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

Diff of /code/trunk/ChangeLog

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

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8        was out-of-date, and there was no check on the pcre_compile() error code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

Legend:
Removed from v.55  
changed lines
  Added in v.448

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12