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

Diff of /code/trunk/ChangeLog

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

revision 43 by nigel, Sat Feb 24 21:39:21 2007 UTC revision 430 by ph10, Tue Sep 1 16:29:46 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8        was out-of-date, and there was no check on the pcre_compile() error code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. Chaos could result.
47    
48    9.  The restrictions on what a pattern can contain when partial matching is
49        requested for pcre_exec() have been removed. All patterns can now be
50        partially matched by this function. In addition, if there are at least two
51        slots in the offset vector, the offsets of the first-encountered partial
52        match are set in them when PCRE_ERROR_PARTIAL is returned.
53    
54    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
55        synonymous with PCRE_PARTIAL, for backwards compatibility, and
56        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
57        and may be more useful for multi-segment matching, especially with
58        pcre_exec().
59    
60    11. Partial matching with pcre_exec() is now more intuitive. A partial match
61        used to be given if ever the end of the subject was reached; now it is
62        given only if matching could not proceed because another character was
63        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
64        string "Z", which now yields "no match" instead of "partial match". In the
65        case of pcre_dfa_exec(), "no match" is given if every matching path for the
66        final character ended with (*FAIL).
67    
68    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
69        if the pattern had a "must contain" character that was already found in the
70        earlier partial match, unless partial matching was again requested. For
71        example, with the pattern /dog.(body)?/, the "must contain" character is
72        "g". If the first part-match was for the string "dog", restarting with
73        "sbody" failed.
74    
75    13. Added a pcredemo man page, created automatically from the pcredemo.c file,
76        so that the demonstration program is easily available in environments where
77        PCRE has not been installed from source.
78    
79    14. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
80        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
81        library.
82    
83    
84    Version 7.9 11-Apr-09
85    ---------------------
86    
87    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
88        (pcretest), all targets were linked against these libraries. This included
89        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
90        libraries. This caused unwanted dependencies to be created. This problem
91        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
92        pcretest is linked with readline.
93    
94    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
95        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
96        moved outside it again, because FALSE and TRUE are already defined in AIX,
97        but BOOL is not.
98    
99    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
100        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
101    
102    4.  The pcregrep documentation said spaces were inserted as well as colons (or
103        hyphens) following file names and line numbers when outputting matching
104        lines. This is not true; no spaces are inserted. I have also clarified the
105        wording for the --colour (or --color) option.
106    
107    5.  In pcregrep, when --colour was used with -o, the list of matching strings
108        was not coloured; this is different to GNU grep, so I have changed it to be
109        the same.
110    
111    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
112        each matching line was coloured. Now it goes on to look for further matches
113        of any of the test patterns, which is the same behaviour as GNU grep.
114    
115    7.  A pattern that could match an empty string could cause pcregrep to loop; it
116        doesn't make sense to accept an empty string match in pcregrep, so I have
117        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
118        seems to be how GNU grep behaves.
119    
120    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
121        start or after a newline", because the conditional assertion was not being
122        correctly handled. The rule now is that both the assertion and what follows
123        in the first alternative must satisfy the test.
124    
125    9.  If auto-callout was enabled in a pattern with a conditional group whose
126        condition was an assertion, PCRE could crash during matching, both with
127        pcre_exec() and pcre_dfa_exec().
128    
129    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
130        used for matching.
131    
132    11. Unicode property support in character classes was not working for
133        characters (bytes) greater than 127 when not in UTF-8 mode.
134    
135    12. Added the -M command line option to pcretest.
136    
137    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
138    
139    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
140    
141    16. Added comments and documentation about mis-use of no_arg in the C++
142        wrapper.
143    
144    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
145        from Martin Jerabek that uses macro names for all relevant character and
146        string constants.
147    
148    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
149        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
150        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
151        these, but not everybody uses configure.
152    
153    19. A conditional group that had only one branch was not being correctly
154        recognized as an item that could match an empty string. This meant that an
155        enclosing group might also not be so recognized, causing infinite looping
156        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
157        with the subject "ab", where knowledge that the repeated group can match
158        nothing is needed in order to break the loop.
159    
160    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
161        exec(), but without supplying a callout function, matching went wrong.
162    
163    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
164        leak if the size of the offset vector was greater than 30. When the vector
165        is smaller, the saved offsets during recursion go onto a local stack
166        vector, but for larger vectors malloc() is used. It was failing to free
167        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
168        error, in fact).
169    
170    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
171        heapframe that is used only when UTF-8 support is enabled. This caused no
172        problem, but was untidy.
173    
174    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
175        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
176        included within another project.
177    
178    24. Steven Van Ingelgem's patches to add more options to the CMake support,
179        slightly modified by me:
180    
181          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
182              not building pcregrep.
183    
184          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
185              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
186    
187    25. Forward references, both numeric and by name, in patterns that made use of
188        duplicate group numbers, could behave incorrectly or give incorrect errors,
189        because when scanning forward to find the reference group, PCRE was not
190        taking into account the duplicate group numbers. A pattern such as
191        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
192    
193    26. Changed a few more instances of "const unsigned char *" to USPTR, making
194        the feature of a custom pointer more persuasive (as requested by a user).
195    
196    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
197        pcretest.c, inside #ifndefs, because it seems they are sometimes already
198        pre-defined.
199    
200    28. Added support for (*UTF8) at the start of a pattern.
201    
202    29. Arrange for flags added by the "release type" setting in CMake to be shown
203        in the configuration summary.
204    
205    
206    Version 7.8 05-Sep-08
207    ---------------------
208    
209    1.  Replaced UCP searching code with optimized version as implemented for Ad
210        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
211        stage table and inline lookup instead of a function, giving speed ups of 2
212        to 5 times on some simple patterns that I tested. Permission was given to
213        distribute the MultiStage2.py script that generates the tables (it's not in
214        the tarball, but is in the Subversion repository).
215    
216    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
217        scripts.
218    
219    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
220        a group with a zero qualifier. The result of the study could be incorrect,
221        or the function might crash, depending on the pattern.
222    
223    4.  Caseless matching was not working for non-ASCII characters in back
224        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
225        It now works when Unicode Property Support is available.
226    
227    5.  In pcretest, an escape such as \x{de} in the data was always generating
228        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
229        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
230        truncation.
231    
232    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
233    
234    7.  Added two (int) casts to pcregrep when printing the difference of two
235        pointers, in case they are 64-bit values.
236    
237    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
238        test 2 if it fails.
239    
240    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
241        and a #define of that name to empty if it is not externally set. This is to
242        allow users of MSVC to set it if necessary.
243    
244    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
245        the convenience functions in the pcre_get.c source file.
246    
247    11. An option change at the start of a pattern that had top-level alternatives
248        could cause overwriting and/or a crash. This command provoked a crash in
249        some environments:
250    
251          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
252    
253        This potential security problem was recorded as CVE-2008-2371.
254    
255    12. For a pattern where the match had to start at the beginning or immediately
256        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
257        pcre_dfa_exec() could read past the end of the passed subject if there was
258        no match. To help with detecting such bugs (e.g. with valgrind), I modified
259        pcretest so that it places the subject at the end of its malloc-ed buffer.
260    
261    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
262        exec() might read past the end of the data buffer in UTF-8 mode.
263    
264    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
265        the data contained the byte 0x85 as part of a UTF-8 character within its
266        first line. This applied both to normal and DFA matching.
267    
268    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
269        /^[^d]*?$/8 failed to match "abc".
270    
271    16. Added a missing copyright notice to pcrecpp_internal.h.
272    
273    17. Make it more clear in the documentation that values returned from
274        pcre_exec() in ovector are byte offsets, not character counts.
275    
276    18. Tidied a few places to stop certain compilers from issuing warnings.
277    
278    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
279        supplied by Stefan Weber. I made a further small update for 7.8 because
280        there is a change of source arrangements: the pcre_searchfuncs.c module is
281        replaced by pcre_ucd.c.
282    
283    
284    Version 7.7 07-May-08
285    ---------------------
286    
287    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
288        a string to a long long, pretend we don't even have a long long." This is
289        done by checking for the strtoq, strtoll, and _strtoi64 functions.
290    
291    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
292        pre-7.6 versions, which defined a global no_arg variable instead of putting
293        it in the RE class. (See also #8 below.)
294    
295    3.  Remove a line of dead code, identified by coverity and reported by Nuno
296        Lopes.
297    
298    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
299    
300        (1) The include/exclude patterns were being applied to the whole pathnames
301            of files, instead of just to the final components.
302    
303        (2) If there was more than one level of directory, the subdirectories were
304            skipped unless they satisfied the include/exclude conditions. This is
305            inconsistent with GNU grep (and could even be seen as contrary to the
306            pcregrep specification - which I improved to make it absolutely clear).
307            The action now is always to scan all levels of directory, and just
308            apply the include/exclude patterns to regular files.
309    
310    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
311        --exclude_dir in the tests to avoid scanning .svn directories.
312    
313    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
314        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
315        doesn't support NULs in patterns.
316    
317    7.  Added some missing "const"s to declarations of static tables in
318        pcre_compile.c and pcre_dfa_exec.c.
319    
320    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
321        caused by fix #2  above. (Subsequently also a second patch to fix the
322        first patch. And a third patch - this was a messy problem.)
323    
324    9.  Applied Craig's patch to remove the use of push_back().
325    
326    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
327        matching function regexec().
328    
329    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
330        which, however, unlike Perl's \g{...}, are subroutine calls, not back
331        references. PCRE supports relative numbers with this syntax (I don't think
332        Oniguruma does).
333    
334    12. Previously, a group with a zero repeat such as (...){0} was completely
335        omitted from the compiled regex. However, this means that if the group
336        was called as a subroutine from elsewhere in the pattern, things went wrong
337        (an internal error was given). Such groups are now left in the compiled
338        pattern, with a new opcode that causes them to be skipped at execution
339        time.
340    
341    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
342        to the way PCRE behaves:
343    
344        (a) A lone ] character is dis-allowed (Perl treats it as data).
345    
346        (b) A back reference to an unmatched subpattern matches an empty string
347            (Perl fails the current match path).
348    
349        (c) A data ] in a character class must be notated as \] because if the
350            first data character in a class is ], it defines an empty class. (In
351            Perl it is not possible to have an empty class.) The empty class []
352            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
353            The negative empty class [^] matches any one character, independently
354            of the DOTALL setting.
355    
356    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
357        non-existent subpattern following a character class starting with ']' and
358        containing () gave an internal compiling error instead of "reference to
359        non-existent subpattern". Fortunately, when the pattern did exist, the
360        compiled code was correct. (When scanning forwards to check for the
361        existencd of the subpattern, it was treating the data ']' as terminating
362        the class, so got the count wrong. When actually compiling, the reference
363        was subsequently set up correctly.)
364    
365    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
366        it was being rejected as not supported by pcre_dfa_exec(), even though
367        other assertions are supported. I have made pcre_dfa_exec() support
368        (*FAIL).
369    
370    16. The implementation of 13c above involved the invention of a new opcode,
371        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
372        cannot be changed at match time, I realized I could make a small
373        improvement to matching performance by compiling OP_ALLANY instead of
374        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
375        on the OP_ANY path.
376    
377    17. Compiling pcretest on Windows with readline support failed without the
378        following two fixes: (1) Make the unistd.h include conditional on
379        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
380    
381    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
382        ncurses library to be included for pcretest when ReadLine support is
383        requested, but also to allow for it to be overridden. This patch came from
384        Daniel Bergström.
385    
386    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
387        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
388        any errors with the current Unicode tables. Thanks to Peter Kankowski for
389        spotting this.
390    
391    
392    Version 7.6 28-Jan-08
393    ---------------------
394    
395    1.  A character class containing a very large number of characters with
396        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
397        overflow.
398    
399    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
400        HAVE_LONG_LONG is not defined.
401    
402    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
403        bring them up to date and include new features. This patch includes:
404    
405        - Fixed PH's badly added libz and libbz2 support.
406        - Fixed a problem with static linking.
407        - Added pcredemo. [But later removed - see 7 below.]
408        - Fixed dftables problem and added an option.
409        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
410            HAVE_LONG_LONG.
411        - Added readline support for pcretest.
412        - Added an listing of the option settings after cmake has run.
413    
414    4.  A user submitted a patch to Makefile that makes it easy to create
415        "pcre.dll" under mingw when using Configure/Make. I added stuff to
416        Makefile.am that cause it to include this special target, without
417        affecting anything else. Note that the same mingw target plus all
418        the other distribution libraries and programs are now supported
419        when configuring with CMake (see 6 below) instead of with
420        Configure/Make.
421    
422    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
423        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
424        exported in the Windows port". It has not yet been confirmed that the patch
425        solves the problem, but it does no harm.
426    
427    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
428        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
429        with CMake, and also correct the comment about stack recursion.
430    
431    7.  Remove the automatic building of pcredemo from the ./configure system and
432        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
433        of a program that users should build themselves after PCRE is installed, so
434        building it automatically is not really right. What is more, it gave
435        trouble in some build environments.
436    
437    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
438    
439    
440    Version 7.5 10-Jan-08
441    ---------------------
442    
443    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
444        values in parens when parsing an RE using the C++ wrapper."
445    
446    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
447        Characters greater than 255 were excluded from the class instead of being
448        included.
449    
450    3.  The same bug as (2) above applied to negated POSIX classes such as
451        [:^space:].
452    
453    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
454        defined or documented. It seems to have been a typo for PCRE_STATIC, so
455        I have changed it.
456    
457    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
458        first named subpattern) and a construct such as (?&a) would reference the
459        first named subpattern whose name started with "a" (in other words, the
460        length check was missing). Both these problems are fixed. "Subpattern name
461        expected" is now given for (?&) (a zero-length name), and this patch also
462        makes it give the same error for \k'' (previously it complained that that
463        was a reference to a non-existent subpattern).
464    
465    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
466        this is right because (?- can be followed by option settings as well as by
467        digits. I have, however, made the messages clearer.
468    
469    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
470        than the number used in the conditional) now cause a compile-time error.
471        This is actually not compatible with Perl, which accepts such patterns, but
472        treats the conditional as always being FALSE (as PCRE used to), but it
473        seems to me that giving a diagnostic is better.
474    
475    8.  Change "alphameric" to the more common word "alphanumeric" in comments
476        and messages.
477    
478    9.  Fix two occurrences of "backslash" in comments that should have been
479        "backspace".
480    
481    10. Remove two redundant lines of code that can never be obeyed (their function
482        was moved elsewhere).
483    
484    11. The program that makes PCRE's Unicode character property table had a bug
485        which caused it to generate incorrect table entries for sequences of
486        characters that have the same character type, but are in different scripts.
487        It amalgamated them into a single range, with the script of the first of
488        them. In other words, some characters were in the wrong script. There were
489        thirteen such cases, affecting characters in the following ranges:
490    
491          U+002b0 - U+002c1
492          U+0060c - U+0060d
493          U+0061e - U+00612
494          U+0064b - U+0065e
495          U+0074d - U+0076d
496          U+01800 - U+01805
497          U+01d00 - U+01d77
498          U+01d9b - U+01dbf
499          U+0200b - U+0200f
500          U+030fc - U+030fe
501          U+03260 - U+0327f
502          U+0fb46 - U+0fbb1
503          U+10450 - U+1049d
504    
505    12. The -o option (show only the matching part of a line) for pcregrep was not
506        compatible with GNU grep in that, if there was more than one match in a
507        line, it showed only the first of them. It now behaves in the same way as
508        GNU grep.
509    
510    13. If the -o and -v options were combined for pcregrep, it printed a blank
511        line for every non-matching line. GNU grep prints nothing, and pcregrep now
512        does the same. The return code can be used to tell if there were any
513        non-matching lines.
514    
515    14. Added --file-offsets and --line-offsets to pcregrep.
516    
517    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
518        infinitely looping recursion. The bug was that positive lookaheads were not
519        being skipped when checking for a possible empty match (negative lookaheads
520        and both kinds of lookbehind were skipped).
521    
522    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
523        inclusion of <windows.h> to before rather than after the definition of
524        INVALID_FILE_ATTRIBUTES (patch from David Byron).
525    
526    17. Specifying a possessive quantifier with a specific limit for a Unicode
527        character property caused pcre_compile() to compile bad code, which led at
528        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
529        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
530        caused the error; without that there was no problem.
531    
532    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
533    
534    19. Added --enable-pcretest-libreadline.
535    
536    20. In pcrecpp.cc, the variable 'count' was incremented twice in
537        RE::GlobalReplace(). As a result, the number of replacements returned was
538        double what it should be. I removed one of the increments, but Craig sent a
539        later patch that removed the other one (the right fix) and added unit tests
540        that check the return values (which was not done before).
541    
542    21. Several CMake things:
543    
544        (1) Arranged that, when cmake is used on Unix, the libraries end up with
545            the names libpcre and libpcreposix, not just pcre and pcreposix.
546    
547        (2) The above change means that pcretest and pcregrep are now correctly
548            linked with the newly-built libraries, not previously installed ones.
549    
550        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
551    
552    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
553        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
554        UTF-8 newline character). The key issue is that the pattern starts .*;
555        this means that the match must be either at the beginning, or after a
556        newline. The bug was in the code for advancing after a failed match and
557        checking that the new position followed a newline. It was not taking
558        account of UTF-8 characters correctly.
559    
560    23. PCRE was behaving differently from Perl in the way it recognized POSIX
561        character classes. PCRE was not treating the sequence [:...:] as a
562        character class unless the ... were all letters. Perl, however, seems to
563        allow any characters between [: and :], though of course it rejects as
564        unknown any "names" that contain non-letters, because all the known class
565        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
566        for example, whereas PCRE did not - it did not recognize a POSIX character
567        class. This seemed a bit dangerous, so the code has been changed to be
568        closer to Perl. The behaviour is not identical to Perl, because PCRE will
569        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
570        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
571        Perl does, and where it didn't before.
572    
573    24. Rewrite so as to remove the single use of %n from pcregrep because in some
574        Windows environments %n is disabled by default.
575    
576    
577    Version 7.4 21-Sep-07
578    ---------------------
579    
580    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
581        means that a class such as [\s] counted as "explicit reference to CR or
582        LF". That isn't really right - the whole point of the change was to try to
583        help when there was an actual mention of one of the two characters. So now
584        the change happens only if \r or \n (or a literal CR or LF) character is
585        encountered.
586    
587    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
588        of both had grown to the point where there were only 3 bits left.
589        Fortunately, there was spare space in the data structure, and so I have
590        moved the internal flags into a new 16-bit field to free up more option
591        bits.
592    
593    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
594        but did not set the internal JCHANGED flag - either of these is enough to
595        control the way the "get" function works - but the PCRE_INFO_JCHANGED
596        facility is supposed to tell if (?J) was ever used, so now (?J) at the
597        start sets both bits.
598    
599    4.  Added options (at build time, compile time, exec time) to change \R from
600        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
601    
602    5.  doc/pcresyntax.html was missing from the distribution.
603    
604    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
605        compatibility, even though it is no longer used.
606    
607    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
608        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
609        windows.h file is present (where different names are used). [This was
610        reversed later after testing - see 16 below.]
611    
612    8.  Changed all #include <config.h> to #include "config.h". There were also
613        some further <pcre.h> cases that I changed to "pcre.h".
614    
615    9.  When pcregrep was used with the --colour option, it missed the line ending
616        sequence off the lines that it output.
617    
618    10. It was pointed out to me that arrays of string pointers cause lots of
619        relocations when a shared library is dynamically loaded. A technique of
620        using a single long string with a table of offsets can drastically reduce
621        these. I have refactored PCRE in four places to do this. The result is
622        dramatic:
623    
624          Originally:                          290
625          After changing UCP table:            187
626          After changing error message table:   43
627          After changing table of "verbs"       36
628          After changing table of Posix names   22
629    
630        Thanks to the folks working on Gregex for glib for this insight.
631    
632    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
633        unicode-properties was also set.
634    
635    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
636    
637    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
638        checked only for CRLF.
639    
640    14. Added casts to pcretest.c to avoid compiler warnings.
641    
642    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
643    
644    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
645        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
646        entirely. This removes changes made in 7 above.
647    
648    17. The CMake files have been updated, and there is now more information about
649        building with CMake in the NON-UNIX-USE document.
650    
651    
652    Version 7.3 28-Aug-07
653    ---------------------
654    
655     1. In the rejigging of the build system that eventually resulted in 7.1, the
656        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
657        brackets there is not right, since it causes compilers to look for an
658        installed pcre.h, not the version that is in the source that is being
659        compiled (which of course may be different). I have changed it back to:
660    
661          #include "pcre.h"
662    
663        I have a vague recollection that the change was concerned with compiling in
664        different directories, but in the new build system, that is taken care of
665        by the VPATH setting the Makefile.
666    
667     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
668        when the subject happened to end in the byte 0x85 (e.g. if the last
669        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
670        characters but of course it shouldn't be taken as a newline when it is part
671        of another character. The bug was that, for an unlimited repeat of . in
672        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
673        characters when looking for a newline.
674    
675     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
676    
677     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
678        in debug output.
679    
680     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
681        long printing in the pcrecpp unittest when running under MinGW.
682    
683     6. ESC_K was left out of the EBCDIC table.
684    
685     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
686        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
687        limit also applies to "virtual nesting" when a pattern is recursive, and in
688        this case 1000 isn't so big. I have been able to remove this limit at the
689        expense of backing off one optimization in certain circumstances. Normally,
690        when pcre_exec() would call its internal match() function recursively and
691        immediately return the result unconditionally, it uses a "tail recursion"
692        feature to save stack. However, when a subpattern that can match an empty
693        string has an unlimited repetition quantifier, it no longer makes this
694        optimization. That gives it a stack frame in which to save the data for
695        checking that an empty string has been matched. Previously this was taken
696        from the 1000-entry workspace that had been reserved. So now there is no
697        explicit limit, but more stack is used.
698    
699     8. Applied Daniel's patches to solve problems with the import/export magic
700        syntax that is required for Windows, and which was going wrong for the
701        pcreposix and pcrecpp parts of the library. These were overlooked when this
702        problem was solved for the main library.
703    
704     9. There were some crude static tests to avoid integer overflow when computing
705        the size of patterns that contain repeated groups with explicit upper
706        limits. As the maximum quantifier is 65535, the maximum group length was
707        set at 30,000 so that the product of these two numbers did not overflow a
708        32-bit integer. However, it turns out that people want to use groups that
709        are longer than 30,000 bytes (though not repeat them that many times).
710        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
711        made it possible to implement the integer overflow checks in a much more
712        dynamic way, which I have now done. The artificial limitation on group
713        length has been removed - we now have only the limit on the total length of
714        the compiled pattern, which depends on the LINK_SIZE setting.
715    
716    10. Fixed a bug in the documentation for get/copy named substring when
717        duplicate names are permitted. If none of the named substrings are set, the
718        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
719        empty string.
720    
721    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
722        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
723        because the ] is interpreted as the first data character and the
724        terminating ] is not found. PCRE has been made compatible with Perl in this
725        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
726        cause memory overwriting.
727    
728    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
729        string has been matched (to stop an infinite loop). It was not recognizing
730        a conditional subpattern that could match an empty string if that
731        subpattern was within another subpattern. For example, it looped when
732        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
733        condition was not nested. This bug has been fixed.
734    
735    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
736        past the start of the subject in the presence of bytes with the top bit
737        set, for example "\x8aBCD".
738    
739    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
740        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
741    
742    14. Optimized (?!) to (*FAIL).
743    
744    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
745        This restricts code points to be within the range 0 to 0x10FFFF, excluding
746        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
747        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
748        does: it's just the validity check that is more restrictive.
749    
750    16. Inserted checks for integer overflows during escape sequence (backslash)
751        processing, and also fixed erroneous offset values for syntax errors during
752        backslash processing.
753    
754    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
755        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
756    
757    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
758        caused an overrun.
759    
760    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
761        something other than just ASCII characters) inside a group that had an
762        unlimited repeat caused a loop at compile time (while checking to see
763        whether the group could match an empty string).
764    
765    20. Debugging a pattern containing \p or \P could cause a crash. For example,
766        [\P{Any}] did so. (Error in the code for printing property names.)
767    
768    21. An orphan \E inside a character class could cause a crash.
769    
770    22. A repeated capturing bracket such as (A)? could cause a wild memory
771        reference during compilation.
772    
773    23. There are several functions in pcre_compile() that scan along a compiled
774        expression for various reasons (e.g. to see if it's fixed length for look
775        behind). There were bugs in these functions when a repeated \p or \P was
776        present in the pattern. These operators have additional parameters compared
777        with \d, etc, and these were not being taken into account when moving along
778        the compiled data. Specifically:
779    
780        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
781            length.
782    
783        (b) An item such as \pL+ within a repeated group could cause crashes or
784            loops.
785    
786        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
787            "reference to non-existent subpattern" error.
788    
789        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
790    
791    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
792        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
793    
794    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
795    
796    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
797        character were causing crashes (broken optimization).
798    
799    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
800        \p or \P) caused a compile-time loop.
801    
802    28. More problems have arisen in unanchored patterns when CRLF is a valid line
803        break. For example, the unstudied pattern [\r\n]A does not match the string
804        "\r\nA" because change 7.0/46 below moves the current point on by two
805        characters after failing to match at the start. However, the pattern \nA
806        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
807        the same is true. There doesn't seem any very clean way out of this, but
808        what I have chosen to do makes the common cases work: PCRE now takes note
809        of whether there can be an explicit match for \r or \n anywhere in the
810        pattern, and if so, 7.0/46 no longer applies. As part of this change,
811        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
812        pattern has explicit CR or LF references.
813    
814    29. Added (*CR) etc for changing newline setting at start of pattern.
815    
816    
817    Version 7.2 19-Jun-07
818    ---------------------
819    
820     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
821        which is apparently normally available under Windows.
822    
823     2. Re-jig the pcregrep tests with different newline settings in an attempt
824        to make them independent of the local environment's newline setting.
825    
826     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
827    
828     4. Some of the "internals" tests were previously cut out when the link size
829        was not 2, because the output contained actual offsets. The recent new
830        "Z" feature of pcretest means that these can be cut out, making the tests
831        usable with all link sizes.
832    
833     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
834        stack recursion. This gives a massive performance boost under BSD, but just
835        a small improvement under Linux. However, it saves one field in the frame
836        in all cases.
837    
838     6. Added more features from the forthcoming Perl 5.10:
839    
840        (a) (?-n) (where n is a string of digits) is a relative subroutine or
841            recursion call. It refers to the nth most recently opened parentheses.
842    
843        (b) (?+n) is also a relative subroutine call; it refers to the nth next
844            to be opened parentheses.
845    
846        (c) Conditions that refer to capturing parentheses can be specified
847            relatively, for example, (?(-2)... or (?(+3)...
848    
849        (d) \K resets the start of the current match so that everything before
850            is not part of it.
851    
852        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
853    
854        (f) \g{name} is another synonym - part of Perl 5.10's unification of
855            reference syntax.
856    
857        (g) (?| introduces a group in which the numbering of parentheses in each
858            alternative starts with the same number.
859    
860        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
861    
862     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
863        PCRE_INFO_JCHANGED.
864    
865     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
866        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
867        for detecting groups that can match an empty string.
868    
869     9. A pattern with a very large number of alternatives (more than several
870        hundred) was running out of internal workspace during the pre-compile
871        phase, where pcre_compile() figures out how much memory will be needed. A
872        bit of new cunning has reduced the workspace needed for groups with
873        alternatives. The 1000-alternative test pattern now uses 12 bytes of
874        workspace instead of running out of the 4096 that are available.
875    
876    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
877    
878    11. Applied patch from Google to remove an optimization that didn't quite work.
879        The report of the bug said:
880    
881          pcrecpp::RE("a*").FullMatch("aaa") matches, while
882          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
883          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
884    
885    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
886        it matched the wrong number of bytes.
887    
888    
889    Version 7.1 24-Apr-07
890    ---------------------
891    
892     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
893        that is more "standard", making use of automake and other Autotools. There
894        is some re-arrangement of the files and adjustment of comments consequent
895        on this.
896    
897     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
898        for recursive directory scanning broke on some systems because the files
899        are not scanned in any specific order and on different systems the order
900        was different. A call to "sort" has been inserted into RunGrepTest for the
901        approprate test as a short-term fix. In the longer term there may be an
902        alternative.
903    
904     3. I had an email from Eric Raymond about problems translating some of PCRE's
905        man pages to HTML (despite the fact that I distribute HTML pages, some
906        people do their own conversions for various reasons). The problems
907        concerned the use of low-level troff macros .br and .in. I have therefore
908        removed all such uses from the man pages (some were redundant, some could
909        be replaced by .nf/.fi pairs). The 132html script that I use to generate
910        HTML has been updated to handle .nf/.fi and to complain if it encounters
911        .br or .in.
912    
913     4. Updated comments in configure.ac that get placed in config.h.in and also
914        arranged for config.h to be included in the distribution, with the name
915        config.h.generic, for the benefit of those who have to compile without
916        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
917    
918     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
919        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
920        makevp.bat for the current PCRE, using the additional files
921        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
922    
923     6. A Windows user reported a minor discrepancy with test 2, which turned out
924        to be caused by a trailing space on an input line that had got lost in his
925        copy. The trailing space was an accident, so I've just removed it.
926    
927     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
928        that is needed.
929    
930     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
931        as "const" (a) because they are and (b) because it helps the PHP
932        maintainers who have recently made a script to detect big data structures
933        in the php code that should be moved to the .rodata section. I remembered
934        to update Builducptable as well, so it won't revert if ucptable.h is ever
935        re-created.
936    
937     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
938        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
939        order to be able to cut out the UTF-8 tables in the latter when UTF-8
940        support is not required. This saves 1.5-2K of code, which is important in
941        some applications.
942    
943        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
944        so as not to refer to the tables, even though these functions will never be
945        called when UTF-8 support is disabled. Otherwise there are problems with a
946        shared library.
947    
948    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
949    
950        (a) It was defining its arguments as char * instead of void *.
951    
952        (b) It was assuming that all moves were upwards in memory; this was true
953            a long time ago when I wrote it, but is no longer the case.
954    
955        The emulated memove() is provided for those environments that have neither
956        memmove() nor bcopy(). I didn't think anyone used it these days, but that
957        is clearly not the case, as these two bugs were recently reported.
958    
959    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
960        and Detrail to create the HTML documentation, the .txt form of the man
961        pages, and it removes trailing spaces from listed files. It also creates
962        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
963        case, it wraps all the #defines with #ifndefs. This script should be run
964        before "make dist".
965    
966    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
967        with Unicode property support.
968    
969        (a) For a maximizing quantifier, if the two different cases of the
970            character were of different lengths in their UTF-8 codings (there are
971            some cases like this - I found 11), and the matching function had to
972            back up over a mixture of the two cases, it incorrectly assumed they
973            were both the same length.
974    
975        (b) When PCRE was configured to use the heap rather than the stack for
976            recursion during matching, it was not correctly preserving the data for
977            the other case of a UTF-8 character when checking ahead for a match
978            while processing a minimizing repeat. If the check also involved
979            matching a wide character, but failed, corruption could cause an
980            erroneous result when trying to check for a repeat of the original
981            character.
982    
983    13. Some tidying changes to the testing mechanism:
984    
985        (a) The RunTest script now detects the internal link size and whether there
986            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
987            values substituted by "configure". (The RunGrepTest script already did
988            this for UTF-8.) The configure.ac script no longer substitutes the
989            relevant variables.
990    
991        (b) The debugging options /B and /D in pcretest show the compiled bytecode
992            with length and offset values. This means that the output is different
993            for different internal link sizes. Test 2 is skipped for link sizes
994            other than 2 because of this, bypassing the problem. Unfortunately,
995            there was also a test in test 3 (the locale tests) that used /B and
996            failed for link sizes other than 2. Rather than cut the whole test out,
997            I have added a new /Z option to pcretest that replaces the length and
998            offset values with spaces. This is now used to make test 3 independent
999            of link size. (Test 2 will be tidied up later.)
1000    
1001    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1002        segmentation fault instead of returning the appropriate error message.
1003    
1004    15. In multiline mode when the newline sequence was set to "any", the pattern
1005        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1006        This doesn't seem right; it now treats the CRLF combination as the line
1007        ending, and so does not match in that case. It's only a pattern such as ^$
1008        that would hit this one: something like ^ABC$ would have failed after \r
1009        and then tried again after \r\n.
1010    
1011    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1012        in an attempt to make files that differ only in their line terminators
1013        compare equal. This works on Linux.
1014    
1015    17. Under certain error circumstances pcregrep might try to free random memory
1016        as it exited. This is now fixed, thanks to valgrind.
1017    
1018    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1019        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1020        was because its rules for how to advance for /g after matching an empty
1021        string at the end of a line did not allow for this case. They now check for
1022        it specially.
1023    
1024    20. pcretest is supposed to handle patterns and data of any length, by
1025        extending its buffers when necessary. It was getting this wrong when the
1026        buffer for a data line had to be extended.
1027    
1028    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1029        CRLF as a newline sequence.
1030    
1031    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1032        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1033        I have nevertheless tidied it up.
1034    
1035    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1036    
1037    24. Added a man page for pcre-config.
1038    
1039    
1040    Version 7.0 19-Dec-06
1041    ---------------------
1042    
1043     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1044        moving to gcc 4.1.1.
1045    
1046     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1047        sys/time.h, which is documented as needed for this function. It doesn't
1048        seem to matter on Linux, but it showed up on some releases of OS X.
1049    
1050     3. It seems that there are systems where bytes whose values are greater than
1051        127 match isprint() in the "C" locale. The "C" locale should be the
1052        default when a C program starts up. In most systems, only ASCII printing
1053        characters match isprint(). This difference caused the output from pcretest
1054        to vary, making some of the tests fail. I have changed pcretest so that:
1055    
1056        (a) When it is outputting text in the compiled version of a pattern, bytes
1057            other than 32-126 are always shown as hex escapes.
1058    
1059        (b) When it is outputting text that is a matched part of a subject string,
1060            it does the same, unless a different locale has been set for the match
1061            (using the /L modifier). In this case, it uses isprint() to decide.
1062    
1063     4. Fixed a major bug that caused incorrect computation of the amount of memory
1064        required for a compiled pattern when options that changed within the
1065        pattern affected the logic of the preliminary scan that determines the
1066        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1067        that the computed length was too small. The symptoms of this bug were
1068        either the PCRE error "internal error: code overflow" from pcre_compile(),
1069        or a glibc crash with a message such as "pcretest: free(): invalid next
1070        size (fast)". Examples of patterns that provoked this bug (shown in
1071        pcretest format) are:
1072    
1073          /(?-x: )/x
1074          /(?x)(?-x: \s*#\s*)/
1075          /((?i)[\x{c0}])/8
1076          /(?i:[\x{c0}])/8
1077    
1078        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1079        is now done differently.
1080    
1081     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1082        wrapper classes; (b) implement a new function in the C++ scanner that is
1083        more efficient than the old way of doing things because it avoids levels of
1084        recursion in the regex matching; (c) add a paragraph to the documentation
1085        for the FullMatch() function.
1086    
1087     6. The escape sequence \n was being treated as whatever was defined as
1088        "newline". Not only was this contrary to the documentation, which states
1089        that \n is character 10 (hex 0A), but it also went horribly wrong when
1090        "newline" was defined as CRLF. This has been fixed.
1091    
1092     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1093        was being set to -1 for the "end of line" case (supposedly a value that no
1094        character can have). Though this value is never used (the check for end of
1095        line is "zero bytes in current character"), it caused compiler complaints.
1096        I've changed it to 0xffffffff.
1097    
1098     8. In pcre_version.c, the version string was being built by a sequence of
1099        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1100        string (as it is for production releases) called a macro with an empty
1101        argument. The C standard says the result of this is undefined. The gcc
1102        compiler treats it as an empty string (which was what was wanted) but it is
1103        reported that Visual C gives an error. The source has been hacked around to
1104        avoid this problem.
1105    
1106     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1107        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1108        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1109        of them did).
1110    
1111    10. Originally, pcretest opened its input and output without "b"; then I was
1112        told that "b" was needed in some environments, so it was added for release
1113        5.0 to both the input and output. (It makes no difference on Unix-like
1114        systems.) Later I was told that it is wrong for the input on Windows. I've
1115        now abstracted the modes into two macros, to make it easier to fiddle with
1116        them, and removed "b" from the input mode under Windows.
1117    
1118    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1119    
1120    12. Added -help and --help to pcretest as an official way of being reminded
1121        of the options.
1122    
1123    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1124        and pcrecpp.cc because they annoy compilers at high warning levels.
1125    
1126    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1127    
1128    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1129        scripts are not C programs :-) and which was not noticed because it works
1130        on Linux.
1131    
1132    16. pcretest is supposed to handle any length of pattern and data line (as one
1133        line or as a continued sequence of lines) by extending its input buffer if
1134        necessary. This feature was broken for very long pattern lines, leading to
1135        a string of junk being passed to pcre_compile() if the pattern was longer
1136        than about 50K.
1137    
1138    17. I have done a major re-factoring of the way pcre_compile() computes the
1139        amount of memory needed for a compiled pattern. Previously, there was code
1140        that made a preliminary scan of the pattern in order to do this. That was
1141        OK when PCRE was new, but as the facilities have expanded, it has become
1142        harder and harder to keep it in step with the real compile phase, and there
1143        have been a number of bugs (see for example, 4 above). I have now found a
1144        cunning way of running the real compile function in a "fake" mode that
1145        enables it to compute how much memory it would need, while actually only
1146        ever using a few hundred bytes of working memory and without too many
1147        tests of the mode. This should make future maintenance and development
1148        easier. A side effect of this work is that the limit of 200 on the nesting
1149        depth of parentheses has been removed (though this was never a serious
1150        limitation, I suspect). However, there is a downside: pcre_compile() now
1151        runs more slowly than before (30% or more, depending on the pattern). I
1152        hope this isn't a big issue. There is no effect on runtime performance.
1153    
1154    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1155        newline (only possible for the last line of a file) and it was a
1156        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1157    
1158    19. Added additional timing features to pcretest. (1) The -tm option now times
1159        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1160        separate command line item, by a number that specifies the number of
1161        repeats to use when timing. The default is 50000; this gives better
1162        precision, but takes uncomfortably long for very large patterns.
1163    
1164    20. Extended pcre_study() to be more clever in cases where a branch of a
1165        subpattern has no definite first character. For example, (a*|b*)[cd] would
1166        previously give no result from pcre_study(). Now it recognizes that the
1167        first character must be a, b, c, or d.
1168    
1169    21. There was an incorrect error "recursive call could loop indefinitely" if
1170        a subpattern (or the entire pattern) that was being tested for matching an
1171        empty string contained only one non-empty item after a nested subpattern.
1172        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1173        incorrectly, because the \d was being skipped in the check.
1174    
1175    22. The pcretest program now has a new pattern option /B and a command line
1176        option -b, which is equivalent to adding /B to every pattern. This causes
1177        it to show the compiled bytecode, without the additional information that
1178        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1179        is the same as /B/I).
1180    
1181    23. A new optimization is now able automatically to treat some sequences such
1182        as a*b as a*+b. More specifically, if something simple (such as a character
1183        or a simple class like \d) has an unlimited quantifier, and is followed by
1184        something that cannot possibly match the quantified thing, the quantifier
1185        is automatically "possessified".
1186    
1187    24. A recursive reference to a subpattern whose number was greater than 39
1188        went wrong under certain circumstances in UTF-8 mode. This bug could also
1189        have affected the operation of pcre_study().
1190    
1191    25. Realized that a little bit of performance could be had by replacing
1192        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1193    
1194    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1195    
1196    27. Possessive quantifiers such as a++ were previously implemented by turning
1197        them into atomic groups such as ($>a+). Now they have their own opcodes,
1198        which improves performance. This includes the automatically created ones
1199        from 23 above.
1200    
1201    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1202        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1203        the first matched character to be a colon. This applied both to named and
1204        numbered groups.
1205    
1206    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1207    
1208    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1209        building PCRE on Windows easier, so I have included it in the distribution.
1210    
1211    31. There is now a check in pcretest against a ridiculously large number being
1212        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1213        loop, the loop is abandoned.
1214    
1215    32. Forward references to subpatterns in conditions such as (?(2)...) where
1216        subpattern 2 is defined later cause pcre_compile() to search forwards in
1217        the pattern for the relevant set of parentheses. This search went wrong
1218        when there were unescaped parentheses in a character class, parentheses
1219        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1220    
1221    33. "Subroutine" calls and backreferences were previously restricted to
1222        referencing subpatterns earlier in the regex. This restriction has now
1223        been removed.
1224    
1225    34. Added a number of extra features that are going to be in Perl 5.10. On the
1226        whole, these are just syntactic alternatives for features that PCRE had
1227        previously implemented using the Python syntax or my own invention. The
1228        other formats are all retained for compatibility.
1229    
1230        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1231            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1232            also .NET compatible.
1233    
1234        (b) A recursion or subroutine call to a named group can now be defined as
1235            (?&name) as well as (?P>name).
1236    
1237        (c) A backreference to a named group can now be defined as \k<name> or
1238            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1239            5.10, are also .NET compatible.
1240    
1241        (d) A conditional reference to a named group can now use the syntax
1242            (?(<name>) or (?('name') as well as (?(name).
1243    
1244        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1245            groups (named and numbered) that are never evaluated inline, but can be
1246            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1247            is always false. There may be only one alternative in such a group.
1248    
1249        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1250            as the simple (?(R). The condition is true only if the most recent
1251            recursion is that of the given number or name. It does not search out
1252            through the entire recursion stack.
1253    
1254        (g) The escape \gN or \g{N} has been added, where N is a positive or
1255            negative number, specifying an absolute or relative reference.
1256    
1257    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1258        some "unreachable code" warnings.
1259    
1260    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1261        things, this adds five new scripts.
1262    
1263    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1264        There were also incompatibilities regarding the handling of \Q..\E inside
1265        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1266        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1267    
1268    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1269        matches an empty string, and forcibly breaks the loop. There were bugs in
1270        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1271        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1272        separate and independent bugs (that affected different cases) have been
1273        fixed.
1274    
1275    39. Refactored the code to abolish the use of different opcodes for small
1276        capturing bracket numbers. This is a tidy that I avoided doing when I
1277        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1278        The new approach is not only tidier, it makes it possible to reduce the
1279        memory needed to fix the previous bug (38).
1280    
1281    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1282        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1283        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1284        mode.
1285    
1286    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1287        report.
1288    
1289    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1290        copy construction and assignment in the C++ wrapper.
1291    
1292    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1293        couple of bugs that could have given wrong results in the "--newline=crlf"
1294        case.
1295    
1296    44. Added a number of casts and did some reorganization of signed/unsigned int
1297        variables following suggestions from Dair Grant. Also renamed the variable
1298        "this" as "item" because it is a C++ keyword.
1299    
1300    45. Arranged for dftables to add
1301    
1302          #include "pcre_internal.h"
1303    
1304        to pcre_chartables.c because without it, gcc 4.x may remove the array
1305        definition from the final binary if PCRE is built into a static library and
1306        dead code stripping is activated.
1307    
1308    46. For an unanchored pattern, if a match attempt fails at the start of a
1309        newline sequence, and the newline setting is CRLF or ANY, and the next two
1310        characters are CRLF, advance by two characters instead of one.
1311    
1312    
1313    Version 6.7 04-Jul-06
1314    ---------------------
1315    
1316     1. In order to handle tests when input lines are enormously long, pcretest has
1317        been re-factored so that it automatically extends its buffers when
1318        necessary. The code is crude, but this _is_ just a test program. The
1319        default size has been increased from 32K to 50K.
1320    
1321     2. The code in pcre_study() was using the value of the re argument before
1322        testing it for NULL. (Of course, in any sensible call of the function, it
1323        won't be NULL.)
1324    
1325     3. The memmove() emulation function in pcre_internal.h, which is used on
1326        systems that lack both memmove() and bcopy() - that is, hardly ever -
1327        was missing a "static" storage class specifier.
1328    
1329     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1330        containing an extended class (one that cannot be represented by a bitmap
1331        because it contains high-valued characters or Unicode property items, e.g.
1332        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1333        pattern, but PCRE should not loop if you do not (it no longer does).
1334        [Detail: two cases were found: (a) a repeated subpattern containing an
1335        extended class; (b) a recursive reference to a subpattern that followed a
1336        previous extended class. It wasn't skipping over the extended class
1337        correctly when UTF-8 mode was not set.]
1338    
1339     5. A negated single-character class was not being recognized as fixed-length
1340        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1341        compile error "lookbehind assertion is not fixed length".
1342    
1343     6. The RunPerlTest auxiliary script was showing an unexpected difference
1344        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1345        write a Perl script that can interpret lines of an input file either as
1346        byte characters or as UTF-8, which is what "perltest" was being required to
1347        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1348        can't do is switch easily at run time between having the "use utf8;" pragma
1349        or not. In the end, I fudged it by using the RunPerlTest script to insert
1350        "use utf8;" explicitly for the UTF-8 tests.
1351    
1352     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1353        the end of the subject string, contrary to the documentation and to what
1354        Perl does. This was true of both matching functions. Now it matches only at
1355        the start of the subject and immediately after *internal* newlines.
1356    
1357     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1358        a pointer to an int instead of a pointer to an unsigned long int. This
1359        caused problems on 64-bit systems.
1360    
1361     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1362        instance of the 'standard' template library not being so standard".
1363    
1364    10. There was no check on the number of named subpatterns nor the maximum
1365        length of a subpattern name. The product of these values is used to compute
1366        the size of the memory block for a compiled pattern. By supplying a very
1367        long subpattern name and a large number of named subpatterns, the size
1368        computation could be caused to overflow. This is now prevented by limiting
1369        the length of names to 32 characters, and the number of named subpatterns
1370        to 10,000.
1371    
1372    11. Subpatterns that are repeated with specific counts have to be replicated in
1373        the compiled pattern. The size of memory for this was computed from the
1374        length of the subpattern and the repeat count. The latter is limited to
1375        65535, but there was no limit on the former, meaning that integer overflow
1376        could in principle occur. The compiled length of a repeated subpattern is
1377        now limited to 30,000 bytes in order to prevent this.
1378    
1379    12. Added the optional facility to have named substrings with the same name.
1380    
1381    13. Added the ability to use a named substring as a condition, using the
1382        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1383        are numbers (not recommended). Forward references are permitted.
1384    
1385    14. Added forward references in named backreferences (if you see what I mean).
1386    
1387    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1388        pattern could run off the end of the subject. For example, the pattern
1389        "(?s)(.{1,5})"8 did this with the subject "ab".
1390    
1391    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1392        PCRE_CASELESS was set when matching characters that were quantified with ?
1393        or *.
1394    
1395    17. A character class other than a single negated character that had a minimum
1396        but no maximum quantifier - for example [ab]{6,} - was not handled
1397        correctly by pce_dfa_exec(). It would match only one character.
1398    
1399    18. A valid (though odd) pattern that looked like a POSIX character
1400        class but used an invalid character after [ (for example [[,abc,]]) caused
1401        pcre_compile() to give the error "Failed: internal error: code overflow" or
1402        in some cases to crash with a glibc free() error. This could even happen if
1403        the pattern terminated after [[ but there just happened to be a sequence of
1404        letters, a binary zero, and a closing ] in the memory that followed.
1405    
1406    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1407        over the years. Originally (before any Unicode support), just the bottom 8
1408        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1409        output from "man perlunicode" includes this:
1410    
1411          The regular expression compiler produces polymorphic opcodes.  That
1412          is, the pattern adapts to the data and automatically switches to
1413          the Unicode character scheme when presented with Unicode data--or
1414          instead uses a traditional byte scheme when presented with byte
1415          data.
1416    
1417        Sadly, a wide octal escape does not cause a switch, and in a string with
1418        no other multibyte characters, these octal escapes are treated as before.
1419        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1420        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1421        Unicode string.
1422    
1423        I have not perpetrated such confusion in PCRE. Up till now, it took just
1424        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1425        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1426        translate to the appropriate multibyte character.
1427    
1428    29. Applied some refactoring to reduce the number of warnings from Microsoft
1429        and Borland compilers. This has included removing the fudge introduced
1430        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1431        a warning about an unused variable.
1432    
1433    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1434        characters since release 4.0, because Perl (from release 5.004) does not.
1435        [Or at least, is documented not to: some releases seem to be in conflict
1436        with the documentation.] However, when a pattern was studied with
1437        pcre_study() and all its branches started with \s, PCRE still included VT
1438        as a possible starting character. Of course, this did no harm; it just
1439        caused an unnecessary match attempt.
1440    
1441    22. Removed a now-redundant internal flag bit that recorded the fact that case
1442        dependency changed within the pattern. This was once needed for "required
1443        byte" processing, but is no longer used. This recovers a now-scarce options
1444        bit. Also moved the least significant internal flag bit to the most-
1445        significant bit of the word, which was not previously used (hangover from
1446        the days when it was an int rather than a uint) to free up another bit for
1447        the future.
1448    
1449    23. Added support for CRLF line endings as well as CR and LF. As well as the
1450        default being selectable at build time, it can now be changed at runtime
1451        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1452        specify that it is scanning data with non-default line endings.
1453    
1454    24. Changed the definition of CXXLINK to make it agree with the definition of
1455        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1456    
1457    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1458        recursions. This makes a big different to stack usage for some patterns.
1459    
1460    26. If a subpattern containing a named recursion or subroutine reference such
1461        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1462        the space required for the compiled pattern went wrong and gave too small a
1463        value. Depending on the environment, this could lead to "Failed: internal
1464        error: code overflow at offset 49" or "glibc detected double free or
1465        corruption" errors.
1466    
1467    27. Applied patches from Google (a) to support the new newline modes and (b) to
1468        advance over multibyte UTF-8 characters in GlobalReplace.
1469    
1470    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1471        difference for some implementation of PCRE in some Windows version.
1472    
1473    29. Added some extra testing facilities to pcretest:
1474    
1475        \q<number>   in a data line sets the "match limit" value
1476        \Q<number>   in a data line sets the "match recursion limt" value
1477        -S <number>  sets the stack size, where <number> is in megabytes
1478    
1479        The -S option isn't available for Windows.
1480    
1481    
1482    Version 6.6 06-Feb-06
1483    ---------------------
1484    
1485     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1486        in pcreposix.h. I have copied the definition from pcre.h.
1487    
1488     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1489        because pcre.h is no longer a built file.
1490    
1491     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1492        not normally included in the compiled code.
1493    
1494    
1495    Version 6.5 01-Feb-06
1496    ---------------------
1497    
1498     1. When using the partial match feature with pcre_dfa_exec(), it was not
1499        anchoring the second and subsequent partial matches at the new starting
1500        point. This could lead to incorrect results. For example, with the pattern
1501        /1234/, partially matching against "123" and then "a4" gave a match.
1502    
1503     2. Changes to pcregrep:
1504    
1505        (a) All non-match returns from pcre_exec() were being treated as failures
1506            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1507            error message is output. Some extra information is given for the
1508            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1509            probably the only errors that are likely to be caused by users (by
1510            specifying a regex that has nested indefinite repeats, for instance).
1511            If there are more than 20 of these errors, pcregrep is abandoned.
1512    
1513        (b) A binary zero was treated as data while matching, but terminated the
1514            output line if it was written out. This has been fixed: binary zeroes
1515            are now no different to any other data bytes.
1516    
1517        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1518            used to set a locale for matching. The --locale=xxxx long option has
1519            been added (no short equivalent) to specify a locale explicitly on the
1520            pcregrep command, overriding the environment variables.
1521    
1522        (d) When -B was used with -n, some line numbers in the output were one less
1523            than they should have been.
1524    
1525        (e) Added the -o (--only-matching) option.
1526    
1527        (f) If -A or -C was used with -c (count only), some lines of context were
1528            accidentally printed for the final match.
1529    
1530        (g) Added the -H (--with-filename) option.
1531    
1532        (h) The combination of options -rh failed to suppress file names for files
1533            that were found from directory arguments.
1534    
1535        (i) Added the -D (--devices) and -d (--directories) options.
1536    
1537        (j) Added the -F (--fixed-strings) option.
1538    
1539        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1540    
1541        (l) Added the --colo(u)r option.
1542    
1543        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1544            is not present by default.
1545    
1546     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1547        items such as (?R) or (?1), when the recursion could match a number of
1548        alternatives. If it matched one of the alternatives, but subsequently,
1549        outside the recursion, there was a failure, the code tried to back up into
1550        the recursion. However, because of the way PCRE is implemented, this is not
1551        possible, and the result was an incorrect result from the match.
1552    
1553        In order to prevent this happening, the specification of recursion has
1554        been changed so that all such subpatterns are automatically treated as
1555        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1556    
1557     4. I had overlooked the fact that, in some locales, there are characters for
1558        which isalpha() is true but neither isupper() nor islower() are true. In
1559        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1560        and ordfeminine) are like this. This affected the treatment of \w and \W
1561        when they appeared in character classes, but not when they appeared outside
1562        a character class. The bit map for "word" characters is now created
1563        separately from the results of isalnum() instead of just taking it from the
1564        upper, lower, and digit maps. (Plus the underscore character, of course.)
1565    
1566     5. The above bug also affected the handling of POSIX character classes such as
1567        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1568        permanent tables. Instead, the bit maps for such a class were previously
1569        created as the appropriate unions of the upper, lower, and digit bitmaps.
1570        Now they are created by subtraction from the [[:word:]] class, which has
1571        its own bitmap.
1572    
1573     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1574        It is created by subtracting the vertical space characters (\x09, \x0a,
1575        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1576        subtraction was done in the overall bitmap for a character class, meaning
1577        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1578        be recognized. This bug has been fixed.
1579    
1580     7. Patches from the folks at Google:
1581    
1582          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1583          real life, but is still worth protecting against".
1584    
1585          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1586          regular expressions".
1587    
1588          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1589          have it.
1590    
1591          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1592          "configure" and the latter not, in order to fix a problem somebody had
1593          with compiling the Arg class on HP-UX.
1594    
1595          (e) Improve the error-handling of the C++ wrapper a little bit.
1596    
1597          (f) New tests for checking recursion limiting.
1598    
1599     8. The pcre_memmove() function, which is used only if the environment does not
1600        have a standard memmove() function (and is therefore rarely compiled),
1601        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1602        returning a result (though PCRE never actually uses the result).
1603    
1604     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1605        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1606        returned instead of calling malloc() with an overflowing number that would
1607        most likely cause subsequent chaos.
1608    
1609    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1610    
1611    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1612        with this option is matched, the nmatch and pmatch options of regexec() are
1613        ignored.
1614    
1615    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1616        provided in case anyone wants to the the POSIX interface with UTF-8
1617        strings.
1618    
1619    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1620        C++ linking (needed for some HP-UX environments).
1621    
1622    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1623        (unused parameter) and in the pcre_printint() function (omitted "default"
1624        switch label when the default is to do nothing).
1625    
1626    15. Added some code to make it possible, when PCRE is compiled as a C++
1627        library, to replace subject pointers for pcre_exec() with a smart pointer
1628        class, thus making it possible to process discontinuous strings.
1629    
1630    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1631        much the same function. They were added by different people who were trying
1632        to make PCRE easy to compile on non-Unix systems. It has been suggested
1633        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1634        for compiling on Windows systems. I have therefore replaced it with
1635        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1636        defaults to "extern" for C or "extern C" for C++, which works fine on
1637        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1638        SCOPE with something explicit in config.h. In addition:
1639    
1640        (a) pcreposix.h still had just "extern" instead of either of these macros;
1641            I have replaced it with PCRE_DATA_SCOPE.
1642    
1643        (b) Functions such as _pcre_xclass(), which are internal to the library,
1644            but external in the C sense, all had PCRE_EXPORT in their definitions.
1645            This is apparently wrong for the Windows case, so I have removed it.
1646            (It makes no difference on Unix-like systems.)
1647    
1648    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1649        of recursive calls to match(). This is different to MATCH_LIMIT because
1650        that limits the total number of calls to match(), not all of which increase
1651        the depth of recursion. Limiting the recursion depth limits the amount of
1652        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1653        when PCRE is compiled, and changed at run time. A patch from Google adds
1654        this functionality to the C++ interface.
1655    
1656    18. Changes to the handling of Unicode character properties:
1657    
1658        (a) Updated the table to Unicode 4.1.0.
1659    
1660        (b) Recognize characters that are not in the table as "Cn" (undefined).
1661    
1662        (c) I revised the way the table is implemented to a much improved format
1663            which includes recognition of ranges. It now supports the ranges that
1664            are defined in UnicodeData.txt, and it also amalgamates other
1665            characters into ranges. This has reduced the number of entries in the
1666            table from around 16,000 to around 3,000, thus reducing its size
1667            considerably. I realized I did not need to use a tree structure after
1668            all - a binary chop search is just as efficient. Having reduced the
1669            number of entries, I extended their size from 6 bytes to 8 bytes to
1670            allow for more data.
1671    
1672        (d) Added support for Unicode script names via properties such as \p{Han}.
1673    
1674    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1675        matching that character.
1676    
1677    20. When matching a repeated Unicode property with a minimum greater than zero,
1678        (for example \pL{2,}), PCRE could look past the end of the subject if it
1679        reached it while seeking the minimum number of characters. This could
1680        happen only if some of the characters were more than one byte long, because
1681        there is a check for at least the minimum number of bytes.
1682    
1683    21. Refactored the implementation of \p and \P so as to be more general, to
1684        allow for more different types of property in future. This has changed the
1685        compiled form incompatibly. Anybody with saved compiled patterns that use
1686        \p or \P will have to recompile them.
1687    
1688    22. Added "Any" and "L&" to the supported property types.
1689    
1690    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1691        but give a compile time error if the value is greater than 0xff.
1692    
1693    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1694        accidentally not being installed or uninstalled.
1695    
1696    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1697        made were to insert the current release number. This seemed silly, because
1698        it made things harder for people building PCRE on systems that don't run
1699        "configure". I have turned pcre.h into a distributed file, no longer built
1700        by "configure", with the version identification directly included. There is
1701        no longer a pcre.h.in file.
1702    
1703        However, this change necessitated a change to the pcre-config script as
1704        well. It is built from pcre-config.in, and one of the substitutions was the
1705        release number. I have updated configure.ac so that ./configure now finds
1706        the release number by grepping pcre.h.
1707    
1708    26. Added the ability to run the tests under valgrind.
1709    
1710    
1711    Version 6.4 05-Sep-05
1712    ---------------------
1713    
1714     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1715        "--" to be printed when multiple files were scanned, even when none of the
1716        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1717        consider it to be a bug, and have restored the previous behaviour.
1718    
1719     2. A couple of code tidies to get rid of compiler warnings.
1720    
1721     3. The pcretest program used to cheat by referring to symbols in the library
1722        whose names begin with _pcre_. These are internal symbols that are not
1723        really supposed to be visible externally, and in some environments it is
1724        possible to suppress them. The cheating is now confined to including
1725        certain files from the library's source, which is a bit cleaner.
1726    
1727     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1728        file's purpose clearer.
1729    
1730     5. Reorganized pcre_ucp_findchar().
1731    
1732    
1733    Version 6.3 15-Aug-05
1734    ---------------------
1735    
1736     1. The file libpcre.pc.in did not have general read permission in the tarball.
1737    
1738     2. There were some problems when building without C++ support:
1739    
1740        (a) If C++ support was not built, "make install" and "make test" still
1741            tried to test it.
1742    
1743        (b) There were problems when the value of CXX was explicitly set. Some
1744            changes have been made to try to fix these, and ...
1745    
1746        (c) --disable-cpp can now be used to explicitly disable C++ support.
1747    
1748        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1749            backslash in a target when C++ was disabled. This confuses some
1750            versions of "make", apparently. Using an intermediate variable solves
1751            this. (Same for CPP_LOBJ.)
1752    
1753     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1754        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1755        necessary on certain architectures.
1756    
1757     4. Added a setting of -export-symbols-regex to the link command to remove
1758        those symbols that are exported in the C sense, but actually are local
1759        within the library, and not documented. Their names all begin with
1760        "_pcre_". This is not a perfect job, because (a) we have to except some
1761        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1762        available (and never for static libraries). I have made a note to try to
1763        find a way round (a) in the future.
1764    
1765    
1766    Version 6.2 01-Aug-05
1767    ---------------------
1768    
1769     1. There was no test for integer overflow of quantifier values. A construction
1770        such as {1111111111111111} would give undefined results. What is worse, if
1771        a minimum quantifier for a parenthesized subpattern overflowed and became
1772        negative, the calculation of the memory size went wrong. This could have
1773        led to memory overwriting.
1774    
1775     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1776    
1777     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1778        operating environments where this matters.
1779    
1780     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1781        PCRE options from within the C++ wrapper.
1782    
1783     5. Named capturing subpatterns were not being correctly counted when a pattern
1784        was compiled. This caused two problems: (a) If there were more than 100
1785        such subpatterns, the calculation of the memory needed for the whole
1786        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1787        back references of the form \12, where the number was greater than 9, were
1788        not recognized as back references, even though there were sufficient
1789        previous subpatterns.
1790    
1791     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1792        versions of gcc, e.g. 2.95.4.
1793    
1794    
1795    Version 6.1 21-Jun-05
1796    ---------------------
1797    
1798     1. There was one reference to the variable "posix" in pcretest.c that was not
1799        surrounded by "#if !defined NOPOSIX".
1800    
1801     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1802        the cross-check on the old pcre_info() function, for the benefit of the
1803        cut-down version of PCRE that is currently imported into Exim.
1804    
1805     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1806        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1807        patterns that start (?i) but I don't think that matters. The use of (?i) is
1808        just an example; this all applies to the other options as well.
1809    
1810     4. Since libtool seems to echo the compile commands it is issuing, the output
1811        from "make" can be reduced a bit by putting "@" in front of each libtool
1812        compile command.
1813    
1814     5. Patch from the folks at Google for configure.in to be a bit more thorough
1815        in checking for a suitable C++ installation before trying to compile the
1816        C++ stuff. This should fix a reported problem when a compiler was present,
1817        but no suitable headers.
1818    
1819     6. The man pages all had just "PCRE" as their title. I have changed them to
1820        be the relevant file name. I have also arranged that these names are
1821        retained in the file doc/pcre.txt, which is a concatenation in text format
1822        of all the man pages except the little individual ones for each function.
1823    
1824     7. The NON-UNIX-USE file had not been updated for the different set of source
1825        files that come with release 6. I also added a few comments about the C++
1826        wrapper.
1827    
1828    
1829    Version 6.0 07-Jun-05
1830    ---------------------
1831    
1832     1. Some minor internal re-organization to help with my DFA experiments.
1833    
1834     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1835        didn't matter for the library itself when fully configured, but did matter
1836        when compiling without UCP support, or within Exim, where the ucp files are
1837        not imported.
1838    
1839     3. Refactoring of the library code to split up the various functions into
1840        different source modules. The addition of the new DFA matching code (see
1841        below) to a single monolithic source would have made it really too
1842        unwieldy, quite apart from causing all the code to be include in a
1843        statically linked application, when only some functions are used. This is
1844        relevant even without the DFA addition now that patterns can be compiled in
1845        one application and matched in another.
1846    
1847        The downside of splitting up is that there have to be some external
1848        functions and data tables that are used internally in different modules of
1849        the library but which are not part of the API. These have all had their
1850        names changed to start with "_pcre_" so that they are unlikely to clash
1851        with other external names.
1852    
1853     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1854        a different (DFA) algorithm. Although it is slower than the original
1855        function, it does have some advantages for certain types of matching
1856        problem.
1857    
1858     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1859        including restarting after a partial match.
1860    
1861     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1862        defined when compiling for Windows was sent to me. I have put it into the
1863        code, though I have no means of testing or verifying it.
1864    
1865     7. Added the pcre_refcount() auxiliary function.
1866    
1867     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1868        match before or at the first newline in the subject string. In pcretest,
1869        the /f option on a pattern can be used to set this.
1870    
1871     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1872        would behave wrongly. This has been present in PCRE since release 4.0.
1873    
1874    10. A number of changes to the pcregrep command:
1875    
1876        (a) Refactored how -x works; insert ^(...)$ instead of setting
1877            PCRE_ANCHORED and checking the length, in preparation for adding
1878            something similar for -w.
1879    
1880        (b) Added the -w (match as a word) option.
1881    
1882        (c) Refactored the way lines are read and buffered so as to have more
1883            than one at a time available.
1884    
1885        (d) Implemented a pcregrep test script.
1886    
1887        (e) Added the -M (multiline match) option. This allows patterns to match
1888            over several lines of the subject. The buffering ensures that at least
1889            8K, or the rest of the document (whichever is the shorter) is available
1890            for matching (and similarly the previous 8K for lookbehind assertions).
1891    
1892        (f) Changed the --help output so that it now says
1893    
1894              -w, --word-regex(p)
1895    
1896            instead of two lines, one with "regex" and the other with "regexp"
1897            because that confused at least one person since the short forms are the
1898            same. (This required a bit of code, as the output is generated
1899            automatically from a table. It wasn't just a text change.)
1900    
1901        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1902            option but starts with a hyphen. Could be a pattern or a path name
1903            starting with a hyphen, for instance.
1904    
1905        (h) "-" can be given as a file name to represent stdin.
1906    
1907        (i) When file names are being printed, "(standard input)" is used for
1908            the standard input, for compatibility with GNU grep. Previously
1909            "<stdin>" was used.
1910    
1911        (j) The option --label=xxx can be used to supply a name to be used for
1912            stdin when file names are being printed. There is no short form.
1913    
1914        (k) Re-factored the options decoding logic because we are going to add
1915            two more options that take data. Such options can now be given in four
1916            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1917    
1918        (l) Added the -A, -B, and -C options for requesting that lines of context
1919            around matches be printed.
1920    
1921        (m) Added the -L option to print the names of files that do not contain
1922            any matching lines, that is, the complement of -l.
1923    
1924        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1925            continue to scan other files.
1926    
1927        (o) The -s option was incorrectly implemented. For compatibility with other
1928            greps, it now suppresses the error message for a non-existent or non-
1929            accessible file (but not the return code). There is a new option called
1930            -q that suppresses the output of matching lines, which was what -s was
1931            previously doing.
1932    
1933        (p) Added --include and --exclude options to specify files for inclusion
1934            and exclusion when recursing.
1935    
1936    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1937        Hopefully, it now does.
1938    
1939    12. Missing cast in pcre_study().
1940    
1941    13. Added an "uninstall" target to the makefile.
1942    
1943    14. Replaced "extern" in the function prototypes in Makefile.in with
1944        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1945        world, but is set differently for Windows.
1946    
1947    15. Added a second compiling function called pcre_compile2(). The only
1948        difference is that it has an extra argument, which is a pointer to an
1949        integer error code. When there is a compile-time failure, this is set
1950        non-zero, in addition to the error test pointer being set to point to an
1951        error message. The new argument may be NULL if no error number is required
1952        (but then you may as well call pcre_compile(), which is now just a
1953        wrapper). This facility is provided because some applications need a
1954        numeric error indication, but it has also enabled me to tidy up the way
1955        compile-time errors are handled in the POSIX wrapper.
1956    
1957    16. Added VPATH=.libs to the makefile; this should help when building with one
1958        prefix path and installing with another. (Or so I'm told by someone who
1959        knows more about this stuff than I do.)
1960    
1961    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1962        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1963        match everything, including newlines. This is not POSIX-compatible, but
1964        somebody wanted the feature. From pcretest it can be activated by using
1965        both the P and the s flags.
1966    
1967    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1968    
1969    19. libpcre.pc was being incorrectly installed as executable.
1970    
1971    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1972        it now also looks for '\r' so that it will work unmodified on Windows.
1973    
1974    21. Added Google's contributed C++ wrapper to the distribution.
1975    
1976    22. Added some untidy missing memory free() calls in pcretest, to keep
1977        Electric Fence happy when testing.
1978    
1979    
1980    
1981    Version 5.0 13-Sep-04
1982    ---------------------
1983    
1984     1. Internal change: literal characters are no longer packed up into items
1985        containing multiple characters in a single byte-string. Each character
1986        is now matched using a separate opcode. However, there may be more than one
1987        byte in the character in UTF-8 mode.
1988    
1989     2. The pcre_callout_block structure has two new fields: pattern_position and
1990        next_item_length. These contain the offset in the pattern to the next match
1991        item, and its length, respectively.
1992    
1993     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1994        insertion of callouts before each pattern item. Added the /C option to
1995        pcretest to make use of this.
1996    
1997     4. On the advice of a Windows user, the lines
1998    
1999          #if defined(_WIN32) || defined(WIN32)
2000          _setmode( _fileno( stdout ), 0x8000 );
2001          #endif  /* defined(_WIN32) || defined(WIN32) */
2002    
2003        have been added to the source of pcretest. This apparently does useful
2004        magic in relation to line terminators.
2005    
2006     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2007        for the benefit of those environments where the "b" makes a difference.
2008    
2009     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2010        to know about it. I have put a hack into configure.in that adds in code
2011        to set GCC=yes if CC=icc. This seems to end up at a point in the
2012        generated configure script that is early enough to affect the setting of
2013        compiler options, which is what is needed, but I have no means of testing
2014        whether it really works. (The user who reported this had patched the
2015        generated configure script, which of course I cannot do.)
2016    
2017        LATER: After change 22 below (new libtool files), the configure script
2018        seems to know about icc (and also ecc). Therefore, I have commented out
2019        this hack in configure.in.
2020    
2021     7. Added support for pkg-config (2 patches were sent in).
2022    
2023     8. Negated POSIX character classes that used a combination of internal tables
2024        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2025        [[:^ascii]]. Typically, they would match almost any characters. The other
2026        POSIX classes were not broken in this way.
2027    
2028     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2029        to find the match, as PCRE was deluded into thinking that the match had to
2030        start at the start point or following a newline. The same bug applied to
2031        patterns with negative forward assertions or any backward assertions
2032        preceding ".*" at the start, unless the pattern required a fixed first
2033        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2034    
2035    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2036        starting at the last subject character, bytes beyond the end of the subject
2037        string were read.
2038    
2039    11. Renamed the variable "class" as "classbits" to make life easier for C++
2040        users. (Previously there was a macro definition, but it apparently wasn't
2041        enough.)
2042    
2043    12. Added the new field "tables" to the extra data so that tables can be passed
2044        in at exec time, or the internal tables can be re-selected. This allows
2045        a compiled regex to be saved and re-used at a later time by a different
2046        program that might have everything at different addresses.
2047    
2048    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2049        -R library as well as a -L library.
2050    
2051    14. The debugging options of pcretest (-d on the command line or D on a
2052        pattern) showed incorrect output for anything following an extended class
2053        that contained multibyte characters and which was followed by a quantifier.
2054    
2055    15. Added optional support for general category Unicode character properties
2056        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2057        support. It adds about 90K to the size of the library. The meanings of the
2058        inbuilt class escapes such as \d and \s have NOT been changed.
2059    
2060    16. Updated pcredemo.c to include calls to free() to release the memory for the
2061        compiled pattern.
2062    
2063    17. The generated file chartables.c was being created in the source directory
2064        instead of in the building directory. This caused the build to fail if the
2065        source directory was different from the building directory, and was
2066        read-only.
2067    
2068    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2069        file. No doubt somebody will tell me if they don't make sense... Also added
2070        Dan Mooney's comments about building on OpenVMS.
2071    
2072    19. Added support for partial matching via the PCRE_PARTIAL option for
2073        pcre_exec() and the \P data escape in pcretest.
2074    
2075    20. Extended pcretest with 3 new pattern features:
2076    
2077        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2078              write the compiled pattern to the file whose name is "rest-of-line".
2079              This is a straight binary dump of the data, with the saved pointer to
2080              the character tables forced to be NULL. The study data, if any, is
2081              written too. After writing, pcretest reads a new pattern.
2082    
2083        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2084              compiled pattern from the given file. There must not be any
2085              occurrences of "<" in the file name (pretty unlikely); if there are,
2086              pcretest will instead treat the initial "<" as a pattern delimiter.
2087              After reading in the pattern, pcretest goes on to read data lines as
2088              usual.
2089    
2090        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2091              and 16-bit fields in a compiled pattern, to simulate a pattern that
2092              was compiled on a host of opposite endianness.
2093    
2094    21. The pcre-exec() function can now cope with patterns that were compiled on
2095        hosts of opposite endianness, with this restriction:
2096    
2097          As for any compiled expression that is saved and used later, the tables
2098          pointer field cannot be preserved; the extra_data field in the arguments
2099          to pcre_exec() should be used to pass in a tables address if a value
2100          other than the default internal tables were used at compile time.
2101    
2102    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2103        now diagnosed as an error. Previously, most of the time, a negative number
2104        would have been treated as zero, but if in addition "ovector" was passed as
2105        NULL, a crash could occur.
2106    
2107    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2108        new versions from the libtool 1.5 distribution (the last one is a copy of
2109        a file called libtool.m4). This seems to have fixed the need to patch
2110        "configure" to support Darwin 1.3 (which I used to do). However, I still
2111        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2112        workstation).
2113    
2114    24. Changed the PCRE licence to be the more standard "BSD" licence.
2115    
2116    
2117    Version 4.5 01-Dec-03
2118    ---------------------
2119    
2120     1. There has been some re-arrangement of the code for the match() function so
2121        that it can be compiled in a version that does not call itself recursively.
2122        Instead, it keeps those local variables that need separate instances for
2123        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2124        needs to "recurse". Keeping track of where control must go is done by means
2125        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2126        hide most of the details from the main code, and operates only if
2127        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2128        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2129        operating.
2130    
2131        To make it easier for callers to provide specially tailored get/free
2132        functions for this usage, two new functions, pcre_stack_malloc, and
2133        pcre_stack_free, are used. They are always called in strict stacking order,
2134        and the size of block requested is always the same.
2135    
2136        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2137        PCRE has been compiled to use the stack or the heap for recursion. The
2138        -C option of pcretest uses this to show which version is compiled.
2139    
2140        A new data escape \S, is added to pcretest; it causes the amounts of store
2141        obtained and freed by both kinds of malloc/free at match time to be added
2142        to the output.
2143    
2144     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2145        what's available on my current Linux desktop machine.
2146    
2147     3. When matching a UTF-8 string, the test for a valid string at the start has
2148        been extended. If start_offset is not zero, PCRE now checks that it points
2149        to a byte that is the start of a UTF-8 character. If not, it returns
2150        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2151        this is necessary because there may be backward assertions in the pattern.
2152        When matching the same subject several times, it may save resources to use
2153        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2154    
2155     4. The code for checking the validity of UTF-8 strings has been tightened so
2156        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2157        containing "overlong sequences".
2158    
2159     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2160        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2161        should have been "&", but it just so happened that all the cases this let
2162        through by mistake were picked up later in the function.
2163    
2164     6. I had used a variable called "isblank" - this is a C99 function, causing
2165        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2166    
2167     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2168        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2169        so the version is shown; (c) stop "make test" echoing "./RunTest".
2170    
2171     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2172    
2173     9. The prototype for memmove() for systems that don't have it was using
2174        size_t, but the inclusion of the header that defines size_t was later. I've
2175        moved the #includes for the C headers earlier to avoid this.
2176    
2177    10. Added some adjustments to the code to make it easier to compiler on certain
2178        special systems:
2179    
2180          (a) Some "const" qualifiers were missing.
2181          (b) Added the macro EXPORT before all exported functions; by default this
2182              is defined to be empty.
2183          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2184              that it reads its output file name as an argument instead of writing
2185              to the standard output and assuming this can be redirected.
2186    
2187    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2188        class containing characters with values greater than 255, PCRE compilation
2189        went into a loop.
2190    
2191    12. A recursive reference to a subpattern that was within another subpattern
2192        that had a minimum quantifier of zero caused PCRE to crash. For example,
2193        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2194        recursion. If the recursively-called subpattern itself had a zero repeat,
2195        that was OK.
2196    
2197    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2198        buffer into which it was copied (for escape processing) was still set at
2199        1024, so long lines caused crashes.
2200    
2201    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2202        "internal error: code overflow...". This applied to any character class
2203        that was followed by a possessive quantifier.
2204    
2205    15. Modified the Makefile to add libpcre.la as a prerequisite for
2206        libpcreposix.la because I was told this is needed for a parallel build to
2207        work.
2208    
2209    16. If a pattern that contained .* following optional items at the start was
2210        studied, the wrong optimizing data was generated, leading to matching
2211        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2212        matching string must start with a or b or c. The correct conclusion for
2213        this pattern is that a match can start with any character.
2214    
2215    
2216    Version 4.4 13-Aug-03
2217    ---------------------
2218    
2219     1. In UTF-8 mode, a character class containing characters with values between
2220        127 and 255 was not handled correctly if the compiled pattern was studied.
2221        In fixing this, I have also improved the studying algorithm for such
2222        classes (slightly).
2223    
2224     2. Three internal functions had redundant arguments passed to them. Removal
2225        might give a very teeny performance improvement.
2226    
2227     3. Documentation bug: the value of the capture_top field in a callout is *one
2228        more than* the number of the hightest numbered captured substring.
2229    
2230     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2231        in incorrectly linking with a previously installed version. They now link
2232        explicitly with libpcre.la.
2233    
2234     5. configure.in no longer needs to recognize Cygwin specially.
2235    
2236     6. A problem in pcre.in for Windows platforms is fixed.
2237    
2238     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2239        pcretest, it used to include the size of the study block as part of its
2240        output. Unfortunately, the structure contains a field that has a different
2241        size on different hardware architectures. This meant that the tests that
2242        showed this size failed. As the block is currently always of a fixed size,
2243        this information isn't actually particularly useful in pcretest output, so
2244        I have just removed it.
2245    
2246     8. Three pre-processor statements accidentally did not start in column 1.
2247        Sadly, there are *still* compilers around that complain, even though
2248        standard C has not required this for well over a decade. Sigh.
2249    
2250     9. In pcretest, the code for checking callouts passed small integers in the
2251        callout_data field, which is a void * field. However, some picky compilers
2252        complained about the casts involved for this on 64-bit systems. Now
2253        pcretest passes the address of the small integer instead, which should get
2254        rid of the warnings.
2255    
2256    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2257        both compile and run time, and gives an error if an invalid UTF-8 sequence
2258        is found. There is a option for disabling this check in cases where the
2259        string is known to be correct and/or the maximum performance is wanted.
2260    
2261    11. In response to a bug report, I changed one line in Makefile.in from
2262    
2263            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2264        to
2265            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2266    
2267        to look similar to other lines, but I have no way of telling whether this
2268        is the right thing to do, as I do not use Windows. No doubt I'll get told
2269        if it's wrong...
2270    
2271    
2272    Version 4.3 21-May-03
2273    ---------------------
2274    
2275    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2276       Makefile.
2277    
2278    2. Some refactoring to improve the quality of the code:
2279    
2280       (i)   The utf8_table... variables are now declared "const".
2281    
2282       (ii)  The code for \cx, which used the "case flipping" table to upper case
2283             lower case letters, now just substracts 32. This is ASCII-specific,
2284             but the whole concept of \cx is ASCII-specific, so it seems
2285             reasonable.
2286    
2287       (iii) PCRE was using its character types table to recognize decimal and
2288             hexadecimal digits in the pattern. This is silly, because it handles
2289             only 0-9, a-f, and A-F, but the character types table is locale-
2290             specific, which means strange things might happen. A private
2291             table is now used for this - though it costs 256 bytes, a table is
2292             much faster than multiple explicit tests. Of course, the standard
2293             character types table is still used for matching digits in subject
2294             strings against \d.
2295    
2296       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2297             ending in _t are). So I've renamed it as ESC_tee.
2298    
2299    3. The first argument for regexec() in the POSIX wrapper should have been
2300       defined as "const".
2301    
2302    4. Changed pcretest to use malloc() for its buffers so that they can be
2303       Electric Fenced for debugging.
2304    
2305    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2306       to read one or more bytes before the start of the subject string. Often this
2307       had no effect on PCRE's behaviour, but in some circumstances it could
2308       provoke a segmentation fault.
2309    
2310    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2311       to try to read one or more bytes before the start of the subject string.
2312    
2313    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2314       UTF-8 support could misbehave in various ways if the subject string
2315       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2316       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2317       back over UTF-8 characters.)
2318    
2319    
2320    Version 4.2 14-Apr-03
2321    ---------------------
2322    
2323    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2324    
2325    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2326         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2327         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2328         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2329         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2330           and BUILD_EXEEXT
2331         Note: automatic setting of the BUILD variables is not yet working
2332         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2333           compile-time but not at link-time
2334         [LINK]: use for linking executables only
2335         make different versions for Windows and non-Windows
2336         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2337           libraries
2338         [LINK_FOR_BUILD]: new variable
2339         [OBJEXT]: use throughout
2340         [EXEEXT]: use throughout
2341         <winshared>: new target
2342         <wininstall>: new target
2343         <dftables.o>: use native compiler
2344         <dftables>: use native linker
2345         <install>: handle Windows platform correctly
2346         <clean>: ditto
2347         <check>: ditto
2348         copy DLL to top builddir before testing
2349    
2350       As part of these changes, -no-undefined was removed again. This was reported
2351       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2352       in any case.
2353    
2354    3. Some tidies to get rid of compiler warnings:
2355    
2356       . In the match_data structure, match_limit was an unsigned long int, whereas
2357         match_call_count was an int. I've made them both unsigned long ints.
2358    
2359       . In pcretest the fact that a const uschar * doesn't automatically cast to
2360         a void * provoked a warning.
2361    
2362       . Turning on some more compiler warnings threw up some "shadow" variables
2363         and a few more missing casts.
2364    
2365    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2366       option, a class that contained a single character with a value between 128
2367       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2368    
2369    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2370       option, a class that contained several characters, but with at least one
2371       whose value was between 128 and 255 caused PCRE to crash.
2372    
2373    
2374    Version 4.1 12-Mar-03
2375    ---------------------
2376    
2377    1. Compiling with gcc -pedantic found a couple of places where casts were
2378    needed, and a string in dftables.c that was longer than standard compilers are
2379    required to support.
2380    
2381    2. Compiling with Sun's compiler found a few more places where the code could
2382    be tidied up in order to avoid warnings.
2383    
2384    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2385    first of these names is deprecated in the latest Autoconf in favour of the name
2386    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2387    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2388    analogy I have changed it to CFLAGS_FOR_BUILD.
2389    
2390    4. Added -no-undefined to the linking command in the Makefile, because this is
2391    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2392    linking step for the pcreposix library.
2393    
2394    5. PCRE was failing to diagnose the case of two named groups with the same
2395    name.
2396    
2397    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2398    literal character that is needed in the subject for a match, and scans along to
2399    ensure that it is present before embarking on the full matching process. This
2400    saves time in cases of nested unlimited repeats that are never going to match.
2401    Problem: the scan can take a lot of time if the subject is very long (e.g.
2402    megabytes), thus penalizing straightforward matches. It is now done only if the
2403    amount of subject to be scanned is less than 1000 bytes.
2404    
2405    7. A lesser problem with the same optimization is that it was recording the
2406    first character of an anchored pattern as "needed", thus provoking a search
2407    right along the subject, even when the first match of the pattern was going to
2408    fail. The "needed" character is now not set for anchored patterns, unless it
2409    follows something in the pattern that is of non-fixed length. Thus, it still
2410    fulfils its original purpose of finding quick non-matches in cases of nested
2411    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2412    
2413    
2414    Version 4.0 17-Feb-03
2415    ---------------------
2416    
2417    1. If a comment in an extended regex that started immediately after a meta-item
2418    extended to the end of string, PCRE compiled incorrect data. This could lead to
2419    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2420    
2421    2. Moved to autoconf 2.53 and libtool 1.4.2.
2422    
2423    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2424    the special perltest8 script is no longer needed - all the tests can be run
2425    from a single perltest script.
2426    
2427    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2428    by \s. It has now been removed in PCRE. This means it isn't recognized as
2429    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2430    class [:space:] *does* include VT, thereby creating a mess.
2431    
2432    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2433    space and tab.
2434    
2435    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2436    its new features into the main test script, reducing the number of scripts.
2437    
2438    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2439    were backward compatible, and made the (?i) apply to the whole pattern, as if
2440    /i were given. Now it behaves more logically, and applies the option setting
2441    only to what follows. PCRE has been changed to follow suit. However, if it
2442    finds options settings right at the start of the pattern, it extracts them into
2443    the global options, as before. Thus, they show up in the info data.
2444    
2445    8. Added support for the \Q...\E escape sequence. Characters in between are
2446    treated as literals. This is slightly different from Perl in that $ and @ are
2447    also handled as literals inside the quotes. In Perl, they will cause variable
2448    interpolation. Note the following examples:
2449    
2450        Pattern            PCRE matches      Perl matches
2451    
2452        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2453        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2454        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2455    
2456    For compatibility with Perl, \Q...\E sequences are recognized inside character
2457    classes as well as outside them.
2458    
2459    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2460    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2461    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2462    signed/unsigned warnings.
2463    
2464    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2465    option for pcretest, so I've replaced it by a simple function that does just
2466    that job.
2467    
2468    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2469    "pcregrep -".
2470    
2471    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2472    Java package. This provides some syntactic sugar for simple cases of what my
2473    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2474    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2475    item, you can use this simplified notation. Note that only makes sense with
2476    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2477    greediness, whatever the setting of the PCRE_UNGREEDY option.
2478    
2479    13. A change of greediness default within a pattern was not taking effect at
2480    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2481    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2482    was abstracted outside.
2483    
2484    14. PCRE now supports the \G assertion. It is true when the current matching
2485    position is at the start point of the match. This differs from \A when the
2486    starting offset is non-zero. Used with the /g option of pcretest (or similar
2487    code), it works in the same way as it does for Perl's /g option. If all
2488    alternatives of a regex begin with \G, the expression is anchored to the start
2489    match position, and the "anchored" flag is set in the compiled expression.
2490    
2491    15. Some bugs concerning the handling of certain option changes within patterns
2492    have been fixed. These applied to options other than (?ims). For example,
2493    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2494    been the other way round. Some of this was related to change 7 above.
2495    
2496    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2497    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2498    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2499    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2500    
2501    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2502    mode. Unlike ".", it always matches newline, whatever the setting of
2503    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2504    assertions. Perl allows it, but it doesn't (in general) work because it can't
2505    calculate the length of the lookbehind. At least, that's the case for Perl
2506    5.8.0 - I've been told they are going to document that it doesn't work in
2507    future.
2508    
2509    18. Added an error diagnosis for escapes that PCRE does not support: these are
2510    \L, \l, \N, \P, \p, \U, \u, and \X.
2511    
2512    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2513    reading past the end of the pattern in cases such as /[abcd/.
2514    
2515    20. PCRE was getting more memory than necessary for patterns with classes that
2516    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2517    
2518    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2519    compiling PCRE for use with Virtual Pascal.
2520    
2521    22. Small fix to the Makefile to make it work properly if the build is done
2522    outside the source tree.
2523    
2524    23. Added a new extension: a condition to go with recursion. If a conditional
2525    subpattern starts with (?(R) the "true" branch is used if recursion has
2526    happened, whereas the "false" branch is used only at the top level.
2527    
2528    24. When there was a very long string of literal characters (over 255 bytes
2529    without UTF support, over 250 bytes with UTF support), the computation of how
2530    much memory was required could be incorrect, leading to segfaults or other
2531    strange effects.
2532    
2533    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2534    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2535    there was a subsequent back reference to those brackets. This meant that, for
2536    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2537    possible to check for precisely this case. All we can do is abandon the
2538    optimization if .* occurs inside capturing brackets when there are any back
2539    references whatsoever. (See below for a better fix that came later.)
2540    
2541    26. The handling of the optimization for finding the first character of a
2542    non-anchored pattern, and for finding a character that is required later in the
2543    match were failing in some cases. This didn't break the matching; it just
2544    failed to optimize when it could. The way this is done has been re-implemented.
2545    
2546    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2547    
2548    28. Added a new feature that provides some of the functionality that Perl
2549    provides with (?{...}). The facility is termed a "callout". The way it is done
2550    in PCRE is for the caller to provide an optional function, by setting
2551    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2552    global variable. By default it is unset, which disables all calling out. To get
2553    the function called, the regex must include (?C) at appropriate points. This
2554    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2555    This provides a means of identifying different callout points. When PCRE
2556    reaches such a point in the regex, if pcre_callout has been set, the external
2557    function is called. It is provided with data in a structure called
2558    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2559    matching continues; if it returns a non-zero value, the match at the current
2560    point fails. However, backtracking will occur if possible. [This was changed
2561    later and other features added - see item 49 below.]
2562    
2563    29. pcretest is upgraded to test the callout functionality. It provides a
2564    callout function that displays information. By default, it shows the start of
2565    the match and the current position in the text. There are some new data escapes
2566    to vary what happens:
2567    
2568        \C+         in addition, show current contents of captured substrings
2569        \C-         do not supply a callout function
2570        \C!n        return 1 when callout number n is reached
2571        \C!n!m      return 1 when callout number n is reached for the mth time
2572    
2573    30. If pcregrep was called with the -l option and just a single file name, it
2574    output "<stdin>" if a match was found, instead of the file name.
2575    
2576    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2577    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2578    pcre_exec(). This saves a malloc/free per call. The default value of
2579    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2580    when configuring.
2581    
2582    32. The default maximum size of a compiled pattern is 64K. There have been a
2583    few cases of people hitting this limit. The code now uses macros to handle the
2584    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2585    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2586    configuring. Tests 2 and 5 work only with 2-byte links because they output
2587    debugging information about compiled patterns.
2588    
2589    33. Internal code re-arrangements:
2590    
2591    (a) Moved the debugging function for printing out a compiled regex into
2592        its own source file (printint.c) and used #include to pull it into
2593        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2594        separate copies.
2595    
2596    (b) Defined the list of op-code names for debugging as a macro in
2597        internal.h so that it is next to the definition of the opcodes.
2598    
2599    (c) Defined a table of op-code lengths for simpler skipping along compiled
2600        code. This is again a macro in internal.h so that it is next to the
2601        definition of the opcodes.
2602    
2603    34. Added support for recursive calls to individual subpatterns, along the
2604    lines of Robin Houston's patch (but implemented somewhat differently).
2605    
2606    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2607    allow it to read and process whole directories in Win32. This code was
2608    contributed by Lionel Fourquaux; it has not been tested by me.
2609    
2610    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2611    used to name a group. Names consist of alphanumerics and underscores, and must
2612    be unique. Back references use the syntax (?P=name) and recursive calls use
2613    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2614    numbers. The function pcre_fullinfo() can be used after compilation to extract
2615    a name/number map. There are three relevant calls:
2616    
2617      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2618      PCRE_INFO_NAMECOUNT            yields the number of entries
2619      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2620    
2621    The map is a vector of fixed-size entries. The size of each entry depends on
2622    the length of the longest name used. The first two bytes of each entry are the
2623    group number, most significant byte first. There follows the corresponding
2624    name, zero terminated. The names are in alphabetical order.
2625    
2626    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2627    case instead of 255. Making it the same both with and without UTF-8 support
2628    means that the same test output works with both.
2629    
2630    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2631    calling malloc() with a zero argument.
2632    
2633    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2634    optimization. I've improved things by keeping a bitmap of backreferences with
2635    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2636    fact referenced, the optimization can be applied. It is unlikely that a
2637    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2638    the match to follow \n) will appear inside brackets with a number greater than
2639    31, but if it does, any back reference > 31 suppresses the optimization.
2640    
2641    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2642    of disabling numbered capturing parentheses. Any opening parenthesis that is
2643    not followed by ? behaves as if it were followed by ?: but named parentheses
2644    can still be used for capturing (and they will acquire numbers in the usual
2645    way).
2646    
2647    41. Redesigned the return codes from the match() function into yes/no/error so
2648    that errors can be passed back from deep inside the nested calls. A malloc
2649    failure while inside a recursive subpattern call now causes the
2650    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2651    
2652    42. It is now possible to set a limit on the number of times the match()
2653    function is called in a call to pcre_exec(). This facility makes it possible to
2654    limit the amount of recursion and backtracking, though not in a directly
2655    obvious way, because the match() function is used in a number of different
2656    circumstances. The count starts from zero for each position in the subject
2657    string (for non-anchored patterns). The default limit is, for compatibility, a
2658    large number, namely 10 000 000. You can change this in two ways:
2659    
2660    (a) When configuring PCRE before making, you can use --with-match-limit=n
2661        to set a default value for the compiled library.
2662    
2663    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2664        a different value is set. See 45 below.
2665    
2666    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2667    
2668    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2669    of things that can be changed at compile time. The first argument specifies
2670    what is wanted and the second points to where the information is to be placed.
2671    The current list of available information is:
2672    
2673      PCRE_CONFIG_UTF8
2674    
2675    The output is an integer that is set to one if UTF-8 support is available;
2676    otherwise it is set to zero.
2677    
2678      PCRE_CONFIG_NEWLINE
2679    
2680    The output is an integer that it set to the value of the code that is used for
2681    newline. It is either LF (10) or CR (13).
2682    
2683      PCRE_CONFIG_LINK_SIZE
2684    
2685    The output is an integer that contains the number of bytes used for internal
2686    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2687    
2688      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2689    
2690    The output is an integer that contains the threshold above which the POSIX
2691    interface uses malloc() for output vectors. See item 31 above.
2692    
2693      PCRE_CONFIG_MATCH_LIMIT
2694    
2695    The output is an unsigned integer that contains the default limit of the number
2696    of match() calls in a pcre_exec() execution. See 42 above.
2697    
2698    44. pcretest has been upgraded by the addition of the -C option. This causes it
2699    to extract all the available output from the new pcre_config() function, and to
2700    output it. The program then exits immediately.
2701    
2702    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2703    order to support additional features. One way would have been to define
2704    pcre_exec2() (for example) with extra arguments, but this would not have been
2705    extensible, and would also have required all calls to the original function to
2706    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2707    is used for passing in "extra" data from pcre_study().
2708    
2709    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2710    contains the following fields:
2711    
2712      flags         a bitmap indicating which of the following fields are set
2713      study_data    opaque data from pcre_study()
2714      match_limit   a way of specifying a limit on match() calls for a specific
2715                      call to pcre_exec()
2716      callout_data  data for callouts (see 49 below)
2717    
2718    The flag bits are also defined in pcre.h, and are
2719    
2720      PCRE_EXTRA_STUDY_DATA
2721      PCRE_EXTRA_MATCH_LIMIT
2722      PCRE_EXTRA_CALLOUT_DATA
2723    
2724    The pcre_study() function now returns one of these new pcre_extra blocks, with
2725    the actual study data pointed to by the study_data field, and the
2726    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2727    before. That is, this change is entirely upwards-compatible and requires no
2728    change to existing code.
2729    
2730    If you want to pass in additional data to pcre_exec(), you can either place it
2731    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2732    block.
2733    
2734    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2735    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2736    times with different match limits, until it finds the minimum value needed for
2737    pcre_exec() to complete. The value is then output. This can be instructive; for
2738    most simple matches the number is quite small, but for pathological cases it
2739    gets very large very quickly.
2740    
2741    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2742    returns the size of the data block pointed to by the study_data field in a
2743    pcre_extra block, that is, the value that was passed as the argument to
2744    pcre_malloc() when PCRE was getting memory in which to place the information
2745    created by pcre_study(). The fourth argument should point to a size_t variable.
2746    pcretest has been extended so that this information is shown after a successful
2747    pcre_study() call when information about the compiled regex is being displayed.
2748    
2749    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2750    because what follows is always an absolute path. (Later: it turns out that this
2751    is more than cosmetic for MinGW, because it doesn't like empty path
2752    components.)
2753    
2754    49. Some changes have been made to the callout feature (see 28 above):
2755    
2756    (i)  A callout function now has three choices for what it returns:
2757    
2758           0  =>  success, carry on matching
2759         > 0  =>  failure at this point, but backtrack if possible
2760         < 0  =>  serious error, return this value from pcre_exec()
2761    
2762         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2763         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2764         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2765         use by callout functions. It will never be used by PCRE itself.
2766    
2767    (ii) The pcre_extra structure (see 45 above) has a void * field called
2768         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2769         pcre_callout_block structure has a field of the same name. The contents of
2770         the field passed in the pcre_extra structure are passed to the callout
2771         function in the corresponding field in the callout block. This makes it
2772         easier to use the same callout-containing regex from multiple threads. For
2773         testing, the pcretest program has a new data escape
2774    
2775           \C*n        pass the number n (may be negative) as callout_data
2776    
2777         If the callout function in pcretest receives a non-zero value as
2778         callout_data, it returns that value.
2779    
2780    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2781    there were some redundant $(CFLAGS) in commands that are now specified as
2782    $(LINK), which already includes $(CFLAGS).
2783    
2784    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2785    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2786    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2787    one-byte characters throughout. Note that case-insensitive matching applies
2788    only to characters whose values are less than 256. PCRE doesn't support the
2789    notion of cases for higher-valued characters.
2790    
2791    (i)   A character class whose characters are all within 0-255 is handled as
2792          a bit map, and the map is inverted for negative classes. Previously, a
2793          character > 255 always failed to match such a class; however it should
2794          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2795    
2796    (ii)  A negated character class with a single character < 255 is coded as
2797          "not this character" (OP_NOT). This wasn't working properly when the test
2798          character was multibyte, either singly or repeated.
2799    
2800    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2801          mode, for example: \x{100}{2,3}.
2802    
2803    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2804          singly or repeated) now correctly test multibyte characters. However,
2805          PCRE doesn't recognize any characters with values greater than 255 as
2806          digits, spaces, or word characters. Such characters always match \D, \S,
2807          and \W, and never match \d, \s, or \w.
2808    
2809    (v)   Classes may now contain characters and character ranges with values
2810          greater than 255. For example: [ab\x{100}-\x{400}].
2811    
2812    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2813          PCRE in UTF-8 mode.
2814    
2815    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2816    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2817    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2818    value.)
2819    
2820    53. The single man page has become too large. I have therefore split it up into
2821    a number of separate man pages. These also give rise to individual HTML pages;
2822    these are now put in a separate directory, and there is an index.html page that
2823    lists them all. Some hyperlinking between the pages has been installed.
2824    
2825    54. Added convenience functions for handling named capturing parentheses.
2826    
2827    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2828    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2829    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2830    are faulted.
2831    
2832    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2833    calling configure. These values are used when compiling the dftables.c program
2834    which is run to generate the source of the default character tables. They
2835    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2836    you will need to set these values.
2837    
2838    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2839    
2840    
2841    Version 3.9 02-Jan-02
2842    ---------------------
2843    
2844    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2845    
2846    2. If --disable-static was given, the building process failed when trying to
2847    build pcretest and pcregrep. (For some reason it was using libtool to compile
2848    them, which is not right, as they aren't part of the library.)
2849    
2850    
2851    Version 3.8 18-Dec-01
2852    ---------------------
2853    
2854    1. The experimental UTF-8 code was completely screwed up. It was packing the
2855    bytes in the wrong order. How dumb can you get?
2856    
2857    
2858    Version 3.7 29-Oct-01
2859    ---------------------
2860    
2861    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2862    This caused pcretest, when used on the test data, to segfault. Unfortunately,
2863    this didn't happen under Solaris 8, where I normally test things.
2864    
2865    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2866    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2867    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2868    this fix an hour or so after the initial 3.7 release.)
2869    
2870    
2871    Version 3.6 23-Oct-01
2872    ---------------------
2873    
2874    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2875    offsets passed as NULL with zero offset count.
2876    
2877    2. The config.guess and config.sub files had not been updated when I moved to
2878    the latest autoconf.
2879    
2880    
2881    Version 3.5 15-Aug-01
2882    ---------------------
2883    
2884    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2885    had been forgotten.
2886    
2887    2. By using declared but undefined structures, we can avoid using "void"
2888    definitions in pcre.h while keeping the internal definitions of the structures
2889    private.
2890    
2891    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2892    user point of view, this means that both static and shared libraries are built
2893    by default, but this can be individually controlled. More of the work of
2894    handling this static/shared cases is now inside libtool instead of PCRE's make
2895    file.
2896    
2897    4. The pcretest utility is now installed along with pcregrep because it is
2898    useful for users (to test regexs) and by doing this, it automatically gets
2899    relinked by libtool. The documentation has been turned into a man page, so
2900    there are now .1, .txt, and .html versions in /doc.
2901    
2902    5. Upgrades to pcregrep:
2903       (i)   Added long-form option names like gnu grep.
2904       (ii)  Added --help to list all options with an explanatory phrase.
2905       (iii) Added -r, --recursive to recurse into sub-directories.
2906       (iv)  Added -f, --file to read patterns from a file.
2907    
2908    6. pcre_exec() was referring to its "code" argument before testing that
2909    argument for NULL (and giving an error if it was NULL).
2910    
2911    7. Upgraded Makefile.in to allow for compiling in a different directory from
2912    the source directory.
2913    
2914    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2915    options bits, the pointer it was passed was to an int instead of to an unsigned
2916    long int. This mattered only on 64-bit systems.
2917    
2918    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2919    generated) instead of pcre.in, which it its source. Also made the same change
2920    in several of the .c files.
2921    
2922    10. A new release of gcc defines printf() as a macro, which broke pcretest
2923    because it had an ifdef in the middle of a string argument for printf(). Fixed
2924    by using separate calls to printf().
2925    
2926    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2927    script, to force use of CR or LF instead of \n in the source. On non-Unix
2928    systems, the value can be set in config.h.
2929    
2930    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2931    absolute limit. Changed the text of the error message to make this clear, and
2932    likewise updated the man page.
2933    
2934    13. The limit of 99 on the number of capturing subpatterns has been removed.
2935    The new limit is 65535, which I hope will not be a "real" limit.
2936    
2937    
2938    Version 3.4 22-Aug-00
2939    ---------------------
2940    
2941    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2942    
2943    2. Diagnose condition (?(0) as an error instead of crashing on matching.
2944    
2945    
2946    Version 3.3 01-Aug-00
2947    ---------------------
2948    
2949    1. If an octal character was given, but the value was greater than \377, it
2950    was not getting masked to the least significant bits, as documented. This could
2951    lead to crashes in some systems.
2952    
2953    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2954    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2955    
2956    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2957    These just pass their arguments on to (pcre_free)(), but they are provided
2958    because some uses of PCRE bind it to non-C systems that can call its functions,
2959    but cannot call free() or pcre_free() directly.
2960    
2961    4. Add "make test" as a synonym for "make check". Corrected some comments in
2962    the Makefile.
2963    
2964    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2965    Makefile.
2966    
2967    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2968    command called pgrep for grepping around the active processes.
2969    
2970    7. Added the beginnings of support for UTF-8 character strings.
2971    
2972    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2973    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2974    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2975    out for the ar command.)
2976    
2977    
2978    Version 3.2 12-May-00
2979    ---------------------
2980    
2981    This is purely a bug fixing release.
2982    
2983    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2984    of ZA. This was just one example of several cases that could provoke this bug,
2985    which was introduced by change 9 of version 2.00. The code for breaking
2986    infinite loops after an iteration that matches an empty string was't working
2987    correctly.
2988    
2989    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2990    when matched against abbab (for example). After matching an empty string, it
2991    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2992    caused it to match further down the string than it should.
2993    
2994    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2995    was there because it doesn't seem to be needed, and it causes trouble on some
2996    systems, as it is not a Standard C header. It has been removed.
2997    
2998    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2999    were reported on the Macintosh. The changes were from
3000    
3001      while ((c = *(++ptr)) != 0 && c != '\n');
3002    to
3003      while ((c = *(++ptr)) != 0 && c != '\n') ;
3004    
3005    Totally extraordinary, but if that's what it takes...
3006    
3007    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3008    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3009    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3010    assumes the way PCRE uses memmove() (always moving upwards).
3011    
3012    6. PCRE is being used in one environment where strchr() is not available. There
3013    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3014    faster code anyway.
3015    
3016    
3017    Version 3.1 09-Feb-00
3018    ---------------------
3019    
3020    The only change in this release is the fixing of some bugs in Makefile.in for
3021    the "install" target:
3022    
3023    (1) It was failing to install pcreposix.h.
3024    
3025    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3026    
3027    
3028  Version 3.0 01-Feb-00  Version 3.0 01-Feb-00
3029  ---------------------  ---------------------

Legend:
Removed from v.43  
changed lines
  Added in v.430

  ViewVC Help
Powered by ViewVC 1.1.5