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

Diff of /code/trunk/ChangeLog

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

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