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

Diff of /code/trunk/ChangeLog

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

revision 83 by nigel, Sat Feb 24 21:41:06 2007 UTC revision 492 by ph10, Tue Mar 2 11:26:55 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.02 01-Mar-2010
5    ------------------------
6    
7    1.  The Unicode data tables have been updated to Unicode 5.2.0.
8    
9    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
10        configured.
11    
12    
13    Version 8.01 19-Jan-2010
14    ------------------------
15    
16    1.  If a pattern contained a conditional subpattern with only one branch (in
17        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
18        computed the wrong minimum data length (which is of course zero for such
19        subpatterns). This could cause incorrect "no match" results.
20    
21    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
22        the pattern is reset in the first branch, pcre_compile() failed with
23        "internal error: code overflow at offset...". This happened only when
24        the reset was to the original external option setting. (An optimization
25        abstracts leading options settings into an external setting, which was the
26        cause of this.)
27    
28    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
29        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
30        assertion pattern did not match (meaning that the assertion was true), it
31        was incorrectly treated as false if the SKIP had been reached during the
32        matching. This also applied to assertions used as conditions.
33    
34    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
35        assertion subpattern, including such a pattern used as a condition,
36        unpredictable results occurred, instead of the error return
37        PCRE_ERROR_DFA_UITEM.
38    
39    5.  The C++ GlobalReplace function was not working like Perl for the special
40        situation when an empty string is matched. It now does the fancy magic
41        stuff that is necessary.
42    
43    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
44        removed. (These were left over from very, very early versions of PCRE.)
45    
46    7.  Some cosmetic changes to the code to make life easier when compiling it
47        as part of something else:
48    
49        (a) Change DEBUG to PCRE_DEBUG.
50    
51        (b) In pcre_compile(), rename the member of the "branch_chain" structure
52            called "current" as "current_branch", to prevent a collision with the
53            Linux macro when compiled as a kernel module.
54    
55        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
56            prevent a collision with the Linux macro when compiled as a kernel
57            module.
58    
59    8.  In pcre_compile() there are some checks for integer overflows that used to
60        cast potentially large values to (double). This has been changed to that
61        when building, a check for int64_t is made, and if it is found, it is used
62        instead, thus avoiding the use of floating point arithmetic. (There is no
63        other use of FP in PCRE.) If int64_t is not found, the fallback is to
64        double.
65    
66    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
67        2005 (difference between two addresses compared to an unsigned value).
68    
69    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
70        custom one, because of the following reported problem in Windows:
71    
72          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
73              under Win32.
74          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
75              therefore missing the function definition.
76          - The compiler thus generates a "C" signature for the test function.
77          - The linker fails to find the "C" function.
78          - PCRE fails to configure if asked to do so against libbz2.
79    
80    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
81        messages were output:
82    
83          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
84          rerunning libtoolize, to keep the correct libtool macros in-tree.
85          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
86    
87        I have done both of these things.
88    
89    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
90        most of the time, it *can* run out if it is given a pattern that contains a
91        runaway infinite recursion. I updated the discussion in the pcrestack man
92        page.
93    
94    13. Now that we have gone to the x.xx style of version numbers, the minor
95        version may start with zero. Using 08 or 09 is a bad idea because users
96        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
97        interpreted as invalid octal numbers. I've updated the previous comment in
98        configure.ac, and also added a check that gives an error if 08 or 09 are
99        used.
100    
101    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
102        causing partial matching to fail when the end of the subject matched \W
103        in a UTF-8 pattern where \W was quantified with a minimum of 3.
104    
105    15. There were some discrepancies between the declarations in pcre_internal.h
106        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
107        their definitions. The declarations used "const uschar *" and the
108        definitions used USPTR. Even though USPTR is normally defined as "const
109        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
110        reported that: "This difference in casting confuses some C++ compilers, for
111        example, SunCC recognizes above declarations as different functions and
112        generates broken code for hbpcre." I have changed the declarations to use
113        USPTR.
114    
115    16. GNU libtool is named differently on some systems. The autogen.sh script now
116        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
117        (FreeBSD).
118    
119    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
120        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
121        comment: "Figure out how to create a longlong from a string: strtoll and
122        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
123        instance, but it only takes 2 args instead of 3!"
124    
125    18. A subtle bug concerned with back references has been fixed by a change of
126        specification, with a corresponding code fix. A pattern such as
127        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
128        refers, was giving matches when it shouldn't. For example, xa=xaaa would
129        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
130        same bug. Such groups have to be quantified to be useful, or contained
131        inside another quantified group. (If there's no repetition, the reference
132        can never match.) The problem arises because, having left the group and
133        moved on to the rest of the pattern, a later failure that backtracks into
134        the group uses the captured value from the final iteration of the group
135        rather than the correct earlier one. I have fixed this in PCRE by forcing
136        any group that contains a reference to itself to be an atomic group; that
137        is, there cannot be any backtracking into it once it has completed. This is
138        similar to recursive and subroutine calls.
139    
140    
141    Version 8.00 19-Oct-09
142    ----------------------
143    
144    1.  The table for translating pcre_compile() error codes into POSIX error codes
145        was out-of-date, and there was no check on the pcre_compile() error code
146        being within the table. This could lead to an OK return being given in
147        error.
148    
149    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
150        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
151        in a Windows environment.
152    
153    3.  The pcregrep --count option prints the count for each file even when it is
154        zero, as does GNU grep. However, pcregrep was also printing all files when
155        --files-with-matches was added. Now, when both options are given, it prints
156        counts only for those files that have at least one match. (GNU grep just
157        prints the file name in this circumstance, but including the count seems
158        more useful - otherwise, why use --count?) Also ensured that the
159        combination -clh just lists non-zero counts, with no names.
160    
161    4.  The long form of the pcregrep -F option was incorrectly implemented as
162        --fixed_strings instead of --fixed-strings. This is an incompatible change,
163        but it seems right to fix it, and I didn't think it was worth preserving
164        the old behaviour.
165    
166    5.  The command line items --regex=pattern and --regexp=pattern were not
167        recognized by pcregrep, which required --regex pattern or --regexp pattern
168        (with a space rather than an '='). The man page documented the '=' forms,
169        which are compatible with GNU grep; these now work.
170    
171    6.  No libpcreposix.pc file was created for pkg-config; there was just
172        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
173    
174    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
175        when UCP support is not needed, by modifying the Python script that
176        generates it from Unicode data files. This should not matter if the module
177        is correctly used as a library, but I received one complaint about 50K of
178        unwanted data. My guess is that the person linked everything into his
179        program rather than using a library. Anyway, it does no harm.
180    
181    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
182        was a minimum greater than 1 for a wide character in a possessive
183        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
184        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
185        character. Chaos in the form of incorrect output or a compiling loop could
186        result.
187    
188    9.  The restrictions on what a pattern can contain when partial matching is
189        requested for pcre_exec() have been removed. All patterns can now be
190        partially matched by this function. In addition, if there are at least two
191        slots in the offset vector, the offset of the earliest inspected character
192        for the match and the offset of the end of the subject are set in them when
193        PCRE_ERROR_PARTIAL is returned.
194    
195    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
196        synonymous with PCRE_PARTIAL, for backwards compatibility, and
197        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
198        and may be more useful for multi-segment matching.
199    
200    11. Partial matching with pcre_exec() is now more intuitive. A partial match
201        used to be given if ever the end of the subject was reached; now it is
202        given only if matching could not proceed because another character was
203        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
204        string "Z", which now yields "no match" instead of "partial match". In the
205        case of pcre_dfa_exec(), "no match" is given if every matching path for the
206        final character ended with (*FAIL).
207    
208    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
209        if the pattern had a "must contain" character that was already found in the
210        earlier partial match, unless partial matching was again requested. For
211        example, with the pattern /dog.(body)?/, the "must contain" character is
212        "g". If the first part-match was for the string "dog", restarting with
213        "sbody" failed. This bug has been fixed.
214    
215    13. The string returned by pcre_dfa_exec() after a partial match has been
216        changed so that it starts at the first inspected character rather than the
217        first character of the match. This makes a difference only if the pattern
218        starts with a lookbehind assertion or \b or \B (\K is not supported by
219        pcre_dfa_exec()). It's an incompatible change, but it makes the two
220        matching functions compatible, and I think it's the right thing to do.
221    
222    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
223        so that the demonstration program is easily available in environments where
224        PCRE has not been installed from source.
225    
226    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
227        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
228        library.
229    
230    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
231        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
232        is not the first non-POSIX option to be added. Clearly some people find
233        these options useful.
234    
235    17. If a caller to the POSIX matching function regexec() passes a non-zero
236        value for nmatch with a NULL value for pmatch, the value of
237        nmatch is forced to zero.
238    
239    18. RunGrepTest did not have a test for the availability of the -u option of
240        the diff command, as RunTest does. It now checks in the same way as
241        RunTest, and also checks for the -b option.
242    
243    19. If an odd number of negated classes containing just a single character
244        interposed, within parentheses, between a forward reference to a named
245        subpattern and the definition of the subpattern, compilation crashed with
246        an internal error, complaining that it could not find the referenced
247        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
248        [The bug was that it was starting one character too far in when skipping
249        over the character class, thus treating the ] as data rather than
250        terminating the class. This meant it could skip too much.]
251    
252    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
253        /g option in pcretest when the pattern contains \K, which makes it possible
254        to have an empty string match not at the start, even when the pattern is
255        anchored. Updated pcretest and pcredemo to use this option.
256    
257    21. If the maximum number of capturing subpatterns in a recursion was greater
258        than the maximum at the outer level, the higher number was returned, but
259        with unset values at the outer level. The correct (outer level) value is
260        now given.
261    
262    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
263        PCRE did not set those parentheses (unlike Perl). I have now found a way to
264        make it do so. The string so far is captured, making this feature
265        compatible with Perl.
266    
267    23. The tests have been re-organized, adding tests 11 and 12, to make it
268        possible to check the Perl 5.10 features against Perl 5.10.
269    
270    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
271        pattern matches a fixed length string. PCRE did not allow this; now it
272        does. Neither allows recursion.
273    
274    25. I finally figured out how to implement a request to provide the minimum
275        length of subject string that was needed in order to match a given pattern.
276        (It was back references and recursion that I had previously got hung up
277        on.) This code has now been added to pcre_study(); it finds a lower bound
278        to the length of subject needed. It is not necessarily the greatest lower
279        bound, but using it to avoid searching strings that are too short does give
280        some useful speed-ups. The value is available to calling programs via
281        pcre_fullinfo().
282    
283    26. While implementing 25, I discovered to my embarrassment that pcretest had
284        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
285        study optimizations had never been tested with that matching function.
286        Oops. What is worse, even when it was passed study data, there was a bug in
287        pcre_dfa_exec() that meant it never actually used it. Double oops. There
288        were also very few tests of studied patterns with pcre_dfa_exec().
289    
290    27. If (?| is used to create subpatterns with duplicate numbers, they are now
291        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
292        on the other side of the coin, they are no longer allowed to have different
293        names, because these cannot be distinguished in PCRE, and this has caused
294        confusion. (This is a difference from Perl.)
295    
296    28. When duplicate subpattern names are present (necessarily with different
297        numbers, as required by 27 above), and a test is made by name in a
298        conditional pattern, either for a subpattern having been matched, or for
299        recursion in such a pattern, all the associated numbered subpatterns are
300        tested, and the overall condition is true if the condition is true for any
301        one of them. This is the way Perl works, and is also more like the way
302        testing by number works.
303    
304    
305    Version 7.9 11-Apr-09
306    ---------------------
307    
308    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
309        (pcretest), all targets were linked against these libraries. This included
310        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
311        libraries. This caused unwanted dependencies to be created. This problem
312        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
313        pcretest is linked with readline.
314    
315    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
316        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
317        moved outside it again, because FALSE and TRUE are already defined in AIX,
318        but BOOL is not.
319    
320    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
321        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
322    
323    4.  The pcregrep documentation said spaces were inserted as well as colons (or
324        hyphens) following file names and line numbers when outputting matching
325        lines. This is not true; no spaces are inserted. I have also clarified the
326        wording for the --colour (or --color) option.
327    
328    5.  In pcregrep, when --colour was used with -o, the list of matching strings
329        was not coloured; this is different to GNU grep, so I have changed it to be
330        the same.
331    
332    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
333        each matching line was coloured. Now it goes on to look for further matches
334        of any of the test patterns, which is the same behaviour as GNU grep.
335    
336    7.  A pattern that could match an empty string could cause pcregrep to loop; it
337        doesn't make sense to accept an empty string match in pcregrep, so I have
338        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
339        seems to be how GNU grep behaves.
340    
341    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
342        start or after a newline", because the conditional assertion was not being
343        correctly handled. The rule now is that both the assertion and what follows
344        in the first alternative must satisfy the test.
345    
346    9.  If auto-callout was enabled in a pattern with a conditional group whose
347        condition was an assertion, PCRE could crash during matching, both with
348        pcre_exec() and pcre_dfa_exec().
349    
350    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
351        used for matching.
352    
353    11. Unicode property support in character classes was not working for
354        characters (bytes) greater than 127 when not in UTF-8 mode.
355    
356    12. Added the -M command line option to pcretest.
357    
358    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
359    
360    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
361    
362    16. Added comments and documentation about mis-use of no_arg in the C++
363        wrapper.
364    
365    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
366        from Martin Jerabek that uses macro names for all relevant character and
367        string constants.
368    
369    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
370        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
371        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
372        these, but not everybody uses configure.
373    
374    19. A conditional group that had only one branch was not being correctly
375        recognized as an item that could match an empty string. This meant that an
376        enclosing group might also not be so recognized, causing infinite looping
377        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
378        with the subject "ab", where knowledge that the repeated group can match
379        nothing is needed in order to break the loop.
380    
381    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
382        exec(), but without supplying a callout function, matching went wrong.
383    
384    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
385        leak if the size of the offset vector was greater than 30. When the vector
386        is smaller, the saved offsets during recursion go onto a local stack
387        vector, but for larger vectors malloc() is used. It was failing to free
388        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
389        error, in fact).
390    
391    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
392        heapframe that is used only when UTF-8 support is enabled. This caused no
393        problem, but was untidy.
394    
395    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
396        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
397        included within another project.
398    
399    24. Steven Van Ingelgem's patches to add more options to the CMake support,
400        slightly modified by me:
401    
402          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
403              not building pcregrep.
404    
405          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
406              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
407    
408    25. Forward references, both numeric and by name, in patterns that made use of
409        duplicate group numbers, could behave incorrectly or give incorrect errors,
410        because when scanning forward to find the reference group, PCRE was not
411        taking into account the duplicate group numbers. A pattern such as
412        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
413    
414    26. Changed a few more instances of "const unsigned char *" to USPTR, making
415        the feature of a custom pointer more persuasive (as requested by a user).
416    
417    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
418        pcretest.c, inside #ifndefs, because it seems they are sometimes already
419        pre-defined.
420    
421    28. Added support for (*UTF8) at the start of a pattern.
422    
423    29. Arrange for flags added by the "release type" setting in CMake to be shown
424        in the configuration summary.
425    
426    
427    Version 7.8 05-Sep-08
428    ---------------------
429    
430    1.  Replaced UCP searching code with optimized version as implemented for Ad
431        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
432        stage table and inline lookup instead of a function, giving speed ups of 2
433        to 5 times on some simple patterns that I tested. Permission was given to
434        distribute the MultiStage2.py script that generates the tables (it's not in
435        the tarball, but is in the Subversion repository).
436    
437    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
438        scripts.
439    
440    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
441        a group with a zero qualifier. The result of the study could be incorrect,
442        or the function might crash, depending on the pattern.
443    
444    4.  Caseless matching was not working for non-ASCII characters in back
445        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
446        It now works when Unicode Property Support is available.
447    
448    5.  In pcretest, an escape such as \x{de} in the data was always generating
449        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
450        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
451        truncation.
452    
453    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
454    
455    7.  Added two (int) casts to pcregrep when printing the difference of two
456        pointers, in case they are 64-bit values.
457    
458    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
459        test 2 if it fails.
460    
461    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
462        and a #define of that name to empty if it is not externally set. This is to
463        allow users of MSVC to set it if necessary.
464    
465    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
466        the convenience functions in the pcre_get.c source file.
467    
468    11. An option change at the start of a pattern that had top-level alternatives
469        could cause overwriting and/or a crash. This command provoked a crash in
470        some environments:
471    
472          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
473    
474        This potential security problem was recorded as CVE-2008-2371.
475    
476    12. For a pattern where the match had to start at the beginning or immediately
477        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
478        pcre_dfa_exec() could read past the end of the passed subject if there was
479        no match. To help with detecting such bugs (e.g. with valgrind), I modified
480        pcretest so that it places the subject at the end of its malloc-ed buffer.
481    
482    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
483        exec() might read past the end of the data buffer in UTF-8 mode.
484    
485    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
486        the data contained the byte 0x85 as part of a UTF-8 character within its
487        first line. This applied both to normal and DFA matching.
488    
489    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
490        /^[^d]*?$/8 failed to match "abc".
491    
492    16. Added a missing copyright notice to pcrecpp_internal.h.
493    
494    17. Make it more clear in the documentation that values returned from
495        pcre_exec() in ovector are byte offsets, not character counts.
496    
497    18. Tidied a few places to stop certain compilers from issuing warnings.
498    
499    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
500        supplied by Stefan Weber. I made a further small update for 7.8 because
501        there is a change of source arrangements: the pcre_searchfuncs.c module is
502        replaced by pcre_ucd.c.
503    
504    
505    Version 7.7 07-May-08
506    ---------------------
507    
508    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
509        a string to a long long, pretend we don't even have a long long." This is
510        done by checking for the strtoq, strtoll, and _strtoi64 functions.
511    
512    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
513        pre-7.6 versions, which defined a global no_arg variable instead of putting
514        it in the RE class. (See also #8 below.)
515    
516    3.  Remove a line of dead code, identified by coverity and reported by Nuno
517        Lopes.
518    
519    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
520    
521        (1) The include/exclude patterns were being applied to the whole pathnames
522            of files, instead of just to the final components.
523    
524        (2) If there was more than one level of directory, the subdirectories were
525            skipped unless they satisfied the include/exclude conditions. This is
526            inconsistent with GNU grep (and could even be seen as contrary to the
527            pcregrep specification - which I improved to make it absolutely clear).
528            The action now is always to scan all levels of directory, and just
529            apply the include/exclude patterns to regular files.
530    
531    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
532        --exclude_dir in the tests to avoid scanning .svn directories.
533    
534    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
535        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
536        doesn't support NULs in patterns.
537    
538    7.  Added some missing "const"s to declarations of static tables in
539        pcre_compile.c and pcre_dfa_exec.c.
540    
541    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
542        caused by fix #2  above. (Subsequently also a second patch to fix the
543        first patch. And a third patch - this was a messy problem.)
544    
545    9.  Applied Craig's patch to remove the use of push_back().
546    
547    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
548        matching function regexec().
549    
550    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
551        which, however, unlike Perl's \g{...}, are subroutine calls, not back
552        references. PCRE supports relative numbers with this syntax (I don't think
553        Oniguruma does).
554    
555    12. Previously, a group with a zero repeat such as (...){0} was completely
556        omitted from the compiled regex. However, this means that if the group
557        was called as a subroutine from elsewhere in the pattern, things went wrong
558        (an internal error was given). Such groups are now left in the compiled
559        pattern, with a new opcode that causes them to be skipped at execution
560        time.
561    
562    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
563        to the way PCRE behaves:
564    
565        (a) A lone ] character is dis-allowed (Perl treats it as data).
566    
567        (b) A back reference to an unmatched subpattern matches an empty string
568            (Perl fails the current match path).
569    
570        (c) A data ] in a character class must be notated as \] because if the
571            first data character in a class is ], it defines an empty class. (In
572            Perl it is not possible to have an empty class.) The empty class []
573            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
574            The negative empty class [^] matches any one character, independently
575            of the DOTALL setting.
576    
577    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
578        non-existent subpattern following a character class starting with ']' and
579        containing () gave an internal compiling error instead of "reference to
580        non-existent subpattern". Fortunately, when the pattern did exist, the
581        compiled code was correct. (When scanning forwards to check for the
582        existencd of the subpattern, it was treating the data ']' as terminating
583        the class, so got the count wrong. When actually compiling, the reference
584        was subsequently set up correctly.)
585    
586    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
587        it was being rejected as not supported by pcre_dfa_exec(), even though
588        other assertions are supported. I have made pcre_dfa_exec() support
589        (*FAIL).
590    
591    16. The implementation of 13c above involved the invention of a new opcode,
592        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
593        cannot be changed at match time, I realized I could make a small
594        improvement to matching performance by compiling OP_ALLANY instead of
595        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
596        on the OP_ANY path.
597    
598    17. Compiling pcretest on Windows with readline support failed without the
599        following two fixes: (1) Make the unistd.h include conditional on
600        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
601    
602    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
603        ncurses library to be included for pcretest when ReadLine support is
604        requested, but also to allow for it to be overridden. This patch came from
605        Daniel Bergström.
606    
607    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
608        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
609        any errors with the current Unicode tables. Thanks to Peter Kankowski for
610        spotting this.
611    
612    
613    Version 7.6 28-Jan-08
614    ---------------------
615    
616    1.  A character class containing a very large number of characters with
617        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
618        overflow.
619    
620    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
621        HAVE_LONG_LONG is not defined.
622    
623    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
624        bring them up to date and include new features. This patch includes:
625    
626        - Fixed PH's badly added libz and libbz2 support.
627        - Fixed a problem with static linking.
628        - Added pcredemo. [But later removed - see 7 below.]
629        - Fixed dftables problem and added an option.
630        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
631            HAVE_LONG_LONG.
632        - Added readline support for pcretest.
633        - Added an listing of the option settings after cmake has run.
634    
635    4.  A user submitted a patch to Makefile that makes it easy to create
636        "pcre.dll" under mingw when using Configure/Make. I added stuff to
637        Makefile.am that cause it to include this special target, without
638        affecting anything else. Note that the same mingw target plus all
639        the other distribution libraries and programs are now supported
640        when configuring with CMake (see 6 below) instead of with
641        Configure/Make.
642    
643    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
644        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
645        exported in the Windows port". It has not yet been confirmed that the patch
646        solves the problem, but it does no harm.
647    
648    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
649        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
650        with CMake, and also correct the comment about stack recursion.
651    
652    7.  Remove the automatic building of pcredemo from the ./configure system and
653        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
654        of a program that users should build themselves after PCRE is installed, so
655        building it automatically is not really right. What is more, it gave
656        trouble in some build environments.
657    
658    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
659    
660    
661    Version 7.5 10-Jan-08
662    ---------------------
663    
664    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
665        values in parens when parsing an RE using the C++ wrapper."
666    
667    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
668        Characters greater than 255 were excluded from the class instead of being
669        included.
670    
671    3.  The same bug as (2) above applied to negated POSIX classes such as
672        [:^space:].
673    
674    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
675        defined or documented. It seems to have been a typo for PCRE_STATIC, so
676        I have changed it.
677    
678    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
679        first named subpattern) and a construct such as (?&a) would reference the
680        first named subpattern whose name started with "a" (in other words, the
681        length check was missing). Both these problems are fixed. "Subpattern name
682        expected" is now given for (?&) (a zero-length name), and this patch also
683        makes it give the same error for \k'' (previously it complained that that
684        was a reference to a non-existent subpattern).
685    
686    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
687        this is right because (?- can be followed by option settings as well as by
688        digits. I have, however, made the messages clearer.
689    
690    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
691        than the number used in the conditional) now cause a compile-time error.
692        This is actually not compatible with Perl, which accepts such patterns, but
693        treats the conditional as always being FALSE (as PCRE used to), but it
694        seems to me that giving a diagnostic is better.
695    
696    8.  Change "alphameric" to the more common word "alphanumeric" in comments
697        and messages.
698    
699    9.  Fix two occurrences of "backslash" in comments that should have been
700        "backspace".
701    
702    10. Remove two redundant lines of code that can never be obeyed (their function
703        was moved elsewhere).
704    
705    11. The program that makes PCRE's Unicode character property table had a bug
706        which caused it to generate incorrect table entries for sequences of
707        characters that have the same character type, but are in different scripts.
708        It amalgamated them into a single range, with the script of the first of
709        them. In other words, some characters were in the wrong script. There were
710        thirteen such cases, affecting characters in the following ranges:
711    
712          U+002b0 - U+002c1
713          U+0060c - U+0060d
714          U+0061e - U+00612
715          U+0064b - U+0065e
716          U+0074d - U+0076d
717          U+01800 - U+01805
718          U+01d00 - U+01d77
719          U+01d9b - U+01dbf
720          U+0200b - U+0200f
721          U+030fc - U+030fe
722          U+03260 - U+0327f
723          U+0fb46 - U+0fbb1
724          U+10450 - U+1049d
725    
726    12. The -o option (show only the matching part of a line) for pcregrep was not
727        compatible with GNU grep in that, if there was more than one match in a
728        line, it showed only the first of them. It now behaves in the same way as
729        GNU grep.
730    
731    13. If the -o and -v options were combined for pcregrep, it printed a blank
732        line for every non-matching line. GNU grep prints nothing, and pcregrep now
733        does the same. The return code can be used to tell if there were any
734        non-matching lines.
735    
736    14. Added --file-offsets and --line-offsets to pcregrep.
737    
738    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
739        infinitely looping recursion. The bug was that positive lookaheads were not
740        being skipped when checking for a possible empty match (negative lookaheads
741        and both kinds of lookbehind were skipped).
742    
743    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
744        inclusion of <windows.h> to before rather than after the definition of
745        INVALID_FILE_ATTRIBUTES (patch from David Byron).
746    
747    17. Specifying a possessive quantifier with a specific limit for a Unicode
748        character property caused pcre_compile() to compile bad code, which led at
749        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
750        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
751        caused the error; without that there was no problem.
752    
753    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
754    
755    19. Added --enable-pcretest-libreadline.
756    
757    20. In pcrecpp.cc, the variable 'count' was incremented twice in
758        RE::GlobalReplace(). As a result, the number of replacements returned was
759        double what it should be. I removed one of the increments, but Craig sent a
760        later patch that removed the other one (the right fix) and added unit tests
761        that check the return values (which was not done before).
762    
763    21. Several CMake things:
764    
765        (1) Arranged that, when cmake is used on Unix, the libraries end up with
766            the names libpcre and libpcreposix, not just pcre and pcreposix.
767    
768        (2) The above change means that pcretest and pcregrep are now correctly
769            linked with the newly-built libraries, not previously installed ones.
770    
771        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
772    
773    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
774        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
775        UTF-8 newline character). The key issue is that the pattern starts .*;
776        this means that the match must be either at the beginning, or after a
777        newline. The bug was in the code for advancing after a failed match and
778        checking that the new position followed a newline. It was not taking
779        account of UTF-8 characters correctly.
780    
781    23. PCRE was behaving differently from Perl in the way it recognized POSIX
782        character classes. PCRE was not treating the sequence [:...:] as a
783        character class unless the ... were all letters. Perl, however, seems to
784        allow any characters between [: and :], though of course it rejects as
785        unknown any "names" that contain non-letters, because all the known class
786        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
787        for example, whereas PCRE did not - it did not recognize a POSIX character
788        class. This seemed a bit dangerous, so the code has been changed to be
789        closer to Perl. The behaviour is not identical to Perl, because PCRE will
790        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
791        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
792        Perl does, and where it didn't before.
793    
794    24. Rewrite so as to remove the single use of %n from pcregrep because in some
795        Windows environments %n is disabled by default.
796    
797    
798    Version 7.4 21-Sep-07
799    ---------------------
800    
801    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
802        means that a class such as [\s] counted as "explicit reference to CR or
803        LF". That isn't really right - the whole point of the change was to try to
804        help when there was an actual mention of one of the two characters. So now
805        the change happens only if \r or \n (or a literal CR or LF) character is
806        encountered.
807    
808    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
809        of both had grown to the point where there were only 3 bits left.
810        Fortunately, there was spare space in the data structure, and so I have
811        moved the internal flags into a new 16-bit field to free up more option
812        bits.
813    
814    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
815        but did not set the internal JCHANGED flag - either of these is enough to
816        control the way the "get" function works - but the PCRE_INFO_JCHANGED
817        facility is supposed to tell if (?J) was ever used, so now (?J) at the
818        start sets both bits.
819    
820    4.  Added options (at build time, compile time, exec time) to change \R from
821        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
822    
823    5.  doc/pcresyntax.html was missing from the distribution.
824    
825    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
826        compatibility, even though it is no longer used.
827    
828    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
829        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
830        windows.h file is present (where different names are used). [This was
831        reversed later after testing - see 16 below.]
832    
833    8.  Changed all #include <config.h> to #include "config.h". There were also
834        some further <pcre.h> cases that I changed to "pcre.h".
835    
836    9.  When pcregrep was used with the --colour option, it missed the line ending
837        sequence off the lines that it output.
838    
839    10. It was pointed out to me that arrays of string pointers cause lots of
840        relocations when a shared library is dynamically loaded. A technique of
841        using a single long string with a table of offsets can drastically reduce
842        these. I have refactored PCRE in four places to do this. The result is
843        dramatic:
844    
845          Originally:                          290
846          After changing UCP table:            187
847          After changing error message table:   43
848          After changing table of "verbs"       36
849          After changing table of Posix names   22
850    
851        Thanks to the folks working on Gregex for glib for this insight.
852    
853    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
854        unicode-properties was also set.
855    
856    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
857    
858    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
859        checked only for CRLF.
860    
861    14. Added casts to pcretest.c to avoid compiler warnings.
862    
863    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
864    
865    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
866        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
867        entirely. This removes changes made in 7 above.
868    
869    17. The CMake files have been updated, and there is now more information about
870        building with CMake in the NON-UNIX-USE document.
871    
872    
873    Version 7.3 28-Aug-07
874    ---------------------
875    
876     1. In the rejigging of the build system that eventually resulted in 7.1, the
877        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
878        brackets there is not right, since it causes compilers to look for an
879        installed pcre.h, not the version that is in the source that is being
880        compiled (which of course may be different). I have changed it back to:
881    
882          #include "pcre.h"
883    
884        I have a vague recollection that the change was concerned with compiling in
885        different directories, but in the new build system, that is taken care of
886        by the VPATH setting the Makefile.
887    
888     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
889        when the subject happened to end in the byte 0x85 (e.g. if the last
890        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
891        characters but of course it shouldn't be taken as a newline when it is part
892        of another character. The bug was that, for an unlimited repeat of . in
893        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
894        characters when looking for a newline.
895    
896     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
897    
898     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
899        in debug output.
900    
901     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
902        long printing in the pcrecpp unittest when running under MinGW.
903    
904     6. ESC_K was left out of the EBCDIC table.
905    
906     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
907        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
908        limit also applies to "virtual nesting" when a pattern is recursive, and in
909        this case 1000 isn't so big. I have been able to remove this limit at the
910        expense of backing off one optimization in certain circumstances. Normally,
911        when pcre_exec() would call its internal match() function recursively and
912        immediately return the result unconditionally, it uses a "tail recursion"
913        feature to save stack. However, when a subpattern that can match an empty
914        string has an unlimited repetition quantifier, it no longer makes this
915        optimization. That gives it a stack frame in which to save the data for
916        checking that an empty string has been matched. Previously this was taken
917        from the 1000-entry workspace that had been reserved. So now there is no
918        explicit limit, but more stack is used.
919    
920     8. Applied Daniel's patches to solve problems with the import/export magic
921        syntax that is required for Windows, and which was going wrong for the
922        pcreposix and pcrecpp parts of the library. These were overlooked when this
923        problem was solved for the main library.
924    
925     9. There were some crude static tests to avoid integer overflow when computing
926        the size of patterns that contain repeated groups with explicit upper
927        limits. As the maximum quantifier is 65535, the maximum group length was
928        set at 30,000 so that the product of these two numbers did not overflow a
929        32-bit integer. However, it turns out that people want to use groups that
930        are longer than 30,000 bytes (though not repeat them that many times).
931        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
932        made it possible to implement the integer overflow checks in a much more
933        dynamic way, which I have now done. The artificial limitation on group
934        length has been removed - we now have only the limit on the total length of
935        the compiled pattern, which depends on the LINK_SIZE setting.
936    
937    10. Fixed a bug in the documentation for get/copy named substring when
938        duplicate names are permitted. If none of the named substrings are set, the
939        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
940        empty string.
941    
942    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
943        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
944        because the ] is interpreted as the first data character and the
945        terminating ] is not found. PCRE has been made compatible with Perl in this
946        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
947        cause memory overwriting.
948    
949    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
950        string has been matched (to stop an infinite loop). It was not recognizing
951        a conditional subpattern that could match an empty string if that
952        subpattern was within another subpattern. For example, it looped when
953        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
954        condition was not nested. This bug has been fixed.
955    
956    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
957        past the start of the subject in the presence of bytes with the top bit
958        set, for example "\x8aBCD".
959    
960    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
961        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
962    
963    14. Optimized (?!) to (*FAIL).
964    
965    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
966        This restricts code points to be within the range 0 to 0x10FFFF, excluding
967        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
968        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
969        does: it's just the validity check that is more restrictive.
970    
971    16. Inserted checks for integer overflows during escape sequence (backslash)
972        processing, and also fixed erroneous offset values for syntax errors during
973        backslash processing.
974    
975    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
976        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
977    
978    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
979        caused an overrun.
980    
981    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
982        something other than just ASCII characters) inside a group that had an
983        unlimited repeat caused a loop at compile time (while checking to see
984        whether the group could match an empty string).
985    
986    20. Debugging a pattern containing \p or \P could cause a crash. For example,
987        [\P{Any}] did so. (Error in the code for printing property names.)
988    
989    21. An orphan \E inside a character class could cause a crash.
990    
991    22. A repeated capturing bracket such as (A)? could cause a wild memory
992        reference during compilation.
993    
994    23. There are several functions in pcre_compile() that scan along a compiled
995        expression for various reasons (e.g. to see if it's fixed length for look
996        behind). There were bugs in these functions when a repeated \p or \P was
997        present in the pattern. These operators have additional parameters compared
998        with \d, etc, and these were not being taken into account when moving along
999        the compiled data. Specifically:
1000    
1001        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1002            length.
1003    
1004        (b) An item such as \pL+ within a repeated group could cause crashes or
1005            loops.
1006    
1007        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1008            "reference to non-existent subpattern" error.
1009    
1010        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1011    
1012    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1013        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1014    
1015    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1016    
1017    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1018        character were causing crashes (broken optimization).
1019    
1020    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1021        \p or \P) caused a compile-time loop.
1022    
1023    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1024        break. For example, the unstudied pattern [\r\n]A does not match the string
1025        "\r\nA" because change 7.0/46 below moves the current point on by two
1026        characters after failing to match at the start. However, the pattern \nA
1027        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1028        the same is true. There doesn't seem any very clean way out of this, but
1029        what I have chosen to do makes the common cases work: PCRE now takes note
1030        of whether there can be an explicit match for \r or \n anywhere in the
1031        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1032        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1033        pattern has explicit CR or LF references.
1034    
1035    29. Added (*CR) etc for changing newline setting at start of pattern.
1036    
1037    
1038    Version 7.2 19-Jun-07
1039    ---------------------
1040    
1041     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1042        which is apparently normally available under Windows.
1043    
1044     2. Re-jig the pcregrep tests with different newline settings in an attempt
1045        to make them independent of the local environment's newline setting.
1046    
1047     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1048    
1049     4. Some of the "internals" tests were previously cut out when the link size
1050        was not 2, because the output contained actual offsets. The recent new
1051        "Z" feature of pcretest means that these can be cut out, making the tests
1052        usable with all link sizes.
1053    
1054     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1055        stack recursion. This gives a massive performance boost under BSD, but just
1056        a small improvement under Linux. However, it saves one field in the frame
1057        in all cases.
1058    
1059     6. Added more features from the forthcoming Perl 5.10:
1060    
1061        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1062            recursion call. It refers to the nth most recently opened parentheses.
1063    
1064        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1065            to be opened parentheses.
1066    
1067        (c) Conditions that refer to capturing parentheses can be specified
1068            relatively, for example, (?(-2)... or (?(+3)...
1069    
1070        (d) \K resets the start of the current match so that everything before
1071            is not part of it.
1072    
1073        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1074    
1075        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1076            reference syntax.
1077    
1078        (g) (?| introduces a group in which the numbering of parentheses in each
1079            alternative starts with the same number.
1080    
1081        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1082    
1083     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1084        PCRE_INFO_JCHANGED.
1085    
1086     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1087        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1088        for detecting groups that can match an empty string.
1089    
1090     9. A pattern with a very large number of alternatives (more than several
1091        hundred) was running out of internal workspace during the pre-compile
1092        phase, where pcre_compile() figures out how much memory will be needed. A
1093        bit of new cunning has reduced the workspace needed for groups with
1094        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1095        workspace instead of running out of the 4096 that are available.
1096    
1097    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1098    
1099    11. Applied patch from Google to remove an optimization that didn't quite work.
1100        The report of the bug said:
1101    
1102          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1103          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1104          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1105    
1106    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1107        it matched the wrong number of bytes.
1108    
1109    
1110    Version 7.1 24-Apr-07
1111    ---------------------
1112    
1113     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1114        that is more "standard", making use of automake and other Autotools. There
1115        is some re-arrangement of the files and adjustment of comments consequent
1116        on this.
1117    
1118     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1119        for recursive directory scanning broke on some systems because the files
1120        are not scanned in any specific order and on different systems the order
1121        was different. A call to "sort" has been inserted into RunGrepTest for the
1122        approprate test as a short-term fix. In the longer term there may be an
1123        alternative.
1124    
1125     3. I had an email from Eric Raymond about problems translating some of PCRE's
1126        man pages to HTML (despite the fact that I distribute HTML pages, some
1127        people do their own conversions for various reasons). The problems
1128        concerned the use of low-level troff macros .br and .in. I have therefore
1129        removed all such uses from the man pages (some were redundant, some could
1130        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1131        HTML has been updated to handle .nf/.fi and to complain if it encounters
1132        .br or .in.
1133    
1134     4. Updated comments in configure.ac that get placed in config.h.in and also
1135        arranged for config.h to be included in the distribution, with the name
1136        config.h.generic, for the benefit of those who have to compile without
1137        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1138    
1139     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1140        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1141        makevp.bat for the current PCRE, using the additional files
1142        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1143    
1144     6. A Windows user reported a minor discrepancy with test 2, which turned out
1145        to be caused by a trailing space on an input line that had got lost in his
1146        copy. The trailing space was an accident, so I've just removed it.
1147    
1148     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1149        that is needed.
1150    
1151     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1152        as "const" (a) because they are and (b) because it helps the PHP
1153        maintainers who have recently made a script to detect big data structures
1154        in the php code that should be moved to the .rodata section. I remembered
1155        to update Builducptable as well, so it won't revert if ucptable.h is ever
1156        re-created.
1157    
1158     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1159        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1160        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1161        support is not required. This saves 1.5-2K of code, which is important in
1162        some applications.
1163    
1164        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1165        so as not to refer to the tables, even though these functions will never be
1166        called when UTF-8 support is disabled. Otherwise there are problems with a
1167        shared library.
1168    
1169    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1170    
1171        (a) It was defining its arguments as char * instead of void *.
1172    
1173        (b) It was assuming that all moves were upwards in memory; this was true
1174            a long time ago when I wrote it, but is no longer the case.
1175    
1176        The emulated memove() is provided for those environments that have neither
1177        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1178        is clearly not the case, as these two bugs were recently reported.
1179    
1180    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1181        and Detrail to create the HTML documentation, the .txt form of the man
1182        pages, and it removes trailing spaces from listed files. It also creates
1183        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1184        case, it wraps all the #defines with #ifndefs. This script should be run
1185        before "make dist".
1186    
1187    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1188        with Unicode property support.
1189    
1190        (a) For a maximizing quantifier, if the two different cases of the
1191            character were of different lengths in their UTF-8 codings (there are
1192            some cases like this - I found 11), and the matching function had to
1193            back up over a mixture of the two cases, it incorrectly assumed they
1194            were both the same length.
1195    
1196        (b) When PCRE was configured to use the heap rather than the stack for
1197            recursion during matching, it was not correctly preserving the data for
1198            the other case of a UTF-8 character when checking ahead for a match
1199            while processing a minimizing repeat. If the check also involved
1200            matching a wide character, but failed, corruption could cause an
1201            erroneous result when trying to check for a repeat of the original
1202            character.
1203    
1204    13. Some tidying changes to the testing mechanism:
1205    
1206        (a) The RunTest script now detects the internal link size and whether there
1207            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1208            values substituted by "configure". (The RunGrepTest script already did
1209            this for UTF-8.) The configure.ac script no longer substitutes the
1210            relevant variables.
1211    
1212        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1213            with length and offset values. This means that the output is different
1214            for different internal link sizes. Test 2 is skipped for link sizes
1215            other than 2 because of this, bypassing the problem. Unfortunately,
1216            there was also a test in test 3 (the locale tests) that used /B and
1217            failed for link sizes other than 2. Rather than cut the whole test out,
1218            I have added a new /Z option to pcretest that replaces the length and
1219            offset values with spaces. This is now used to make test 3 independent
1220            of link size. (Test 2 will be tidied up later.)
1221    
1222    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1223        segmentation fault instead of returning the appropriate error message.
1224    
1225    15. In multiline mode when the newline sequence was set to "any", the pattern
1226        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1227        This doesn't seem right; it now treats the CRLF combination as the line
1228        ending, and so does not match in that case. It's only a pattern such as ^$
1229        that would hit this one: something like ^ABC$ would have failed after \r
1230        and then tried again after \r\n.
1231    
1232    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1233        in an attempt to make files that differ only in their line terminators
1234        compare equal. This works on Linux.
1235    
1236    17. Under certain error circumstances pcregrep might try to free random memory
1237        as it exited. This is now fixed, thanks to valgrind.
1238    
1239    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1240        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1241        was because its rules for how to advance for /g after matching an empty
1242        string at the end of a line did not allow for this case. They now check for
1243        it specially.
1244    
1245    20. pcretest is supposed to handle patterns and data of any length, by
1246        extending its buffers when necessary. It was getting this wrong when the
1247        buffer for a data line had to be extended.
1248    
1249    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1250        CRLF as a newline sequence.
1251    
1252    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1253        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1254        I have nevertheless tidied it up.
1255    
1256    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1257    
1258    24. Added a man page for pcre-config.
1259    
1260    
1261    Version 7.0 19-Dec-06
1262    ---------------------
1263    
1264     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1265        moving to gcc 4.1.1.
1266    
1267     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1268        sys/time.h, which is documented as needed for this function. It doesn't
1269        seem to matter on Linux, but it showed up on some releases of OS X.
1270    
1271     3. It seems that there are systems where bytes whose values are greater than
1272        127 match isprint() in the "C" locale. The "C" locale should be the
1273        default when a C program starts up. In most systems, only ASCII printing
1274        characters match isprint(). This difference caused the output from pcretest
1275        to vary, making some of the tests fail. I have changed pcretest so that:
1276    
1277        (a) When it is outputting text in the compiled version of a pattern, bytes
1278            other than 32-126 are always shown as hex escapes.
1279    
1280        (b) When it is outputting text that is a matched part of a subject string,
1281            it does the same, unless a different locale has been set for the match
1282            (using the /L modifier). In this case, it uses isprint() to decide.
1283    
1284     4. Fixed a major bug that caused incorrect computation of the amount of memory
1285        required for a compiled pattern when options that changed within the
1286        pattern affected the logic of the preliminary scan that determines the
1287        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1288        that the computed length was too small. The symptoms of this bug were
1289        either the PCRE error "internal error: code overflow" from pcre_compile(),
1290        or a glibc crash with a message such as "pcretest: free(): invalid next
1291        size (fast)". Examples of patterns that provoked this bug (shown in
1292        pcretest format) are:
1293    
1294          /(?-x: )/x
1295          /(?x)(?-x: \s*#\s*)/
1296          /((?i)[\x{c0}])/8
1297          /(?i:[\x{c0}])/8
1298    
1299        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1300        is now done differently.
1301    
1302     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1303        wrapper classes; (b) implement a new function in the C++ scanner that is
1304        more efficient than the old way of doing things because it avoids levels of
1305        recursion in the regex matching; (c) add a paragraph to the documentation
1306        for the FullMatch() function.
1307    
1308     6. The escape sequence \n was being treated as whatever was defined as
1309        "newline". Not only was this contrary to the documentation, which states
1310        that \n is character 10 (hex 0A), but it also went horribly wrong when
1311        "newline" was defined as CRLF. This has been fixed.
1312    
1313     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1314        was being set to -1 for the "end of line" case (supposedly a value that no
1315        character can have). Though this value is never used (the check for end of
1316        line is "zero bytes in current character"), it caused compiler complaints.
1317        I've changed it to 0xffffffff.
1318    
1319     8. In pcre_version.c, the version string was being built by a sequence of
1320        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1321        string (as it is for production releases) called a macro with an empty
1322        argument. The C standard says the result of this is undefined. The gcc
1323        compiler treats it as an empty string (which was what was wanted) but it is
1324        reported that Visual C gives an error. The source has been hacked around to
1325        avoid this problem.
1326    
1327     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1328        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1329        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1330        of them did).
1331    
1332    10. Originally, pcretest opened its input and output without "b"; then I was
1333        told that "b" was needed in some environments, so it was added for release
1334        5.0 to both the input and output. (It makes no difference on Unix-like
1335        systems.) Later I was told that it is wrong for the input on Windows. I've
1336        now abstracted the modes into two macros, to make it easier to fiddle with
1337        them, and removed "b" from the input mode under Windows.
1338    
1339    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1340    
1341    12. Added -help and --help to pcretest as an official way of being reminded
1342        of the options.
1343    
1344    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1345        and pcrecpp.cc because they annoy compilers at high warning levels.
1346    
1347    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1348    
1349    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1350        scripts are not C programs :-) and which was not noticed because it works
1351        on Linux.
1352    
1353    16. pcretest is supposed to handle any length of pattern and data line (as one
1354        line or as a continued sequence of lines) by extending its input buffer if
1355        necessary. This feature was broken for very long pattern lines, leading to
1356        a string of junk being passed to pcre_compile() if the pattern was longer
1357        than about 50K.
1358    
1359    17. I have done a major re-factoring of the way pcre_compile() computes the
1360        amount of memory needed for a compiled pattern. Previously, there was code
1361        that made a preliminary scan of the pattern in order to do this. That was
1362        OK when PCRE was new, but as the facilities have expanded, it has become
1363        harder and harder to keep it in step with the real compile phase, and there
1364        have been a number of bugs (see for example, 4 above). I have now found a
1365        cunning way of running the real compile function in a "fake" mode that
1366        enables it to compute how much memory it would need, while actually only
1367        ever using a few hundred bytes of working memory and without too many
1368        tests of the mode. This should make future maintenance and development
1369        easier. A side effect of this work is that the limit of 200 on the nesting
1370        depth of parentheses has been removed (though this was never a serious
1371        limitation, I suspect). However, there is a downside: pcre_compile() now
1372        runs more slowly than before (30% or more, depending on the pattern). I
1373        hope this isn't a big issue. There is no effect on runtime performance.
1374    
1375    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1376        newline (only possible for the last line of a file) and it was a
1377        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1378    
1379    19. Added additional timing features to pcretest. (1) The -tm option now times
1380        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1381        separate command line item, by a number that specifies the number of
1382        repeats to use when timing. The default is 50000; this gives better
1383        precision, but takes uncomfortably long for very large patterns.
1384    
1385    20. Extended pcre_study() to be more clever in cases where a branch of a
1386        subpattern has no definite first character. For example, (a*|b*)[cd] would
1387        previously give no result from pcre_study(). Now it recognizes that the
1388        first character must be a, b, c, or d.
1389    
1390    21. There was an incorrect error "recursive call could loop indefinitely" if
1391        a subpattern (or the entire pattern) that was being tested for matching an
1392        empty string contained only one non-empty item after a nested subpattern.
1393        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1394        incorrectly, because the \d was being skipped in the check.
1395    
1396    22. The pcretest program now has a new pattern option /B and a command line
1397        option -b, which is equivalent to adding /B to every pattern. This causes
1398        it to show the compiled bytecode, without the additional information that
1399        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1400        is the same as /B/I).
1401    
1402    23. A new optimization is now able automatically to treat some sequences such
1403        as a*b as a*+b. More specifically, if something simple (such as a character
1404        or a simple class like \d) has an unlimited quantifier, and is followed by
1405        something that cannot possibly match the quantified thing, the quantifier
1406        is automatically "possessified".
1407    
1408    24. A recursive reference to a subpattern whose number was greater than 39
1409        went wrong under certain circumstances in UTF-8 mode. This bug could also
1410        have affected the operation of pcre_study().
1411    
1412    25. Realized that a little bit of performance could be had by replacing
1413        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1414    
1415    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1416    
1417    27. Possessive quantifiers such as a++ were previously implemented by turning
1418        them into atomic groups such as ($>a+). Now they have their own opcodes,
1419        which improves performance. This includes the automatically created ones
1420        from 23 above.
1421    
1422    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1423        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1424        the first matched character to be a colon. This applied both to named and
1425        numbered groups.
1426    
1427    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1428    
1429    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1430        building PCRE on Windows easier, so I have included it in the distribution.
1431    
1432    31. There is now a check in pcretest against a ridiculously large number being
1433        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1434        loop, the loop is abandoned.
1435    
1436    32. Forward references to subpatterns in conditions such as (?(2)...) where
1437        subpattern 2 is defined later cause pcre_compile() to search forwards in
1438        the pattern for the relevant set of parentheses. This search went wrong
1439        when there were unescaped parentheses in a character class, parentheses
1440        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1441    
1442    33. "Subroutine" calls and backreferences were previously restricted to
1443        referencing subpatterns earlier in the regex. This restriction has now
1444        been removed.
1445    
1446    34. Added a number of extra features that are going to be in Perl 5.10. On the
1447        whole, these are just syntactic alternatives for features that PCRE had
1448        previously implemented using the Python syntax or my own invention. The
1449        other formats are all retained for compatibility.
1450    
1451        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1452            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1453            also .NET compatible.
1454    
1455        (b) A recursion or subroutine call to a named group can now be defined as
1456            (?&name) as well as (?P>name).
1457    
1458        (c) A backreference to a named group can now be defined as \k<name> or
1459            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1460            5.10, are also .NET compatible.
1461    
1462        (d) A conditional reference to a named group can now use the syntax
1463            (?(<name>) or (?('name') as well as (?(name).
1464    
1465        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1466            groups (named and numbered) that are never evaluated inline, but can be
1467            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1468            is always false. There may be only one alternative in such a group.
1469    
1470        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1471            as the simple (?(R). The condition is true only if the most recent
1472            recursion is that of the given number or name. It does not search out
1473            through the entire recursion stack.
1474    
1475        (g) The escape \gN or \g{N} has been added, where N is a positive or
1476            negative number, specifying an absolute or relative reference.
1477    
1478    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1479        some "unreachable code" warnings.
1480    
1481    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1482        things, this adds five new scripts.
1483    
1484    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1485        There were also incompatibilities regarding the handling of \Q..\E inside
1486        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1487        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1488    
1489    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1490        matches an empty string, and forcibly breaks the loop. There were bugs in
1491        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1492        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1493        separate and independent bugs (that affected different cases) have been
1494        fixed.
1495    
1496    39. Refactored the code to abolish the use of different opcodes for small
1497        capturing bracket numbers. This is a tidy that I avoided doing when I
1498        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1499        The new approach is not only tidier, it makes it possible to reduce the
1500        memory needed to fix the previous bug (38).
1501    
1502    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1503        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1504        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1505        mode.
1506    
1507    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1508        report.
1509    
1510    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1511        copy construction and assignment in the C++ wrapper.
1512    
1513    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1514        couple of bugs that could have given wrong results in the "--newline=crlf"
1515        case.
1516    
1517    44. Added a number of casts and did some reorganization of signed/unsigned int
1518        variables following suggestions from Dair Grant. Also renamed the variable
1519        "this" as "item" because it is a C++ keyword.
1520    
1521    45. Arranged for dftables to add
1522    
1523          #include "pcre_internal.h"
1524    
1525        to pcre_chartables.c because without it, gcc 4.x may remove the array
1526        definition from the final binary if PCRE is built into a static library and
1527        dead code stripping is activated.
1528    
1529    46. For an unanchored pattern, if a match attempt fails at the start of a
1530        newline sequence, and the newline setting is CRLF or ANY, and the next two
1531        characters are CRLF, advance by two characters instead of one.
1532    
1533    
1534    Version 6.7 04-Jul-06
1535    ---------------------
1536    
1537     1. In order to handle tests when input lines are enormously long, pcretest has
1538        been re-factored so that it automatically extends its buffers when
1539        necessary. The code is crude, but this _is_ just a test program. The
1540        default size has been increased from 32K to 50K.
1541    
1542     2. The code in pcre_study() was using the value of the re argument before
1543        testing it for NULL. (Of course, in any sensible call of the function, it
1544        won't be NULL.)
1545    
1546     3. The memmove() emulation function in pcre_internal.h, which is used on
1547        systems that lack both memmove() and bcopy() - that is, hardly ever -
1548        was missing a "static" storage class specifier.
1549    
1550     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1551        containing an extended class (one that cannot be represented by a bitmap
1552        because it contains high-valued characters or Unicode property items, e.g.
1553        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1554        pattern, but PCRE should not loop if you do not (it no longer does).
1555        [Detail: two cases were found: (a) a repeated subpattern containing an
1556        extended class; (b) a recursive reference to a subpattern that followed a
1557        previous extended class. It wasn't skipping over the extended class
1558        correctly when UTF-8 mode was not set.]
1559    
1560     5. A negated single-character class was not being recognized as fixed-length
1561        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1562        compile error "lookbehind assertion is not fixed length".
1563    
1564     6. The RunPerlTest auxiliary script was showing an unexpected difference
1565        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1566        write a Perl script that can interpret lines of an input file either as
1567        byte characters or as UTF-8, which is what "perltest" was being required to
1568        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1569        can't do is switch easily at run time between having the "use utf8;" pragma
1570        or not. In the end, I fudged it by using the RunPerlTest script to insert
1571        "use utf8;" explicitly for the UTF-8 tests.
1572    
1573     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1574        the end of the subject string, contrary to the documentation and to what
1575        Perl does. This was true of both matching functions. Now it matches only at
1576        the start of the subject and immediately after *internal* newlines.
1577    
1578     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1579        a pointer to an int instead of a pointer to an unsigned long int. This
1580        caused problems on 64-bit systems.
1581    
1582     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1583        instance of the 'standard' template library not being so standard".
1584    
1585    10. There was no check on the number of named subpatterns nor the maximum
1586        length of a subpattern name. The product of these values is used to compute
1587        the size of the memory block for a compiled pattern. By supplying a very
1588        long subpattern name and a large number of named subpatterns, the size
1589        computation could be caused to overflow. This is now prevented by limiting
1590        the length of names to 32 characters, and the number of named subpatterns
1591        to 10,000.
1592    
1593    11. Subpatterns that are repeated with specific counts have to be replicated in
1594        the compiled pattern. The size of memory for this was computed from the
1595        length of the subpattern and the repeat count. The latter is limited to
1596        65535, but there was no limit on the former, meaning that integer overflow
1597        could in principle occur. The compiled length of a repeated subpattern is
1598        now limited to 30,000 bytes in order to prevent this.
1599    
1600    12. Added the optional facility to have named substrings with the same name.
1601    
1602    13. Added the ability to use a named substring as a condition, using the
1603        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1604        are numbers (not recommended). Forward references are permitted.
1605    
1606    14. Added forward references in named backreferences (if you see what I mean).
1607    
1608    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1609        pattern could run off the end of the subject. For example, the pattern
1610        "(?s)(.{1,5})"8 did this with the subject "ab".
1611    
1612    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1613        PCRE_CASELESS was set when matching characters that were quantified with ?
1614        or *.
1615    
1616    17. A character class other than a single negated character that had a minimum
1617        but no maximum quantifier - for example [ab]{6,} - was not handled
1618        correctly by pce_dfa_exec(). It would match only one character.
1619    
1620    18. A valid (though odd) pattern that looked like a POSIX character
1621        class but used an invalid character after [ (for example [[,abc,]]) caused
1622        pcre_compile() to give the error "Failed: internal error: code overflow" or
1623        in some cases to crash with a glibc free() error. This could even happen if
1624        the pattern terminated after [[ but there just happened to be a sequence of
1625        letters, a binary zero, and a closing ] in the memory that followed.
1626    
1627    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1628        over the years. Originally (before any Unicode support), just the bottom 8
1629        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1630        output from "man perlunicode" includes this:
1631    
1632          The regular expression compiler produces polymorphic opcodes.  That
1633          is, the pattern adapts to the data and automatically switches to
1634          the Unicode character scheme when presented with Unicode data--or
1635          instead uses a traditional byte scheme when presented with byte
1636          data.
1637    
1638        Sadly, a wide octal escape does not cause a switch, and in a string with
1639        no other multibyte characters, these octal escapes are treated as before.
1640        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1641        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1642        Unicode string.
1643    
1644        I have not perpetrated such confusion in PCRE. Up till now, it took just
1645        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1646        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1647        translate to the appropriate multibyte character.
1648    
1649    29. Applied some refactoring to reduce the number of warnings from Microsoft
1650        and Borland compilers. This has included removing the fudge introduced
1651        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1652        a warning about an unused variable.
1653    
1654    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1655        characters since release 4.0, because Perl (from release 5.004) does not.
1656        [Or at least, is documented not to: some releases seem to be in conflict
1657        with the documentation.] However, when a pattern was studied with
1658        pcre_study() and all its branches started with \s, PCRE still included VT
1659        as a possible starting character. Of course, this did no harm; it just
1660        caused an unnecessary match attempt.
1661    
1662    22. Removed a now-redundant internal flag bit that recorded the fact that case
1663        dependency changed within the pattern. This was once needed for "required
1664        byte" processing, but is no longer used. This recovers a now-scarce options
1665        bit. Also moved the least significant internal flag bit to the most-
1666        significant bit of the word, which was not previously used (hangover from
1667        the days when it was an int rather than a uint) to free up another bit for
1668        the future.
1669    
1670    23. Added support for CRLF line endings as well as CR and LF. As well as the
1671        default being selectable at build time, it can now be changed at runtime
1672        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1673        specify that it is scanning data with non-default line endings.
1674    
1675    24. Changed the definition of CXXLINK to make it agree with the definition of
1676        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1677    
1678    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1679        recursions. This makes a big different to stack usage for some patterns.
1680    
1681    26. If a subpattern containing a named recursion or subroutine reference such
1682        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1683        the space required for the compiled pattern went wrong and gave too small a
1684        value. Depending on the environment, this could lead to "Failed: internal
1685        error: code overflow at offset 49" or "glibc detected double free or
1686        corruption" errors.
1687    
1688    27. Applied patches from Google (a) to support the new newline modes and (b) to
1689        advance over multibyte UTF-8 characters in GlobalReplace.
1690    
1691    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1692        difference for some implementation of PCRE in some Windows version.
1693    
1694    29. Added some extra testing facilities to pcretest:
1695    
1696        \q<number>   in a data line sets the "match limit" value
1697        \Q<number>   in a data line sets the "match recursion limt" value
1698        -S <number>  sets the stack size, where <number> is in megabytes
1699    
1700        The -S option isn't available for Windows.
1701    
1702    
1703    Version 6.6 06-Feb-06
1704    ---------------------
1705    
1706     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1707        in pcreposix.h. I have copied the definition from pcre.h.
1708    
1709     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1710        because pcre.h is no longer a built file.
1711    
1712     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1713        not normally included in the compiled code.
1714    
1715    
1716    Version 6.5 01-Feb-06
1717    ---------------------
1718    
1719     1. When using the partial match feature with pcre_dfa_exec(), it was not
1720        anchoring the second and subsequent partial matches at the new starting
1721        point. This could lead to incorrect results. For example, with the pattern
1722        /1234/, partially matching against "123" and then "a4" gave a match.
1723    
1724     2. Changes to pcregrep:
1725    
1726        (a) All non-match returns from pcre_exec() were being treated as failures
1727            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1728            error message is output. Some extra information is given for the
1729            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1730            probably the only errors that are likely to be caused by users (by
1731            specifying a regex that has nested indefinite repeats, for instance).
1732            If there are more than 20 of these errors, pcregrep is abandoned.
1733    
1734        (b) A binary zero was treated as data while matching, but terminated the
1735            output line if it was written out. This has been fixed: binary zeroes
1736            are now no different to any other data bytes.
1737    
1738        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1739            used to set a locale for matching. The --locale=xxxx long option has
1740            been added (no short equivalent) to specify a locale explicitly on the
1741            pcregrep command, overriding the environment variables.
1742    
1743        (d) When -B was used with -n, some line numbers in the output were one less
1744            than they should have been.
1745    
1746        (e) Added the -o (--only-matching) option.
1747    
1748        (f) If -A or -C was used with -c (count only), some lines of context were
1749            accidentally printed for the final match.
1750    
1751        (g) Added the -H (--with-filename) option.
1752    
1753        (h) The combination of options -rh failed to suppress file names for files
1754            that were found from directory arguments.
1755    
1756        (i) Added the -D (--devices) and -d (--directories) options.
1757    
1758        (j) Added the -F (--fixed-strings) option.
1759    
1760        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1761    
1762        (l) Added the --colo(u)r option.
1763    
1764        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1765            is not present by default.
1766    
1767     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1768        items such as (?R) or (?1), when the recursion could match a number of
1769        alternatives. If it matched one of the alternatives, but subsequently,
1770        outside the recursion, there was a failure, the code tried to back up into
1771        the recursion. However, because of the way PCRE is implemented, this is not
1772        possible, and the result was an incorrect result from the match.
1773    
1774        In order to prevent this happening, the specification of recursion has
1775        been changed so that all such subpatterns are automatically treated as
1776        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1777    
1778     4. I had overlooked the fact that, in some locales, there are characters for
1779        which isalpha() is true but neither isupper() nor islower() are true. In
1780        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1781        and ordfeminine) are like this. This affected the treatment of \w and \W
1782        when they appeared in character classes, but not when they appeared outside
1783        a character class. The bit map for "word" characters is now created
1784        separately from the results of isalnum() instead of just taking it from the
1785        upper, lower, and digit maps. (Plus the underscore character, of course.)
1786    
1787     5. The above bug also affected the handling of POSIX character classes such as
1788        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1789        permanent tables. Instead, the bit maps for such a class were previously
1790        created as the appropriate unions of the upper, lower, and digit bitmaps.
1791        Now they are created by subtraction from the [[:word:]] class, which has
1792        its own bitmap.
1793    
1794     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1795        It is created by subtracting the vertical space characters (\x09, \x0a,
1796        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1797        subtraction was done in the overall bitmap for a character class, meaning
1798        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1799        be recognized. This bug has been fixed.
1800    
1801     7. Patches from the folks at Google:
1802    
1803          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1804          real life, but is still worth protecting against".
1805    
1806          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1807          regular expressions".
1808    
1809          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1810          have it.
1811    
1812          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1813          "configure" and the latter not, in order to fix a problem somebody had
1814          with compiling the Arg class on HP-UX.
1815    
1816          (e) Improve the error-handling of the C++ wrapper a little bit.
1817    
1818          (f) New tests for checking recursion limiting.
1819    
1820     8. The pcre_memmove() function, which is used only if the environment does not
1821        have a standard memmove() function (and is therefore rarely compiled),
1822        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1823        returning a result (though PCRE never actually uses the result).
1824    
1825     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1826        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1827        returned instead of calling malloc() with an overflowing number that would
1828        most likely cause subsequent chaos.
1829    
1830    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1831    
1832    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1833        with this option is matched, the nmatch and pmatch options of regexec() are
1834        ignored.
1835    
1836    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1837        provided in case anyone wants to the the POSIX interface with UTF-8
1838        strings.
1839    
1840    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1841        C++ linking (needed for some HP-UX environments).
1842    
1843    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1844        (unused parameter) and in the pcre_printint() function (omitted "default"
1845        switch label when the default is to do nothing).
1846    
1847    15. Added some code to make it possible, when PCRE is compiled as a C++
1848        library, to replace subject pointers for pcre_exec() with a smart pointer
1849        class, thus making it possible to process discontinuous strings.
1850    
1851    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1852        much the same function. They were added by different people who were trying
1853        to make PCRE easy to compile on non-Unix systems. It has been suggested
1854        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1855        for compiling on Windows systems. I have therefore replaced it with
1856        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1857        defaults to "extern" for C or "extern C" for C++, which works fine on
1858        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1859        SCOPE with something explicit in config.h. In addition:
1860    
1861        (a) pcreposix.h still had just "extern" instead of either of these macros;
1862            I have replaced it with PCRE_DATA_SCOPE.
1863    
1864        (b) Functions such as _pcre_xclass(), which are internal to the library,
1865            but external in the C sense, all had PCRE_EXPORT in their definitions.
1866            This is apparently wrong for the Windows case, so I have removed it.
1867            (It makes no difference on Unix-like systems.)
1868    
1869    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1870        of recursive calls to match(). This is different to MATCH_LIMIT because
1871        that limits the total number of calls to match(), not all of which increase
1872        the depth of recursion. Limiting the recursion depth limits the amount of
1873        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1874        when PCRE is compiled, and changed at run time. A patch from Google adds
1875        this functionality to the C++ interface.
1876    
1877    18. Changes to the handling of Unicode character properties:
1878    
1879        (a) Updated the table to Unicode 4.1.0.
1880    
1881        (b) Recognize characters that are not in the table as "Cn" (undefined).
1882    
1883        (c) I revised the way the table is implemented to a much improved format
1884            which includes recognition of ranges. It now supports the ranges that
1885            are defined in UnicodeData.txt, and it also amalgamates other
1886            characters into ranges. This has reduced the number of entries in the
1887            table from around 16,000 to around 3,000, thus reducing its size
1888            considerably. I realized I did not need to use a tree structure after
1889            all - a binary chop search is just as efficient. Having reduced the
1890            number of entries, I extended their size from 6 bytes to 8 bytes to
1891            allow for more data.
1892    
1893        (d) Added support for Unicode script names via properties such as \p{Han}.
1894    
1895    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1896        matching that character.
1897    
1898    20. When matching a repeated Unicode property with a minimum greater than zero,
1899        (for example \pL{2,}), PCRE could look past the end of the subject if it
1900        reached it while seeking the minimum number of characters. This could
1901        happen only if some of the characters were more than one byte long, because
1902        there is a check for at least the minimum number of bytes.
1903    
1904    21. Refactored the implementation of \p and \P so as to be more general, to
1905        allow for more different types of property in future. This has changed the
1906        compiled form incompatibly. Anybody with saved compiled patterns that use
1907        \p or \P will have to recompile them.
1908    
1909    22. Added "Any" and "L&" to the supported property types.
1910    
1911    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1912        but give a compile time error if the value is greater than 0xff.
1913    
1914    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1915        accidentally not being installed or uninstalled.
1916    
1917    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1918        made were to insert the current release number. This seemed silly, because
1919        it made things harder for people building PCRE on systems that don't run
1920        "configure". I have turned pcre.h into a distributed file, no longer built
1921        by "configure", with the version identification directly included. There is
1922        no longer a pcre.h.in file.
1923    
1924        However, this change necessitated a change to the pcre-config script as
1925        well. It is built from pcre-config.in, and one of the substitutions was the
1926        release number. I have updated configure.ac so that ./configure now finds
1927        the release number by grepping pcre.h.
1928    
1929    26. Added the ability to run the tests under valgrind.
1930    
1931    
1932    Version 6.4 05-Sep-05
1933    ---------------------
1934    
1935     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1936        "--" to be printed when multiple files were scanned, even when none of the
1937        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1938        consider it to be a bug, and have restored the previous behaviour.
1939    
1940     2. A couple of code tidies to get rid of compiler warnings.
1941    
1942     3. The pcretest program used to cheat by referring to symbols in the library
1943        whose names begin with _pcre_. These are internal symbols that are not
1944        really supposed to be visible externally, and in some environments it is
1945        possible to suppress them. The cheating is now confined to including
1946        certain files from the library's source, which is a bit cleaner.
1947    
1948     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1949        file's purpose clearer.
1950    
1951     5. Reorganized pcre_ucp_findchar().
1952    
1953    
1954  Version 6.3 15-Aug-05  Version 6.3 15-Aug-05
1955  ---------------------  ---------------------
1956    

Legend:
Removed from v.83  
changed lines
  Added in v.492

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12