all relevant character and
460        string constants.
461    
462    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
463        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
464        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
465        these, but not everybody uses configure.
466    
467    19. A conditional group that had only one branch was not being correctly
468        recognized as an item that could match an empty string. This meant that an
469        enclosing group might also not be so recognized, causing infinite looping
470        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
471        with the subject "ab", where knowledge that the repeated group can match
472        nothing is needed in order to break the loop.
473    
474    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
475        exec(), but without supplying a callout function, matching went wrong.
476    
477    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
478        leak if the size of the offset vector was greater than 30. When the vector
479        is smaller, the saved offsets during recursion go onto a local stack
480        vector, but for larger vectors malloc() is used. It was failing to free
481        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
482        error, in fact).
483    
484    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
485        heapframe that is used only when UTF-8 support is enabled. This caused no
486        problem, but was untidy.
487    
488    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
489        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
490        included within another project.
491    
492    24. Steven Van Ingelgem's patches to add more options to the CMake support,
493        slightly modified by me:
494    
495          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
496              not building pcregrep.
497    
498          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
499              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
500    
501    25. Forward references, both numeric and by name, in patterns that made use of
502        duplicate group numbers, could behave incorrectly or give incorrect errors,
503        because when scanning forward to find the reference group, PCRE was not
504        taking into account the duplicate group numbers. A pattern such as
505        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
506    
507    26. Changed a few more instances of "const unsigned char *" to USPTR, making
508        the feature of a custom pointer more persuasive (as requested by a user).
509    
510    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
511        pcretest.c, inside #ifndefs, because it seems they are sometimes already
512        pre-defined.
513    
514    28. Added support for (*UTF8) at the start of a pattern.
515    
516    29. Arrange for flags added by the "release type" setting in CMake to be shown
517        in the configuration summary.
518    
519    
520    Version 7.8 05-Sep-08
521    ---------------------
522    
523    1.  Replaced UCP searching code with optimized version as implemented for Ad
524        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
525        stage table and inline lookup instead of a function, giving speed ups of 2
526        to 5 times on some simple patterns that I tested. Permission was given to
527        distribute the MultiStage2.py script that generates the tables (it's not in
528        the tarball, but is in the Subversion repository).
529    
530    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
531        scripts.
532    
533    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
534        a group with a zero qualifier. The result of the study could be incorrect,
535        or the function might crash, depending on the pattern.
536    
537    4.  Caseless matching was not working for non-ASCII characters in back
538        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
539        It now works when Unicode Property Support is available.
540    
541    5.  In pcretest, an escape such as \x{de} in the data was always generating
542        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
543        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
544        truncation.
545    
546    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
547    
548    7.  Added two (int) casts to pcregrep when printing the difference of two
549        pointers, in case they are 64-bit values.
550    
551    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
552        test 2 if it fails.
553    
554    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
555        and a #define of that name to empty if it is not externally set. This is to
556        allow users of MSVC to set it if necessary.
557    
558    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
559        the convenience functions in the pcre_get.c source file.
560    
561    11. An option change at the start of a pattern that had top-level alternatives
562        could cause overwriting and/or a crash. This command provoked a crash in
563        some environments:
564    
565          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
566    
567        This potential security problem was recorded as CVE-2008-2371.
568    
569    12. For a pattern where the match had to start at the beginning or immediately
570        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
571        pcre_dfa_exec() could read past the end of the passed subject if there was
572        no match. To help with detecting such bugs (e.g. with valgrind), I modified
573        pcretest so that it places the subject at the end of its malloc-ed buffer.
574    
575    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
576        exec() might read past the end of the data buffer in UTF-8 mode.
577    
578    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
579        the data contained the byte 0x85 as part of a UTF-8 character within its
580        first line. This applied both to normal and DFA matching.
581    
582    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
583        /^[^d]*?$/8 failed to match "abc".
584    
585    16. Added a missing copyright notice to pcrecpp_internal.h.
586    
587    17. Make it more clear in the documentation that values returned from
588        pcre_exec() in ovector are byte offsets, not character counts.
589    
590    18. Tidied a few places to stop certain compilers from issuing warnings.
591    
592    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
593        supplied by Stefan Weber. I made a further small update for 7.8 because
594        there is a change of source arrangements: the pcre_searchfuncs.c module is
595        replaced by pcre_ucd.c.
596    
597    
598    Version 7.7 07-May-08
599    ---------------------
600    
601    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
602        a string to a long long, pretend we don't even have a long long." This is
603        done by checking for the strtoq, strtoll, and _strtoi64 functions.
604    
605    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
606        pre-7.6 versions, which defined a global no_arg variable instead of putting
607        it in the RE class. (See also #8 below.)
608    
609    3.  Remove a line of dead code, identified by coverity and reported by Nuno
610        Lopes.
611    
612    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
613    
614        (1) The include/exclude patterns were being applied to the whole pathnames
615            of files, instead of just to the final components.
616    
617        (2) If there was more than one level of directory, the subdirectories were
618            skipped unless they satisfied the include/exclude conditions. This is
619            inconsistent with GNU grep (and could even be seen as contrary to the
620            pcregrep specification - which I improved to make it absolutely clear).
621            The action now is always to scan all levels of directory, and just
622            apply the include/exclude patterns to regular files.
623    
624    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
625        --exclude_dir in the tests to avoid scanning .svn directories.
626    
627    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
628        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
629        doesn't support NULs in patterns.
630    
631    7.  Added some missing "const"s to declarations of static tables in
632        pcre_compile.c and pcre_dfa_exec.c.
633    
634    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
635        caused by fix #2  above. (Subsequently also a second patch to fix the
636        first patch. And a third patch - this was a messy problem.)
637    
638    9.  Applied Craig's patch to remove the use of push_back().
639    
640    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
641        matching function regexec().
642    
643    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
644        which, however, unlike Perl's \g{...}, are subroutine calls, not back
645        references. PCRE supports relative numbers with this syntax (I don't think
646        Oniguruma does).
647    
648    12. Previously, a group with a zero repeat such as (...){0} was completely
649        omitted from the compiled regex. However, this means that if the group
650        was called as a subroutine from elsewhere in the pattern, things went wrong
651        (an internal error was given). Such groups are now left in the compiled
652        pattern, with a new opcode that causes them to be skipped at execution
653        time.
654    
655    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
656        to the way PCRE behaves:
657    
658        (a) A lone ] character is dis-allowed (Perl treats it as data).
659    
660        (b) A back reference to an unmatched subpattern matches an empty string
661            (Perl fails the current match path).
662    
663        (c) A data ] in a character class must be notated as \] because if the
664            first data character in a class is ], it defines an empty class. (In
665            Perl it is not possible to have an empty class.) The empty class []
666            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
667            The negative empty class [^] matches any one character, independently
668            of the DOTALL setting.
669    
670    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
671        non-existent subpattern following a character class starting with ']' and
672        containing () gave an internal compiling error instead of "reference to
673        non-existent subpattern". Fortunately, when the pattern did exist, the
674        compiled code was correct. (When scanning forwards to check for the
675        existencd of the subpattern, it was treating the data ']' as terminating
676        the class, so got the count wrong. When actually compiling, the reference
677        was subsequently set up correctly.)
678    
679    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
680        it was being rejected as not supported by pcre_dfa_exec(), even though
681        other assertions are supported. I have made pcre_dfa_exec() support
682        (*FAIL).
683    
684    16. The implementation of 13c above involved the invention of a new opcode,
685        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
686        cannot be changed at match time, I realized I could make a small
687        improvement to matching performance by compiling OP_ALLANY instead of
688        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
689        on the OP_ANY path.
690    
691    17. Compiling pcretest on Windows with readline support failed without the
692        following two fixes: (1) Make the unistd.h include conditional on
693        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
694    
695    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
696        ncurses library to be included for pcretest when ReadLine support is
697        requested, but also to allow for it to be overridden. This patch came from
698        Daniel Bergström.
699    
700    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
701        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
702        any errors with the current Unicode tables. Thanks to Peter Kankowski for
703        spotting this.
704    
705    
706    Version 7.6 28-Jan-08
707    ---------------------
708    
709    1.  A character class containing a very large number of characters with
710        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
711        overflow.
712    
713    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
714        HAVE_LONG_LONG is not defined.
715    
716    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
717        bring them up to date and include new features. This patch includes:
718    
719        - Fixed PH's badly added libz and libbz2 support.
720        - Fixed a problem with static linking.
721        - Added pcredemo. [But later removed - see 7 below.]
722        - Fixed dftables problem and added an option.
723        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
724            HAVE_LONG_LONG.
725        - Added readline support for pcretest.
726        - Added an listing of the option settings after cmake has run.
727    
728    4.  A user submitted a patch to Makefile that makes it easy to create
729        "pcre.dll" under mingw when using Configure/Make. I added stuff to
730        Makefile.am that cause it to include this special target, without
731        affecting anything else. Note that the same mingw target plus all
732        the other distribution libraries and programs are now supported
733        when configuring with CMake (see 6 below) instead of with
734        Configure/Make.
735    
736    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
737        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
738        exported in the Windows port". It has not yet been confirmed that the patch
739        solves the problem, but it does no harm.
740    
741    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
742        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
743        with CMake, and also correct the comment about stack recursion.
744    
745    7.  Remove the automatic building of pcredemo from the ./configure system and
746        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
747        of a program that users should build themselves after PCRE is installed, so
748        building it automatically is not really right. What is more, it gave
749        trouble in some build environments.
750    
751    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
752    
753    
754    Version 7.5 10-Jan-08
755    ---------------------
756    
757    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
758        values in parens when parsing an RE using the C++ wrapper."
759    
760    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
761        Characters greater than 255 were excluded from the class instead of being
762        included.
763    
764    3.  The same bug as (2) above applied to negated POSIX classes such as
765        [:^space:].
766    
767    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
768        defined or documented. It seems to have been a typo for PCRE_STATIC, so
769        I have changed it.
770    
771    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
772        first named subpattern) and a construct such as (?&a) would reference the
773        first named subpattern whose name started with "a" (in other words, the
774        length check was missing). Both these problems are fixed. "Subpattern name
775        expected" is now given for (?&) (a zero-length name), and this patch also
776        makes it give the same error for \k'' (previously it complained that that
777        was a reference to a non-existent subpattern).
778    
779    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
780        this is right because (?- can be followed by option settings as well as by
781        digits. I have, however, made the messages clearer.
782    
783    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
784        than the number used in the conditional) now cause a compile-time error.
785        This is actually not compatible with Perl, which accepts such patterns, but
786        treats the conditional as always being FALSE (as PCRE used to), but it
787        seems to me that giving a diagnostic is better.
788    
789    8.  Change "alphameric" to the more common word "alphanumeric" in comments
790        and messages.
791    
792    9.  Fix two occurrences of "backslash" in comments that should have been
793        "backspace".
794    
795    10. Remove two redundant lines of code that can never be obeyed (their function
796        was moved elsewhere).
797    
798    11. The program that makes PCRE's Unicode character property table had a bug
799        which caused it to generate incorrect table entries for sequences of
800        characters that have the same character type, but are in different scripts.
801        It amalgamated them into a single range, with the script of the first of
802        them. In other words, some characters were in the wrong script. There were
803        thirteen such cases, affecting characters in the following ranges:
804    
805          U+002b0 - U+002c1
806          U+0060c - U+0060d
807          U+0061e - U+00612
808          U+0064b - U+0065e
809          U+0074d - U+0076d
810          U+01800 - U+01805
811          U+01d00 - U+01d77
812          U+01d9b - U+01dbf
813          U+0200b - U+0200f
814          U+030fc - U+030fe
815          U+03260 - U+0327f
816          U+0fb46 - U+0fbb1
817          U+10450 - U+1049d
818    
819    12. The -o option (show only the matching part of a line) for pcregrep was not
820        compatible with GNU grep in that, if there was more than one match in a
821        line, it showed only the first of them. It now behaves in the same way as
822        GNU grep.
823    
824    13. If the -o and -v options were combined for pcregrep, it printed a blank
825        line for every non-matching line. GNU grep prints nothing, and pcregrep now
826        does the same. The return code can be used to tell if there were any
827        non-matching lines.
828    
829    14. Added --file-offsets and --line-offsets to pcregrep.
830    
831    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
832        infinitely looping recursion. The bug was that positive lookaheads were not
833        being skipped when checking for a possible empty match (negative lookaheads
834        and both kinds of lookbehind were skipped).
835    
836    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
837        inclusion of <windows.h> to before rather than after the definition of
838        INVALID_FILE_ATTRIBUTES (patch from David Byron).
839    
840    17. Specifying a possessive quantifier with a specific limit for a Unicode
841        character property caused pcre_compile() to compile bad code, which led at
842        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
843        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
844        caused the error; without that there was no problem.
845    
846    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
847    
848    19. Added --enable-pcretest-libreadline.
849    
850    20. In pcrecpp.cc, the variable 'count' was incremented twice in
851        RE::GlobalReplace(). As a result, the number of replacements returned was
852        double what it should be. I removed one of the increments, but Craig sent a
853        later patch that removed the other one (the right fix) and added unit tests
854        that check the return values (which was not done before).
855    
856    21. Several CMake things:
857    
858        (1) Arranged that, when cmake is used on Unix, the libraries end up with
859            the names libpcre and libpcreposix, not just pcre and pcreposix.
860    
861        (2) The above change means that pcretest and pcregrep are now correctly
862            linked with the newly-built libraries, not previously installed ones.
863    
864        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
865    
866    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
867        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
868        UTF-8 newline character). The key issue is that the pattern starts .*;
869        this means that the match must be either at the beginning, or after a
870        newline. The bug was in the code for advancing after a failed match and
871        checking that the new position followed a newline. It was not taking
872        account of UTF-8 characters correctly.
873    
874    23. PCRE was behaving differently from Perl in the way it recognized POSIX
875        character classes. PCRE was not treating the sequence [:...:] as a
876        character class unless the ... were all letters. Perl, however, seems to
877        allow any characters between [: and :], though of course it rejects as
878        unknown any "names" that contain non-letters, because all the known class
879        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
880        for example, whereas PCRE did not - it did not recognize a POSIX character
881        class. This seemed a bit dangerous, so the code has been changed to be
882        closer to Perl. The behaviour is not identical to Perl, because PCRE will
883        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
884        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
885        Perl does, and where it didn't before.
886    
887    24. Rewrite so as to remove the single use of %n from pcregrep because in some
888        Windows environments %n is disabled by default.
889    
890    
891    Version 7.4 21-Sep-07
892    ---------------------
893    
894    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
895        means that a class such as [\s] counted as "explicit reference to CR or
896        LF". That isn't really right - the whole point of the change was to try to
897        help when there was an actual mention of one of the two characters. So now
898        the change happens only if \r or \n (or a literal CR or LF) character is
899        encountered.
900    
901    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
902        of both had grown to the point where there were only 3 bits left.
903        Fortunately, there was spare space in the data structure, and so I have
904        moved the internal flags into a new 16-bit field to free up more option
905        bits.
906    
907    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
908        but did not set the internal JCHANGED flag - either of these is enough to
909        control the way the "get" function works - but the PCRE_INFO_JCHANGED
910        facility is supposed to tell if (?J) was ever used, so now (?J) at the
911        start sets both bits.
912    
913    4.  Added options (at build time, compile time, exec time) to change \R from
914        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
915    
916    5.  doc/pcresyntax.html was missing from the distribution.
917    
918    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
919        compatibility, even though it is no longer used.
920    
921    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
922        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
923        windows.h file is present (where different names are used). [This was
924        reversed later after testing - see 16 below.]
925    
926    8.  Changed all #include <config.h> to #include "config.h". There were also
927        some further <pcre.h> cases that I changed to "pcre.h".
928    
929    9.  When pcregrep was used with the --colour option, it missed the line ending
930        sequence off the lines that it output.
931    
932    10. It was pointed out to me that arrays of string pointers cause lots of
933        relocations when a shared library is dynamically loaded. A technique of
934        using a single long string with a table of offsets can drastically reduce
935        these. I have refactored PCRE in four places to do this. The result is
936        dramatic:
937    
938          Originally:                          290
939          After changing UCP table:            187
940          After changing error message table:   43
941          After changing table of "verbs"       36
942          After changing table of Posix names   22
943    
944        Thanks to the folks working on Gregex for glib for this insight.
945    
946    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
947        unicode-properties was also set.
948    
949    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
950    
951    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
952        checked only for CRLF.
953    
954    14. Added casts to pcretest.c to avoid compiler warnings.
955    
956    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
957    
958    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
959        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
960        entirely. This removes changes made in 7 above.
961    
962    17. The CMake files have been updated, and there is now more information about
963        building with CMake in the NON-UNIX-USE document.
964    
965    
966    Version 7.3 28-Aug-07
967    ---------------------
968    
969     1. In the rejigging of the build system that eventually resulted in 7.1, the
970        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
971        brackets there is not right, since it causes compilers to look for an
972        installed pcre.h, not the version that is in the source that is being
973        compiled (which of course may be different). I have changed it back to:
974    
975          #include "pcre.h"
976    
977        I have a vague recollection that the change was concerned with compiling in
978        different directories, but in the new build system, that is taken care of
979        by the VPATH setting the Makefile.
980    
981     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
982        when the subject happened to end in the byte 0x85 (e.g. if the last
983        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
984        characters but of course it shouldn't be taken as a newline when it is part
985        of another character. The bug was that, for an unlimited repeat of . in
986        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
987        characters when looking for a newline.
988    
989     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
990    
991     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
992        in debug output.
993    
994     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
995        long printing in the pcrecpp unittest when running under MinGW.
996    
997     6. ESC_K was left out of the EBCDIC table.
998    
999     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1000        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1001        limit also applies to "virtual nesting" when a pattern is recursive, and in
1002        this case 1000 isn't so big. I have been able to remove this limit at the
1003        expense of backing off one optimization in certain circumstances. Normally,
1004        when pcre_exec() would call its internal match() function recursively and
1005        immediately return the result unconditionally, it uses a "tail recursion"
1006        feature to save stack. However, when a subpattern that can match an empty
1007        string has an unlimited repetition quantifier, it no longer makes this
1008        optimization. That gives it a stack frame in which to save the data for
1009        checking that an empty string has been matched. Previously this was taken
1010        from the 1000-entry workspace that had been reserved. So now there is no
1011        explicit limit, but more stack is used.
1012    
1013     8. Applied Daniel's patches to solve problems with the import/export magic
1014        syntax that is required for Windows, and which was going wrong for the
1015        pcreposix and pcrecpp parts of the library. These were overlooked when this
1016        problem was solved for the main library.
1017    
1018     9. There were some crude static tests to avoid integer overflow when computing
1019        the size of patterns that contain repeated groups with explicit upper
1020        limits. As the maximum quantifier is 65535, the maximum group length was
1021        set at 30,000 so that the product of these two numbers did not overflow a
1022        32-bit integer. However, it turns out that people want to use groups that
1023        are longer than 30,000 bytes (though not repeat them that many times).
1024        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1025        made it possible to implement the integer overflow checks in a much more
1026        dynamic way, which I have now done. The artificial limitation on group
1027        length has been removed - we now have only the limit on the total length of
1028        the compiled pattern, which depends on the LINK_SIZE setting.
1029    
1030    10. Fixed a bug in the documentation for get/copy named substring when
1031        duplicate names are permitted. If none of the named substrings are set, the
1032        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1033        empty string.
1034    
1035    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1036        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1037        because the ] is interpreted as the first data character and the
1038        terminating ] is not found. PCRE has been made compatible with Perl in this
1039        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1040        cause memory overwriting.
1041    
1042    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1043        string has been matched (to stop an infinite loop). It was not recognizing
1044        a conditional subpattern that could match an empty string if that
1045        subpattern was within another subpattern. For example, it looped when
1046        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1047        condition was not nested. This bug has been fixed.
1048    
1049    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1050        past the start of the subject in the presence of bytes with the top bit
1051        set, for example "\x8aBCD".
1052    
1053    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1054        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1055    
1056    14. Optimized (?!) to (*FAIL).
1057    
1058    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1059        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1060        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1061        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1062        does: it's just the validity check that is more restrictive.
1063    
1064    16. Inserted checks for integer overflows during escape sequence (backslash)
1065        processing, and also fixed erroneous offset values for syntax errors during
1066        backslash processing.
1067    
1068    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1069        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1070    
1071    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1072        caused an overrun.
1073    
1074    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1075        something other than just ASCII characters) inside a group that had an
1076        unlimited repeat caused a loop at compile time (while checking to see
1077        whether the group could match an empty string).
1078    
1079    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1080        [\P{Any}] did so. (Error in the code for printing property names.)
1081    
1082    21. An orphan \E inside a character class could cause a crash.
1083    
1084    22. A repeated capturing bracket such as (A)? could cause a wild memory
1085        reference during compilation.
1086    
1087    23. There are several functions in pcre_compile() that scan along a compiled
1088        expression for various reasons (e.g. to see if it's fixed length for look
1089        behind). There were bugs in these functions when a repeated \p or \P was
1090        present in the pattern. These operators have additional parameters compared
1091        with \d, etc, and these were not being taken into account when moving along
1092        the compiled data. Specifically:
1093    
1094        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1095            length.
1096    
1097        (b) An item such as \pL+ within a repeated group could cause crashes or
1098            loops.
1099    
1100        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1101            "reference to non-existent subpattern" error.
1102    
1103        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1104    
1105    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1106        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1107    
1108    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1109    
1110    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1111        character were causing crashes (broken optimization).
1112    
1113    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1114        \p or \P) caused a compile-time loop.
1115    
1116    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1117        break. For example, the unstudied pattern [\r\n]A does not match the string
1118        "\r\nA" because change 7.0/46 below moves the current point on by two
1119        characters after failing to match at the start. However, the pattern \nA
1120        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1121        the same is true. There doesn't seem any very clean way out of this, but
1122        what I have chosen to do makes the common cases work: PCRE now takes note
1123        of whether there can be an explicit match for \r or \n anywhere in the
1124        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1125        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1126        pattern has explicit CR or LF references.
1127    
1128    29. Added (*CR) etc for changing newline setting at start of pattern.
1129    
1130    
1131    Version 7.2 19-Jun-07
1132    ---------------------
1133    
1134     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1135        which is apparently normally available under Windows.
1136    
1137     2. Re-jig the pcregrep tests with different newline settings in an attempt
1138        to make them independent of the local environment's newline setting.
1139    
1140     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1141    
1142     4. Some of the "internals" tests were previously cut out when the link size
1143        was not 2, because the output contained actual offsets. The recent new
1144        "Z" feature of pcretest means that these can be cut out, making the tests
1145        usable with all link sizes.
1146    
1147     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1148        stack recursion. This gives a massive performance boost under BSD, but just
1149        a small improvement under Linux. However, it saves one field in the frame
1150        in all cases.
1151    
1152     6. Added more features from the forthcoming Perl 5.10:
1153    
1154        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1155            recursion call. It refers to the nth most recently opened parentheses.
1156    
1157        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1158            to be opened parentheses.
1159    
1160        (c) Conditions that refer to capturing parentheses can be specified
1161            relatively, for example, (?(-2)... or (?(+3)...
1162    
1163        (d) \K resets the start of the current match so that everything before
1164            is not part of it.
1165    
1166        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1167    
1168        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1169            reference syntax.
1170    
1171        (g) (?| introduces a group in which the numbering of parentheses in each
1172            alternative starts with the same number.
1173    
1174        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1175    
1176     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1177        PCRE_INFO_JCHANGED.
1178    
1179     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1180        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1181        for detecting groups that can match an empty string.
1182    
1183     9. A pattern with a very large number of alternatives (more than several
1184        hundred) was running out of internal workspace during the pre-compile
1185        phase, where pcre_compile() figures out how much memory will be needed. A
1186        bit of new cunning has reduced the workspace needed for groups with
1187        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1188        workspace instead of running out of the 4096 that are available.
1189    
1190    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1191    
1192    11. Applied patch from Google to remove an optimization that didn't quite work.
1193        The report of the bug said:
1194    
1195          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1196          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1197          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1198    
1199    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1200        it matched the wrong number of bytes.
1201    
1202    
1203    Version 7.1 24-Apr-07
1204    ---------------------
1205    
1206     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1207        that is more "standard", making use of automake and other Autotools. There
1208        is some re-arrangement of the files and adjustment of comments consequent
1209        on this.
1210    
1211     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1212        for recursive directory scanning broke on some systems because the files
1213        are not scanned in any specific order and on different systems the order
1214        was different. A call to "sort" has been inserted into RunGrepTest for the
1215        approprate test as a short-term fix. In the longer term there may be an
1216        alternative.
1217    
1218     3. I had an email from Eric Raymond about problems translating some of PCRE's
1219        man pages to HTML (despite the fact that I distribute HTML pages, some
1220        people do their own conversions for various reasons). The problems
1221        concerned the use of low-level troff macros .br and .in. I have therefore
1222        removed all such uses from the man pages (some were redundant, some could
1223        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1224        HTML has been updated to handle .nf/.fi and to complain if it encounters
1225        .br or .in.
1226    
1227     4. Updated comments in configure.ac that get placed in config.h.in and also
1228        arranged for config.h to be included in the distribution, with the name
1229        config.h.generic, for the benefit of those who have to compile without
1230        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1231    
1232     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1233        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1234        makevp.bat for the current PCRE, using the additional files
1235        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1236    
1237     6. A Windows user reported a minor discrepancy with test 2, which turned out
1238        to be caused by a trailing space on an input line that had got lost in his
1239        copy. The trailing space was an accident, so I've just removed it.
1240    
1241     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1242        that is needed.
1243    
1244     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1245        as "const" (a) because they are and (b) because it helps the PHP
1246        maintainers who have recently made a script to detect big data structures
1247        in the php code that should be moved to the .rodata section. I remembered
1248        to update Builducptable as well, so it won't revert if ucptable.h is ever
1249        re-created.
1250    
1251     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1252        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1253        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1254        support is not required. This saves 1.5-2K of code, which is important in
1255        some applications.
1256    
1257        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1258        so as not to refer to the tables, even though these functions will never be
1259        called when UTF-8 support is disabled. Otherwise there are problems with a
1260        shared library.
1261    
1262    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1263    
1264        (a) It was defining its arguments as char * instead of void *.
1265    
1266        (b) It was assuming that all moves were upwards in memory; this was true
1267            a long time ago when I wrote it, but is no longer the case.
1268    
1269        The emulated memove() is provided for those environments that have neither
1270        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1271        is clearly not the case, as these two bugs were recently reported.
1272    
1273    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1274        and Detrail to create the HTML documentation, the .txt form of the man
1275        pages, and it removes trailing spaces from listed files. It also creates
1276        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1277        case, it wraps all the #defines with #ifndefs. This script should be run
1278        before "make dist".
1279    
1280    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1281        with Unicode property support.
1282    
1283        (a) For a maximizing quantifier, if the two different cases of the
1284            character were of different lengths in their UTF-8 codings (there are
1285            some cases like this - I found 11), and the matching function had to
1286            back up over a mixture of the two cases, it incorrectly assumed they
1287            were both the same length.
1288    
1289        (b) When PCRE was configured to use the heap rather than the stack for
1290            recursion during matching, it was not correctly preserving the data for
1291            the other case of a UTF-8 character when checking ahead for a match
1292            while processing a minimizing repeat. If the check also involved
1293            matching a wide character, but failed, corruption could cause an
1294            erroneous result when trying to check for a repeat of the original
1295            character.
1296    
1297    13. Some tidying changes to the testing mechanism:
1298    
1299        (a) The RunTest script now detects the internal link size and whether there
1300            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1301            values substituted by "configure". (The RunGrepTest script already did
1302            this for UTF-8.) The configure.ac script no longer substitutes the
1303            relevant variables.
1304    
1305        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1306            with length and offset values. This means that the output is different
1307            for different internal link sizes. Test 2 is skipped for link sizes
1308            other than 2 because of this, bypassing the problem. Unfortunately,
1309            there was also a test in test 3 (the locale tests) that used /B and
1310            failed for link sizes other than 2. Rather than cut the whole test out,
1311            I have added a new /Z option to pcretest that replaces the length and
1312            offset values with spaces. This is now used to make test 3 independent
1313            of link size. (Test 2 will be tidied up later.)
1314    
1315    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1316        segmentation fault instead of returning the appropriate error message.
1317    
1318    15. In multiline mode when the newline sequence was set to "any", the pattern
1319        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1320        This doesn't seem right; it now treats the CRLF combination as the line
1321        ending, and so does not match in that case. It's only a pattern such as ^$
1322        that would hit this one: something like ^ABC$ would have failed after \r
1323        and then tried again after \r\n.
1324    
1325    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1326        in an attempt to make files that differ only in their line terminators
1327        compare equal. This works on Linux.
1328    
1329    17. Under certain error circumstances pcregrep might try to free random memory
1330        as it exited. This is now fixed, thanks to valgrind.
1331    
1332    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1333        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1334        was because its rules for how to advance for /g after matching an empty
1335        string at the end of a line did not allow for this case. They now check for
1336        it specially.
1337    
1338    20. pcretest is supposed to handle patterns and data of any length, by
1339        extending its buffers when necessary. It was getting this wrong when the
1340        buffer for a data line had to be extended.
1341    
1342    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1343        CRLF as a newline sequence.
1344    
1345    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1346        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1347        I have nevertheless tidied it up.
1348    
1349    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1350    
1351    24. Added a man page for pcre-config.
1352    
1353    
1354    Version 7.0 19-Dec-06
1355    ---------------------
1356    
1357     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1358        moving to gcc 4.1.1.
1359    
1360     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1361        sys/time.h, which is documented as needed for this function. It doesn't
1362        seem to matter on Linux, but it showed up on some releases of OS X.
1363    
1364     3. It seems that there are systems where bytes whose values are greater than
1365        127 match isprint() in the "C" locale. The "C" locale should be the
1366        default when a C program starts up. In most systems, only ASCII printing
1367        characters match isprint(). This difference caused the output from pcretest
1368        to vary, making some of the tests fail. I have changed pcretest so that:
1369    
1370        (a) When it is outputting text in the compiled version of a pattern, bytes
1371            other than 32-126 are always shown as hex escapes.
1372    
1373        (b) When it is outputting text that is a matched part of a subject string,
1374            it does the same, unless a different locale has been set for the match
1375            (using the /L modifier). In this case, it uses isprint() to decide.
1376    
1377     4. Fixed a major bug that caused incorrect computation of the amount of memory
1378        required for a compiled pattern when options that changed within the
1379        pattern affected the logic of the preliminary scan that determines the
1380        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1381        that the computed length was too small. The symptoms of this bug were
1382        either the PCRE error "internal error: code overflow" from pcre_compile(),
1383        or a glibc crash with a message such as "pcretest: free(): invalid next
1384        size (fast)". Examples of patterns that provoked this bug (shown in
1385        pcretest format) are:
1386    
1387          /(?-x: )/x
1388          /(?x)(?-x: \s*#\s*)/
1389          /((?i)[\x{c0}])/8
1390          /(?i:[\x{c0}])/8
1391    
1392        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1393        is now done differently.
1394    
1395     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1396        wrapper classes; (b) implement a new function in the C++ scanner that is
1397        more efficient than the old way of doing things because it avoids levels of
1398        recursion in the regex matching; (c) add a paragraph to the documentation
1399        for the FullMatch() function.
1400    
1401     6. The escape sequence \n was being treated as whatever was defined as
1402        "newline". Not only was this contrary to the documentation, which states
1403        that \n is character 10 (hex 0A), but it also went horribly wrong when
1404        "newline" was defined as CRLF. This has been fixed.
1405    
1406     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1407        was being set to -1 for the "end of line" case (supposedly a value that no
1408        character can have). Though this value is never used (the check for end of
1409        line is "zero bytes in current character"), it caused compiler complaints.
1410        I've changed it to 0xffffffff.
1411    
1412     8. In pcre_version.c, the version string was being built by a sequence of
1413        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1414        string (as it is for production releases) called a macro with an empty
1415        argument. The C standard says the result of this is undefined. The gcc
1416        compiler treats it as an empty string (which was what was wanted) but it is
1417        reported that Visual C gives an error. The source has been hacked around to
1418        avoid this problem.
1419    
1420     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1421        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1422        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1423        of them did).
1424    
1425    10. Originally, pcretest opened its input and output without "b"; then I was
1426        told that "b" was needed in some environments, so it was added for release
1427        5.0 to both the input and output. (It makes no difference on Unix-like
1428        systems.) Later I was told that it is wrong for the input on Windows. I've
1429        now abstracted the modes into two macros, to make it easier to fiddle with
1430        them, and removed "b" from the input mode under Windows.
1431    
1432    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1433    
1434    12. Added -help and --help to pcretest as an official way of being reminded
1435        of the options.
1436    
1437    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1438        and pcrecpp.cc because they annoy compilers at high warning levels.
1439    
1440    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1441    
1442    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1443        scripts are not C programs :-) and which was not noticed because it works
1444        on Linux.
1445    
1446    16. pcretest is supposed to handle any length of pattern and data line (as one
1447        line or as a continued sequence of lines) by extending its input buffer if
1448        necessary. This feature was broken for very long pattern lines, leading to
1449        a string of junk being passed to pcre_compile() if the pattern was longer
1450        than about 50K.
1451    
1452    17. I have done a major re-factoring of the way pcre_compile() computes the
1453        amount of memory needed for a compiled pattern. Previously, there was code
1454        that made a preliminary scan of the pattern in order to do this. That was
1455        OK when PCRE was new, but as the facilities have expanded, it has become
1456        harder and harder to keep it in step with the real compile phase, and there
1457        have been a number of bugs (see for example, 4 above). I have now found a
1458        cunning way of running the real compile function in a "fake" mode that
1459        enables it to compute how much memory it would need, while actually only
1460        ever using a few hundred bytes of working memory and without too many
1461        tests of the mode. This should make future maintenance and development
1462        easier. A side effect of this work is that the limit of 200 on the nesting
1463        depth of parentheses has been removed (though this was never a serious
1464        limitation, I suspect). However, there is a downside: pcre_compile() now
1465        runs more slowly than before (30% or more, depending on the pattern). I
1466        hope this isn't a big issue. There is no effect on runtime performance.
1467    
1468    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1469        newline (only possible for the last line of a file) and it was a
1470        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1471    
1472    19. Added additional timing features to pcretest. (1) The -tm option now times
1473        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1474        separate command line item, by a number that specifies the number of
1475        repeats to use when timing. The default is 50000; this gives better
1476        precision, but takes uncomfortably long for very large patterns.
1477    
1478    20. Extended pcre_study() to be more clever in cases where a branch of a
1479        subpattern has no definite first character. For example, (a*|b*)[cd] would
1480        previously give no result from pcre_study(). Now it recognizes that the
1481        first character must be a, b, c, or d.
1482    
1483    21. There was an incorrect error "recursive call could loop indefinitely" if
1484        a subpattern (or the entire pattern) that was being tested for matching an
1485        empty string contained only one non-empty item after a nested subpattern.
1486        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1487        incorrectly, because the \d was being skipped in the check.
1488    
1489    22. The pcretest program now has a new pattern option /B and a command line
1490        option -b, which is equivalent to adding /B to every pattern. This causes
1491        it to show the compiled bytecode, without the additional information that
1492        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1493        is the same as /B/I).
1494    
1495    23. A new optimization is now able automatically to treat some sequences such
1496        as a*b as a*+b. More specifically, if something simple (such as a character
1497        or a simple class like \d) has an unlimited quantifier, and is followed by
1498        something that cannot possibly match the quantified thing, the quantifier
1499        is automatically "possessified".
1500    
1501    24. A recursive reference to a subpattern whose number was greater than 39
1502        went wrong under certain circumstances in UTF-8 mode. This bug could also
1503        have affected the operation of pcre_study().
1504    
1505    25. Realized that a little bit of performance could be had by replacing
1506        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1507    
1508    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1509    
1510    27. Possessive quantifiers such as a++ were previously implemented by turning
1511        them into atomic groups such as ($>a+). Now they have their own opcodes,
1512        which improves performance. This includes the automatically created ones
1513        from 23 above.
1514    
1515    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1516        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1517        the first matched character to be a colon. This applied both to named and
1518        numbered groups.
1519    
1520    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1521    
1522    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1523        building PCRE on Windows easier, so I have included it in the distribution.
1524    
1525    31. There is now a check in pcretest against a ridiculously large number being
1526        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1527        loop, the loop is abandoned.
1528    
1529    32. Forward references to subpatterns in conditions such as (?(2)...) where
1530        subpattern 2 is defined later cause pcre_compile() to search forwards in
1531        the pattern for the relevant set of parentheses. This search went wrong
1532        when there were unescaped parentheses in a character class, parentheses
1533        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1534    
1535    33. "Subroutine" calls and backreferences were previously restricted to
1536        referencing subpatterns earlier in the regex. This restriction has now
1537        been removed.
1538    
1539    34. Added a number of extra features that are going to be in Perl 5.10. On the
1540        whole, these are just syntactic alternatives for features that PCRE had
1541        previously implemented using the Python syntax or my own invention. The
1542        other formats are all retained for compatibility.
1543    
1544        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1545            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1546            also .NET compatible.
1547    
1548        (b) A recursion or subroutine call to a named group can now be defined as
1549            (?&name) as well as (?P>name).
1550    
1551        (c) A backreference to a named group can now be defined as \k<name> or
1552            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1553            5.10, are also .NET compatible.
1554    
1555        (d) A conditional reference to a named group can now use the syntax
1556            (?(<name>) or (?('name') as well as (?(name).
1557    
1558        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1559            groups (named and numbered) that are never evaluated inline, but can be
1560            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1561            is always false. There may be only one alternative in such a group.
1562    
1563        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1564            as the simple (?(R). The condition is true only if the most recent
1565            recursion is that of the given number or name. It does not search out
1566            through the entire recursion stack.
1567    
1568        (g) The escape \gN or \g{N} has been added, where N is a positive or
1569            negative number, specifying an absolute or relative reference.
1570    
1571    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1572        some "unreachable code" warnings.
1573    
1574    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1575        things, this adds five new scripts.
1576    
1577    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1578        There were also incompatibilities regarding the handling of \Q..\E inside
1579        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1580        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1581    
1582    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1583        matches an empty string, and forcibly breaks the loop. There were bugs in
1584        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1585        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1586        separate and independent bugs (that affected different cases) have been
1587        fixed.
1588    
1589    39. Refactored the code to abolish the use of different opcodes for small
1590        capturing bracket numbers. This is a tidy that I avoided doing when I
1591        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1592        The new approach is not only tidier, it makes it possible to reduce the
1593        memory needed to fix the previous bug (38).
1594    
1595    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1596        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1597        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1598        mode.
1599    
1600    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1601        report.
1602    
1603    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1604        copy construction and assignment in the C++ wrapper.
1605    
1606    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1607        couple of bugs that could have given wrong results in the "--newline=crlf"
1608        case.
1609    
1610    44. Added a number of casts and did some reorganization of signed/unsigned int
1611        variables following suggestions from Dair Grant. Also renamed the variable
1612        "this" as "item" because it is a C++ keyword.
1613    
1614    45. Arranged for dftables to add
1615    
1616          #include "pcre_internal.h"
1617    
1618        to pcre_chartables.c because without it, gcc 4.x may remove the array
1619        definition from the final binary if PCRE is built into a static library and
1620        dead code stripping is activated.
1621    
1622    46. For an unanchored pattern, if a match attempt fails at the start of a
1623        newline sequence, and the newline setting is CRLF or ANY, and the next two
1624        characters are CRLF, advance by two characters instead of one.
1625    
1626    
1627    Version 6.7 04-Jul-06
1628    ---------------------
1629    
1630     1. In order to handle tests when input lines are enormously long, pcretest has
1631        been re-factored so that it automatically extends its buffers when
1632        necessary. The code is crude, but this _is_ just a test program. The
1633        default size has been increased from 32K to 50K.
1634    
1635     2. The code in pcre_study() was using the value of the re argument before
1636        testing it for NULL. (Of course, in any sensible call of the function, it
1637        won't be NULL.)
1638    
1639     3. The memmove() emulation function in pcre_internal.h, which is used on
1640        systems that lack both memmove() and bcopy() - that is, hardly ever -
1641        was missing a "static" storage class specifier.
1642    
1643     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1644        containing an extended class (one that cannot be represented by a bitmap
1645        because it contains high-valued characters or Unicode property items, e.g.
1646        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1647        pattern, but PCRE should not loop if you do not (it no longer does).
1648        [Detail: two cases were found: (a) a repeated subpattern containing an
1649        extended class; (b) a recursive reference to a subpattern that followed a
1650        previous extended class. It wasn't skipping over the extended class
1651        correctly when UTF-8 mode was not set.]
1652    
1653     5. A negated single-character class was not being recognized as fixed-length
1654        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1655        compile error "lookbehind assertion is not fixed length".
1656    
1657     6. The RunPerlTest auxiliary script was showing an unexpected difference
1658        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1659        write a Perl script that can interpret lines of an input file either as
1660        byte characters or as UTF-8, which is what "perltest" was being required to
1661        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1662        can't do is switch easily at run time between having the "use utf8;" pragma
1663        or not. In the end, I fudged it by using the RunPerlTest script to insert
1664        "use utf8;" explicitly for the UTF-8 tests.
1665    
1666     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1667        the end of the subject string, contrary to the documentation and to what
1668        Perl does. This was true of both matching functions. Now it matches only at
1669        the start of the subject and immediately after *internal* newlines.
1670    
1671     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1672        a pointer to an int instead of a pointer to an unsigned long int. This
1673        caused problems on 64-bit systems.
1674    
1675     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1676        instance of the 'standard' template library not being so standard".
1677    
1678    10. There was no check on the number of named subpatterns nor the maximum
1679        length of a subpattern name. The product of these values is used to compute
1680        the size of the memory block for a compiled pattern. By supplying a very
1681        long subpattern name and a large number of named subpatterns, the size
1682        computation could be caused to overflow. This is now prevented by limiting
1683        the length of names to 32 characters, and the number of named subpatterns
1684        to 10,000.
1685    
1686    11. Subpatterns that are repeated with specific counts have to be replicated in
1687        the compiled pattern. The size of memory for this was computed from the
1688        length of the subpattern and the repeat count. The latter is limited to
1689        65535, but there was no limit on the former, meaning that integer overflow
1690        could in principle occur. The compiled length of a repeated subpattern is
1691        now limited to 30,000 bytes in order to prevent this.
1692    
1693    12. Added the optional facility to have named substrings with the same name.
1694    
1695    13. Added the ability to use a named substring as a condition, using the
1696        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1697        are numbers (not recommended). Forward references are permitted.
1698    
1699    14. Added forward references in named backreferences (if you see what I mean).
1700    
1701    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1702        pattern could run off the end of the subject. For example, the pattern
1703        "(?s)(.{1,5})"8 did this with the subject "ab".
1704    
1705    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1706        PCRE_CASELESS was set when matching characters that were quantified with ?
1707        or *.
1708    
1709    17. A character class other than a single negated character that had a minimum
1710        but no maximum quantifier - for example [ab]{6,} - was not handled
1711        correctly by pce_dfa_exec(). It would match only one character.
1712    
1713    18. A valid (though odd) pattern that looked like a POSIX character
1714        class but used an invalid character after [ (for example [[,abc,]]) caused
1715        pcre_compile() to give the error "Failed: internal error: code overflow" or
1716        in some cases to crash with a glibc free() error. This could even happen if
1717        the pattern terminated after [[ but there just happened to be a sequence of
1718        letters, a binary zero, and a closing ] in the memory that followed.
1719    
1720    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1721        over the years. Originally (before any Unicode support), just the bottom 8
1722        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1723        output from "man perlunicode" includes this:
1724    
1725          The regular expression compiler produces polymorphic opcodes.  That
1726          is, the pattern adapts to the data and automatically switches to
1727          the Unicode character scheme when presented with Unicode data--or
1728          instead uses a traditional byte scheme when presented with byte
1729          data.
1730    
1731        Sadly, a wide octal escape does not cause a switch, and in a string with
1732        no other multibyte characters, these octal escapes are treated as before.
1733        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1734        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1735        Unicode string.
1736    
1737        I have not perpetrated such confusion in PCRE. Up till now, it took just
1738        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1739        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1740        translate to the appropriate multibyte character.
1741    
1742    29. Applied some refactoring to reduce the number of warnings from Microsoft
1743        and Borland compilers. This has included removing the fudge introduced
1744        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1745        a warning about an unused variable.
1746    
1747    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1748        characters since release 4.0, because Perl (from release 5.004) does not.
1749        [Or at least, is documented not to: some releases seem to be in conflict
1750        with the documentation.] However, when a pattern was studied with
1751        pcre_study() and all its branches started with \s, PCRE still included VT
1752        as a possible starting character. Of course, this did no harm; it just
1753        caused an unnecessary match attempt.
1754    
1755    22. Removed a now-redundant internal flag bit that recorded the fact that case
1756        dependency changed within the pattern. This was once needed for "required
1757        byte" processing, but is no longer used. This recovers a now-scarce options
1758        bit. Also moved the least significant internal flag bit to the most-
1759        significant bit of the word, which was not previously used (hangover from
1760        the days when it was an int rather than a uint) to free up another bit for
1761        the future.
1762    
1763    23. Added support for CRLF line endings as well as CR and LF. As well as the
1764        default being selectable at build time, it can now be changed at runtime
1765        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1766        specify that it is scanning data with non-default line endings.
1767    
1768    24. Changed the definition of CXXLINK to make it agree with the definition of
1769        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1770    
1771    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1772        recursions. This makes a big different to stack usage for some patterns.
1773    
1774    26. If a subpattern containing a named recursion or subroutine reference such
1775        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1776        the space required for the compiled pattern went wrong and gave too small a
1777        value. Depending on the environment, this could lead to "Failed: internal
1778        error: code overflow at offset 49" or "glibc detected double free or
1779        corruption" errors.
1780    
1781    27. Applied patches from Google (a) to support the new newline modes and (b) to
1782        advance over multibyte UTF-8 characters in GlobalReplace.
1783    
1784    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1785        difference for some implementation of PCRE in some Windows version.
1786    
1787    29. Added some extra testing facilities to pcretest:
1788    
1789        \q<number>   in a data line sets the "match limit" value
1790        \Q<number>   in a data line sets the "match recursion limt" value
1791        -S <number>  sets the stack size, where <number> is in megabytes
1792    
1793        The -S option isn't available for Windows.
1794    
1795    
1796    Version 6.6 06-Feb-06
1797    ---------------------
1798    
1799     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1800        in pcreposix.h. I have copied the definition from pcre.h.
1801    
1802     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1803        because pcre.h is no longer a built file.
1804    
1805     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1806        not normally included in the compiled code.
1807    
1808    
1809    Version 6.5 01-Feb-06
1810    ---------------------
1811    
1812     1. When using the partial match feature with pcre_dfa_exec(), it was not
1813        anchoring the second and subsequent partial matches at the new starting
1814        point. This could lead to incorrect results. For example, with the pattern
1815        /1234/, partially matching against "123" and then "a4" gave a match.
1816    
1817     2. Changes to pcregrep:
1818    
1819        (a) All non-match returns from pcre_exec() were being treated as failures
1820            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1821            error message is output. Some extra information is given for the
1822            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1823            probably the only errors that are likely to be caused by users (by
1824            specifying a regex that has nested indefinite repeats, for instance).
1825            If there are more than 20 of these errors, pcregrep is abandoned.
1826    
1827        (b) A binary zero was treated as data while matching, but terminated the
1828            output line if it was written out. This has been fixed: binary zeroes
1829            are now no different to any other data bytes.
1830    
1831        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1832            used to set a locale for matching. The --locale=xxxx long option has
1833            been added (no short equivalent) to specify a locale explicitly on the
1834            pcregrep command, overriding the environment variables.
1835    
1836        (d) When -B was used with -n, some line numbers in the output were one less
1837            than they should have been.
1838    
1839        (e) Added the -o (--only-matching) option.
1840    
1841        (f) If -A or -C was used with -c (count only), some lines of context were
1842            accidentally printed for the final match.
1843    
1844        (g) Added the -H (--with-filename) option.
1845    
1846        (h) The combination of options -rh failed to suppress file names for files
1847            that were found from directory arguments.
1848    
1849        (i) Added the -D (--devices) and -d (--directories) options.
1850    
1851        (j) Added the -F (--fixed-strings) option.
1852    
1853        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1854    
1855        (l) Added the --colo(u)r option.
1856    
1857        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1858            is not present by default.
1859    
1860     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1861        items such as (?R) or (?1), when the recursion could match a number of
1862        alternatives. If it matched one of the alternatives, but subsequently,
1863        outside the recursion, there was a failure, the code tried to back up into
1864        the recursion. However, because of the way PCRE is implemented, this is not
1865        possible, and the result was an incorrect result from the match.
1866    
1867        In order to prevent this happening, the specification of recursion has
1868        been changed so that all such subpatterns are automatically treated as
1869        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1870    
1871     4. I had overlooked the fact that, in some locales, there are characters for
1872        which isalpha() is true but neither isupper() nor islower() are true. In
1873        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1874        and ordfeminine) are like this. This affected the treatment of \w and \W
1875        when they appeared in character classes, but not when they appeared outside
1876        a character class. The bit map for "word" characters is now created
1877        separately from the results of isalnum() instead of just taking it from the
1878        upper, lower, and digit maps. (Plus the underscore character, of course.)
1879    
1880     5. The above bug also affected the handling of POSIX character classes such as
1881        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1882        permanent tables. Instead, the bit maps for such a class were previously
1883        created as the appropriate unions of the upper, lower, and digit bitmaps.
1884        Now they are created by subtraction from the [[:word:]] class, which has
1885        its own bitmap.
1886    
1887     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1888        It is created by subtracting the vertical space characters (\x09, \x0a,
1889        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1890        subtraction was done in the overall bitmap for a character class, meaning
1891        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1892        be recognized. This bug has been fixed.
1893    
1894     7. Patches from the folks at Google:
1895    
1896          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1897          real life, but is still worth protecting against".
1898    
1899          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1900          regular expressions".
1901    
1902          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1903          have it.
1904    
1905          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1906          "configure" and the latter not, in order to fix a problem somebody had
1907          with compiling the Arg class on HP-UX.
1908    
1909          (e) Improve the error-handling of the C++ wrapper a little bit.
1910    
1911          (f) New tests for checking recursion limiting.
1912    
1913     8. The pcre_memmove() function, which is used only if the environment does not
1914        have a standard memmove() function (and is therefore rarely compiled),
1915        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1916        returning a result (though PCRE never actually uses the result).
1917    
1918     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1919        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1920        returned instead of calling malloc() with an overflowing number that would
1921        most likely cause subsequent chaos.
1922    
1923    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1924    
1925    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1926        with this option is matched, the nmatch and pmatch options of regexec() are
1927        ignored.
1928    
1929    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1930        provided in case anyone wants to the the POSIX interface with UTF-8
1931        strings.
1932    
1933    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1934        C++ linking (needed for some HP-UX environments).
1935    
1936    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1937        (unused parameter) and in the pcre_printint() function (omitted "default"
1938        switch label when the default is to do nothing).
1939    
1940    15. Added some code to make it possible, when PCRE is compiled as a C++
1941        library, to replace subject pointers for pcre_exec() with a smart pointer
1942        class, thus making it possible to process discontinuous strings.
1943    
1944    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1945        much the same function. They were added by different people who were trying
1946        to make PCRE easy to compile on non-Unix systems. It has been suggested
1947        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1948        for compiling on Windows systems. I have therefore replaced it with
1949        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1950        defaults to "extern" for C or "extern C" for C++, which works fine on
1951        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1952        SCOPE with something explicit in config.h. In addition:
1953    
1954        (a) pcreposix.h still had just "extern" instead of either of these macros;
1955            I have replaced it with PCRE_DATA_SCOPE.
1956    
1957        (b) Functions such as _pcre_xclass(), which are internal to the library,
1958            but external in the C sense, all had PCRE_EXPORT in their definitions.
1959            This is apparently wrong for the Windows case, so I have removed it.
1960            (It makes no difference on Unix-like systems.)
1961    
1962    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1963        of recursive calls to match(). This is different to MATCH_LIMIT because
1964        that limits the total number of calls to match(), not all of which increase
1965        the depth of recursion. Limiting the recursion depth limits the amount of
1966        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1967        when PCRE is compiled, and changed at run time. A patch from Google adds
1968        this functionality to the C++ interface.
1969    
1970    18. Changes to the handling of Unicode character properties:
1971    
1972        (a) Updated the table to Unicode 4.1.0.
1973    
1974        (b) Recognize characters that are not in the table as "Cn" (undefined).
1975    
1976        (c) I revised the way the table is implemented to a much improved format
1977            which includes recognition of ranges. It now supports the ranges that
1978            are defined in UnicodeData.txt, and it also amalgamates other
1979            characters into ranges. This has reduced the number of entries in the
1980            table from around 16,000 to around 3,000, thus reducing its size
1981            considerably. I realized I did not need to use a tree structure after
1982            all - a binary chop search is just as efficient. Having reduced the
1983            number of entries, I extended their size from 6 bytes to 8 bytes to
1984            allow for more data.
1985    
1986        (d) Added support for Unicode script names via properties such as \p{Han}.
1987    
1988    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1989        matching that character.
1990    
1991    20. When matching a repeated Unicode property with a minimum greater than zero,
1992        (for example \pL{2,}), PCRE could look past the end of the subject if it
1993        reached it while seeking the minimum number of characters. This could
1994        happen only if some of the characters were more than one byte long, because
1995        there is a check for at least the minimum number of bytes.
1996    
1997    21. Refactored the implementation of \p and \P so as to be more general, to
1998        allow for more different types of property in future. This has changed the
1999        compiled form incompatibly. Anybody with saved compiled patterns that use
2000        \p or \P will have to recompile them.
2001    
2002    22. Added "Any" and "L&" to the supported property types.
2003    
2004    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2005        but give a compile time error if the value is greater than 0xff.
2006    
2007    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2008        accidentally not being installed or uninstalled.
2009    
2010    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2011        made were to insert the current release number. This seemed silly, because
2012        it made things harder for people building PCRE on systems that don't run
2013        "configure". I have turned pcre.h into a distributed file, no longer built
2014        by "configure", with the version identification directly included. There is
2015        no longer a pcre.h.in file.
2016    
2017        However, this change necessitated a change to the pcre-config script as
2018        well. It is built from pcre-config.in, and one of the substitutions was the
2019        release number. I have updated configure.ac so that ./configure now finds
2020        the release number by grepping pcre.h.
2021    
2022    26. Added the ability to run the tests under valgrind.
2023    
2024    
2025  Version 6.4 05-Sep-05  Version 6.4 05-Sep-05
2026  ---------------------  ---------------------
2027    

Legend:
Removed from v.85  
changed lines
  Added in v.518

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12