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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.87  
changed lines
  Added in v.503

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12