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

Legend:
Removed from v.85  
changed lines
  Added in v.513

  ViewVC Help
Powered by ViewVC 1.1.5